television - American Radio History
TELEVISION
Volume IV
(1942
-
1946)
Edited by
ALFRED N. GOLDSMITH
ARTHUR F. VAN DYCK
ROBERT S. BURNAP
EDWARD T. DICKEY
GEORGE M. K. BAKER
JANUARY,
1947
Published by
RCA REVIEW
RADIO CORPORATION OF AMERICA
RCA LABORATORIES DIVISION
Princeton, New Jersey
www.americanradiohistory.com
Copyright, 1947, by
Radio Corporation of America,
RCA Laboratories Division
Printed in U.B.A.
www.americanradiohistory.com
Mickey Mouse (60 -line definition)
View of New York (525 -line definition)
COMPARATIVE CATHODE -RAY TUBE PHOTOS OF
TELEVISION IMAGES
www.americanradiohistory.com
.
www.americanradiohistory.com
TELEVISION, Vol. IV
PREFACE
TELEVISION, Vol. IV, covering the period 1942 -1946, is the fourth
volume on television in the RCA REVIEW Technical Book Series. The
first volume was published in 1936 followed by TELEVISION, Vol. II
in 1937. TELEVISION, Vol. III, long -delayed by wartime security
restrictions, appeared in early 1947.
The large number of excellent papers on the subject of television
has made necessary a very stringent selection process. All the available
material can not be included in full form. A number of papers are,
therefore, presented herein in summary form only; it has been necessary to omit others entirely. Suitably balanced presentation of the
various phases of television was the major criterion in deciding which
papers to include in full and which in summary. The presentation of a
paper in summary form (or the non-inclusion of any particular paper)
is not intended to indicate any deficiency in technical accuracy, literary
merit, or impórtance.
*
*
*
The papers in this volume are presented in six sections pickup,
transmission, reception, color television, military television, and
general. As a source of reference material, the Appendix to this book
includes a television bibliography covering the period 1929 -1946.
:
*
*
*
RCA REVIEW gratefully acknowledges the courtesy of The Institute of Radio Engineers, the McGraw -Hill Book Company, and the
Society of Motion Picture Engineers in granting to RCA REVIEW
permission to republish material which has appeared in Their publica-
tions.
The appreciation of RCA REVIEW is extended to all authors whose
papers appear herein, and particularly to those whose papers are being
printed in this book without prior publication.
*
*
*
TELEVISION, Vol. IV, like its predecessors, is being published
for scientists, engineers and others interested in television, with the
sincere hope that the material here assembled may help to speed developments and advance the position of television among companion arts
and services.
RCA Laboratories
Princeton, N. J.
December 31, 1946
The Manager, RCA
www.americanradiohistory.com
REVIEW
www.americanradiohistory.com
TELEVISION
Volume IV
(1942-1946)
CONTENTS
FRONTISPIECE
PREFACE
INTRODUCTION
The Manager, RCA REVIEW
DAVID SARNOFF
PAGE
-V-
-
-vii
-xi-
PICKUP
Contemporary Problems in Television Sound
C L. TOWNSEND
The Focusing View -finder Problem in Television Cameras. G. L. BEERS
Electron Bombardment in Television Tubes
I. G. MALOFF
Image Orthicon Camera
R D. KELL AND G. C. SZIKLAI
Field Television
R E. SHELBY AND H. P. SEE
The Image Orthicon -A Sensitive Television Pickup Tube
A ROSE, P. K. WEIMER, AND H. B. LAW
A Unified Approach to the Performance of Photographic Film, Television Pickup Tubes, and the Human Eye
A ROSE
1
12
28
43
53
70
90
TRANSMISSION
Analysis, Synthesis and Evaluation of Transient Response in Television Apparatus
A V. BEDFORD AND G. L. FREDENDALL
112
G L. FREDENDALL, K. SCHLESINGER, AND A. C. SCHROEDER
A Method of Measuring the Degree of Modulation of a Television
Signal
T J. BUZALSKI
147
Transmission of Television Sound on the Picture Carrier
177
RECEPTION
Factors Governing the Performance of Electron Guns in Television
Cathode -Ray Tubes
R R. LAW 184
Television Reception with Built -in Antennas for Horizontally and Vertically Polarized Waves
W. L. CARLSON 191
Automatic Frequency and Phase Control of Synchronization in Television Receivers
K. R. WENDT AND G. L. FREDENDALL 203
Radio -Frequency -Operated High -Voltage Supplies for Cathode -Ray
Tubes
O H. SCHADE 221
A Type of Light Valve for Television Reproduction
J
S. DONAL, JR., AND D. B. LANGMUIR
235
W. EPSTEIN
250
C. SZIKLAI AND A. C. SCHROEDER
270
D W. EPSTEIN AND L. PENSAK
290
Reflective Optics in Projection Television
I. G. MALOFF
Cathode -Coupled Wide -Band Amplifiers
G
AND D.
Improved Cathode-Ray Tubes with Metal-Backed Luminescent Screens
- ix
www.americanradiohistory.com
CONTENTS (Continued)
Local Oscillator Radiation and Its Effect on Television Picture Contrast
E W. HEROLD
Development of an Ultra Low Loss Transmission Line for Television
E O. JOHNSON
PAGE
296
318
COLOR TELEVISION
An Experimental Color Television System
-
R. D. KELL, G. L. FREDENDALL, A. C. SCHROEDER AND R. C. WEBB
Simultaneous All- Electronic Color Television
DIVISION
RCA
LABORATORIES
327
341
MILITARY TELEVISION
Military Television
GEORGE M. K.
Introduction to Technical Papers on Airborne Television.
BAKER
351
DAVID SARNOFF
357
359
Flying Torpedo with an Electric Eye
V K. ZWORYKIN
Naval Airborne Television Reconnaissance System
R E. SHELBY, F. J. SOMERS, AND L. R. MOFFETT
Miniature Airborne Television Equipment
R D. KELL AND G. C. SZIKLAI
MIMO-Miniature Image Orthicon
P K. WEIMER, H. B. LAW
Television Equipment for
Aircraft... M.
AND S. V. FORGUE
A. TRAINER AND W. J. POCH
369
404
424
433
GENERAL
Television -A Review, 1946
Television Broadcasting-1946
Television Today and Its Problems -1946
E W. ENGSTROM
0 B. HANSON
A N. GOLDSMITH
467
482
486
SUM MARIES
Measurement of the Slope and Duration of Television Synchronizing
Impulses
R A. MONFORT AND F. J. SOMERS 493
The Relative Sensitivities of Television Pickup Tubes, Photographic
Film, and the Human Eye
A ROSE 493
A Portable High- Frequency Square -Wave Oscillograph for Television
R D. KELL, A. V. BEDFORD, AND H. N. KOZANOWSKI
Cathode -Ray Control of Television Light Valves
J S. DONAL, JR.
A Reflective Optical System for Television
E W. WILBY
Projection Television
D W. EPSTEIN AND I. G. MALOFF
Band -Pass Bridged -T Network for Television Intermediate -Frequency
Amplifiers
G C. SZIKLAI AND A. C. SCHROEDER
Input Impedance of Several Receiving -Type Pentodes at FM and Television Frequencies
F MURAL
Television High -Voltage R -F Supplies
R S. MAUTNER AND O. H. SCHADE
494
495
495
496
497
497
498
APPENDIX
TELEVISION -A Bibliography of Technical Papers by RCA Authors,
1929 -1946
499
-X
www.americanradiohistory.com
INTRODUCTION
BY
DAVID SARNOFF
President, Radio
Co
poi at ion of America
INCE the publication of earlier volumes on television in this
RCA REVIEW Technical Book Series, the art has made notable
advances. It has moved ahead from its status as a promising
experiment to its present role as a proved medium of entertainment,
education and news. Aided by research accelerated by the demands of
war, the development of television has reached a point that otherwise
might not have been 'expected at least until 1950. New tubes and new
circuits have greatly improved transmitters and receivers and the
growing interest currently being exhibited in television by the public
is proof that television is being accepted rapidly as a new art and a
promising new service.
For more than 20 years I have repeatedly expressed my confidence
in the future of television. I have not hesitated to forecast that in
due time it will become a greater industry and a greater art than sound
broadcasting. And I believe that during this expansion of television,
broadcasting will continue as a great industry, for the two forms of
communication will supplement each other. There are natural reasons
why sight and sound should be united to form a combination that will
be far more effective than either medium alone, in imparting information, entertainment, culture and understanding of our life and our
Government to all classes.
Television will not reach its full stature overnight. Well- founded
scientific developments do not progress in that manner. History has
proved that it takes about five years for any cycle in radio to translate
itself into practical reality. The vacuum tube did not immediately
supersede the spark in international communications nor did the
superheterodyne become an overnight successor to the crystal receiver.
After research has shown the way, a multitude of problems arise in
the development of suitable programs and merchandising of the product that must be solved before the new service becomes universally
acceptable. Television does not differ from other technical inventions
in this respect.
An estimate at this time of the ultimate effect of television and the
social and scientific consequences which will flow from its introduction
--
-
www.americanradiohistory.com
would be impractical. But we do know from our experience that inventions which give us new powers over natural forces have had farreaching effects on the human race. I need only mention Watts and his
steam engine; Nobel and gunpowder ; Morse's telegraph ; Bell's tele-
phone; Marconi's wireless and the Wright Brothers' aeroplane to support this view. These brain -children have produced and still are producing far -reaching effects on the family, on government, education,
industrial production, yes, even on the habits and the beliefs of people.
Moreover, as history shows, there is a cumulative gain as one development leads to another. This is exemplified in the case of the gas engine
and jet propulsion, the single engined biplane and the stratoliner,
and now in radio and television.
Because of these derivative results of inventions, the full social
effects of a development such as television should be weighed on a
general basis only. For instance, sociologists have called attention to
the growing decentralization of industry. There' are good reasons to
believe that television may hasten this trend by simplifying the remote
control of industrial operations, and by expanding the entertainment
and cultural advantages which the video art will make available in
small communities. As leisure time is increased along with greater
technological efficiency, it is likely that television will be depended upon
to make our hours of ease more profitable and satisfying.
Such a possibility creates the picture of a population which increasingly may center its interest more and more within the home. In such
a setting, television will become a vital element in daily living. To
people with more conveniences and with more time in which to enjoy
them television may well become their principal source of entertainment, education and news.
Television, too, may enable man to keep pace with his thoughts.
The human being has been created with a mind that can encompass the
whole world in the fraction of a second; yet his physical senses lag
woefully behind. With his feet he can walk only a limited distance;
with his hands he can touch only what is within reach. With his
unaided eyes he cannot see beyond the horizon, and his ears are useful
only at short range.
These natural restrictions will vanish with the spread of television.
In the years of its fulfilled destiny, television will bring the boundaries
of the earth itself within the useful limits of Man's several senses.
When this day arrives, there may come also a new philosophy and a
finer, broader understanding between all peoples, whatever their
nationalities or wherever they may live. If the airplane is named as
www.americanradiohistory.com
the invention which annihilates distance, television can be said to annihilate both distance and time.
As this is written, television promises to make big strides in taking
its place alongside the older arts, in many instances giving them new
and modern import. Science has equipped the new industry for its
expanded role. New supersensitive cameras, that make it possible to
pick up scenes under low- lighting conditions; improved kinescope picture tubes that provide the viewer with steadier, brighter and clearer
images on his receiver ; mobile truck units which can speed to remote
points and transmit pictures back to the main television transmitter
for rebroadcasting; and new antennas that increase the area covered
by television stations. To these advances, add the more skillful studio
technique developed by broadcasters and a well- rounded, solid foundation for the new art is assured.
Leaders in education are becoming increasingly interested in television as an invaluable supplement to present school curricula. Through
the medium of television, the skill and knowledge of the best teachers
in the land can easily raise the educational level of the "little red
schoolhouse" to the standing of leading schools in the larger cities.
In religion, also, the impact of television will be powerful and
effective. Nationwide television will bring the services of the great
cathedrals into the homes of the most remote residents. Viewers not
only will hear the minister's words and the music but will see the
preacher face to face as he delivers his sermon, and observe as though
present, the solemn ceremonies at the altar.
Any discussion of television, however brief, should include some
mention of its by- products. Some of the fields in which television
devices may bring about important advances are in marine and aerial
navigation, by permitting vision in fog and darkness through the use
of infrared rays; in metallurgical, chemical, physical and biological
research in certain manufacturing processes as substitutes for human
vision or for control purposes; in department stores for advertising
and display; and in numerous other fields where a substitute for the
eye may be useful.
The new flickerless, all -electronic color television system demonstrated at RCA Laboratories on October 30, 1946, and again on January 29, 1947, is the most recent positive contribution to the television
leadership of our country. In the near future it is expected that outdoor scenes will be televised in color followed in 1948 by electronic
color television on large -size theater screens.
The realization of this universal system of all- electronic color television, accomplished without the outmoded rotating disk or any other
;
www.americanradiohistory.com
moving part, is as far reaching as was the creation of the RCA all electronic television system which supplanted the mechanical disks used
in black- and -white television when it first began.
With interest in television increasing more rapidly now than at any
time since its introduction on a commercial basis in 1939, it is fitting
that this Volume should be issued to continue the permanent history of
research and achievement inaugurated by the preceding Volumes in
the series. Reference to these records will provide those interested in
this subject with useful information.
www.americanradiohistory.com
CONTEMPORARY PROBLEMS IN
TELEVISION SOUND *t
BY
C. L. TOWNSEND
National Broadcasting Company, Inc..
New York, N. Y.
Summary -The present rapid development of television is introducing
new problems in sound pickup and operation. As the art progresses,
engineering tools and methods must not only keep pace with, but generally
anticipate, the needs of the program- producing staff in the production of
more and more intricate material. The nature of the acoustic problems so
raised, and their solutions, are treated in this paper. New tools necessary to
proper operation and the methods of their employment are discussed. For
a better understanding of television requirements, the methods normally
employed in motion pictures and standard radio broadcasting are compared
with those in use in the present television studio. Some indications as to
what may be required in the near future are discussed and possible developments suitable for such use are described.
THE history of every new activity, problems and concepts
peculiar to itself arise. Certainly television is no exception to this
rule nor is that part of television which we are to consider. There
may have been many who felt in the earlier days of the art that television's sound accompaniment could well be expected to care for itself,
for much had been done to perfect a technique of .sound pickup with
action in progress in the motion -picture studios of Hollywood. But
very shortly, marked departures from the accepted methods were found
desirable, and gradually it became clear that good television sound
required not only different treatment but also different tools than were
used at first. As the show -producing workers in television become
familiar with their picture -making equipment, more and more is being
demanded of it, and the sound accompaniment must keep pace. No
consideration of the sound portion of a problem arising in a television
studio is permitted to interfere with the picture technique, since the
production staff has come to rely upon the sound engineer to find a
way around his difficulties. This paper discusses these difficulties, and
considers what may be done to overcome them.
A consideration of the mechanics of television studio operation will
disclose some of the problems arising in sound pickup associated with
N
*
Decimal classification: R583.
from Proc. I.R.E., January, 1943.
t Reprinted
1
www.americanradiohistory.com
TELEVISION, Volume IV
2
visual programs. The National Broadcasting Company's studio is
equipped at present with three television cameras and normal set
lighting requires the use of four floor broads of about 3 kilowatts each.
All of this equipment must be positioned for best advantage as to
camera angles and lighting effects. If no sound equipment were used
at all, the portion of the studio in use would be crowded enough, but
it is necessary for the microphone boom to find a place also. The boom
operator chooses his position with regard not only to his own best
sound requirements but also considers the possible camera movements.
If it is likely that a camera -dolly movement will find him in its way, he
must be able to move the base of the boom sufficiently in advance of
the dolly to clear the necessary space. Thus, the boom operator must
Fig.
-For
good pictures, television cameras require most of the
space. Sound equipment must operate in what remains.
1
not only follow closely the action on the set but must also bear in mind
the exact pattern of off -stage activities. The present operators have
become adept at maintaining the position of the microphone correctly
above the heads of the persons on picture, while stepping from the
boom platform and moving it bodily a sufficient distance to permit
passage of a camera. Often, too, only a few seconds can be allowed for
a complete change from one set to another, necessitating accurate
planning of movements and precise co- operation between sound- and
sight- equipment personnel. To aid in this the boom used is as small as
is presently practicable, having a maximum extension of 14 feet and
being about 4 feet wide across the base. A unidirectional microphone
www.americanradiohistory.com
TELEVISION SOUND
3
is used to aid in reducing off -set noise, but this adds to the precision
necessary in operation, for if close -ups are being used, the microphone
must be aimed at the person being televised. This means that the boom
operator must watch the camera -switching lights and position the
microphone to suit the camera as well as the actor, being careful to
discriminate against off -camera sounds only.
To facilitate scene transitions, or to provide a second pickup in a
set where two widely spaced sound sources act concurrently, a method
of hanging microphones has been devised. The studio ceiling carries
a network of pipes of approximately 21/2 inches in diameter. A special
clamp has been made to fit these pipes. Connected to each clamp is an
adjustable length of light conduit, designed to accept a standard
Fig. 2-Efficient utilization of floor space is a necessity in television.
microphone coupling. The clamp can be operated by twisting the conduit making it unnecessary to climb ladders to hang microphones ; this
greatly increases the all- important factor of speed.
Three types of microphones are normally used in the National
Broadcasting Company's television studio. The unidirectional type
with a cardioid pickup pattern is used for dialogue, mainly because of
its ability to reduce the effect of off -stage noise. Television, unavoidably, has rather more of this than is used on a motion -picture sound
stage, since following scenes must be prepared, equipment moved continuously to new locations, and the show kept running generally. This
contrasts markedly with the complete stopping of all other activity
when a scene is made in motion pictures. Regular velocity microphones
www.americanradiohistory.com
TELEVISION, Vodume IV
4
-A
3
transition from one scene to another may require the use of a.
fixed -position microphone for opening the new scene. Action will be
restricted until the arrival of the boom microphone.
Fig.
are used in cases requiring more reverberation, or when convenient to
use both sides for pickup. Usually this occurs when music is used on
the set, and an acoustically bright effect is desired. A pressure microphone is used when its nondirectional characteristic is advantageous.
The production staff at NBC recognizes that in recent years a micro-
Fig. 4- Musicians must be close to the set for good musical coordination,
introducing problems in balance and overlapping pickups. Unidirectional
microphones awl greatly in such situations.
www.americanradiohistory.com
TELEVISION SOUND
5
phone has become an integral part of some scenes. A supper -club set
may call for several microphones, and if a grouping dictated by picture
requirements is too wide for other types, a pressure microphone will
solve the problem. As these microphones are relatively small, they are
also most suitable for use in positions where they might tend to obscure
the picture, or when a microphone must be held in the hand.
Even with the above variety of tools, situations arise that defy
ordinary "on- the- spot" pickups. These cases generally can be classified
into those in which high scenes limit the possibility of bringing a
microphone close to the action, and those in which the action is too
fast or too complicated to permit its being followed by the micrphone
boom. Both occur usually in the musical production type of scene. It
may be that a large and decorative background has been erected for a
solo song, center stage and low. Obviously, no reasonable balance can
be obtained between voice and accompaniment if the microphone must
be far enough away to be out of the picture when it includes so large
a backdrop. In the second case, trouble usually is encountered when
performers not only sing but also move through a routine of action not
suited to sound pickup. This may include singing while facing away
from the camera, or while moving through a doorway, or perhaps next
to percussion instruments of an orchestra where maintenance of balance
would be impossible. All of these situations call for prerecording, a
technique developed in Hollywood and happily adaptable to television.
Two methods of procedure are available. In the first type mentioned
above, the microphone is located in a suitable position for the making
of the record, usually several hours before show time. The action is
carried out as usual and the timing of the record automatically fits the
scene as it will be broadcast. When the actual show takes place, a cue
from the production director will start the record and kill all sound
pickup in the studio. The record is then not only put on the air, but
also fed back into the studio, where the performers can hear it, and
synchronize their actions to it. When the recorded portion ends, the
studio microphones are opened and the show continues normally. In
the second type mentioned, the action is too detrimental to sound pickup
to permit recording with it in progress even though no picture is required. Hence the action is carefully timed and cues noted. The recording will then be made without action, the setup being entirely to suit
the sound situation. Such a record is then checked for synchronism on
another rehearsal, and used "on the air" as described. A lacquer disk
recording with the NBC Orthacoustic characteristic is used, resulting
in transitions from direct pickup to record and back again with prac-
www.americanradiohistory.com
.
6
TELEVISION, Volume IV
tically no noticeable change in sound quality. With such satisfactory
matching of sound quality available, prerecording is a very useful tool
in television.
Another angle of the studio-mechanics problem is in peculiar contradiction to the case in motion pictures. In some instances, the motion picture- making equipment causes some trouble through making noise
which may interfere with the desired sounds. In television the reverse
is true. Sound in the studio may be of such intensity and frequency
that it will cause spurious signals due to microphonics to appear in the
picture. These generally consist of horizontal bars across the picture,
and result from vibration of elements of vacuum tubes used in the video
preamplifier in the camera. It is necessary to treat the television
camera to keep sound out, rather than in. A heavy material, similar
to roofing felt may be cemented to the inside surfaces of the camera
housing to reduce sound transmission, and particularly to damp vibration occurring in the large plane sections of the present camera's sides
and top. Without such damping, these parts will vibrate very heavily
at their natural periods, making sound crossover almost a certainty.
With sufficient loading the tendency to vibrate disappears almost completely, permitting operation with any normal studio sound level.
The very nature of television is that its appeal must be in the intimate manner. As long as the present methods of picture reproduction
are being used extensively, this will continue to be the case, for picture
size and detail make best use of close -ups and penalize the extreme long
shots. The sound that accompanies these pictures should partake of the
same quality, heightening the tone and mood of a scene. The methods
adopted and the tools used must, then, be suitable for such work.
The National Broadcasting Company's live-talent studio is a room
30 x 50 x 17 feet. Its acoustic treatment differs radically from what a
motion-picture engineer might expect to find on a sound stage, in that
the reverberation constant is not as short as it could be made, but
rather a variable quantity, being in some cases as long as 11/4 seconds,
and in others as short as 1/2 second (over the essential range of frequencies). The reasons for this are close to the heart of the television
problem. In the usual sound -stage case, the studio is a large acoustically dead room, in which relatively permanent sets are erected. It is
normally the intention to permit those sets to exhibit their own characteristic reverberation without much, if any, artificial reinforcement.
The case in television is somewhat different. Our sets are designed for
rapid scene changes, and efficient use of personnel. They are made of
linen stretched on wood frames in the manner of legitimate stage
TELEVISION SOUND
7
scenery. Instead of adding a lifelike reverberation to the sound originating in them, such sets produce undesirable low- frequency resonance
effects, and add large amounts of high-frequency absorption in their
unpainted surfaces. If the studio itself were very "dead" these effects
would add detrimentally. Dialogue equalizers are used, which help to
avoid this trouble, but the less equalization that can be employed, the
better will be the average sound quality.
Studio acoustics also play an important part in television sound for
other reasons. The volume of the sets in use is always a very large
portion of the total studio volume, since many scenes must be set up at
once to provide a continuous performance. Under usual conditions
almost the whole studio is used in a show to run an hour and a half.
With so much absorption added in the sets, much of the original treatment of the studio must be removed to produce anything like normal
reverberation. Most television shows will also present music as well
as speech in the same studio, without a pause between the two portions
of the program. Such a case in motion -picture production would call
for the use of a scoring stage, or a set especially treated for music. In
television, the problem is attacked by making large sections of the
acoustic treatment on the studio walls movable. These panels can be
opened to expose a hard, reflective surface, increasing the reverberation
to an acceptable level. Should an outdoor scene be required, however,
all the absorbing panels would be closed, and equalization added to
produce an essentially reverberation -free pick -up.
It has often been remarked that television even now should use large
studios of the motion-picture sound -stage variety. There are however
certain mechanical and acoustical considerations that make this doubtful. Present television practice, which demands many close -ups and
rather restricted action during most of the show, means that even with
a relatively large set, for the major portion of a "take," the cameras,
lights, and sound equipment must be crowded together to serve best
the particular portion of it used at the moment. It is a provoking fact
that although most of the studio may be empty, the television equipment must be worked in close quarters. Consequently, additional room
would not materially increase the freedom of action of the cameras as
far as any one set is concerned. Mechanical considerations, then, indicate that the size of the studio is determined by the number of sets
which reasonably can be used on one show, or can be served by one
group of equipment. Under present production conditions, this would
result in a studio considerably smaller than the larger motion -picture
sound stages. Acoustically, the smaller studio is desirable, because of
www.americanradiohistory.com
8
TELEVISION, Volume IV
the requirement mentioned above that studio acoustics be adjustable
to compensate for set absorption. If the studio becomes too large, it
cannot contribute usefully to the over -all sound quality, for reverberation as a desirable enhancement is replaced by what is commonly called
room -slap, or echo. Hence, if a very large studio is to be used, it must
be very "dead," which inflexibility seriously limits its usefulness as an
acoustic tool. The answer to this problem seems to be that television
studios should be of a size between those used in radio broadcasting
and the large stages of Hollywood if all the mechanical and acoustical
advantages are to be realized.
The excellent work currently done in the broadcasting studios has
raised to a high stage of refinement the art of producing mood and
atmosphere with sound. Television must offer at least as much facility
for creation of these effects and at the same time must not limit in any
way the freedom of action necessary to good pictorial effect. Some of
the problems encountered in this blending of sound technique and sight
productions are worthy of consideration.
In the television studio, both close -up and long shots must be taken
at the same time. The accompanying sound must not only suit the
apparent distances shown in the picture, but may also be required to
produce an effect complementary to it. At times, perspective in television sound is so important that what would normally be only a
medium long shot can be made to seem very long, if the sound which
accompanies it carries sufficient reverberation to suggest great distance
to the mind of the listener. Since actual long shots are not usually
permitted for long periods of time, such an aid in producing the effect
of distance is a valuable tool. Close -ups, of course, require intimate
sound, and often the change from a distant view to a close -up occurs
too quickly to permit any actual change in microphone placement or
acoustic treatment, so the effect of a change must be produced electrically. Reverberation once added cannot be deleted ; consequently, the
pickup conditions must be set to produce close -up sound. It is then
possible to add reverberation through the use of the standard echochamber method. In the studio, close -up cameras are provided with
long -focal -length lenses, and long -shot cameras have either normal- or
short-focus lenses. Switching between the cameras actuates a set of
relays so connected that amounts of reverberation and volume level can
be adjusted to suit the lens of the camera in use. This is accomplished
by providing for each camera a separate volume control, and a separate
reverberation control. If a camera is to be used to take a long shot,
the volume control associated with it is turned down an amount calcu-
www.americanradiohistory.com
TELEVISION SOUND
9
lated to produce the proper psychological effect, and the reverberation
control is opened to accept a large portion of the output of the echo
chamber. Another camera having an intermediate focal length would
use more direct sound and less reverberation, while the close-up camera
would use full volume and no feed from the echo chamber at all. When
the technical director switches from one camera to another, the sound
is also switched from one set of controls to another, producing instantaneous changes in sound quality to suit the picture requirements. Of
course such artificial correction is confined in its application to interior
shots which would normally exhibit acoustical characteristics similar
to those available from the echo chamber. Corrections can be applied
Fig.
5-Making
close -ups and long shots simultaneously introduce
problems of acoustic realism.
to outdoor scenes by changing the volume level and low-frequency
response of the syst.em to match the camera switching. Thus an
exterior long shot would be accompanied by a reduction in volume -control setting and an increase of equalization designed to remove low and
high frequencies, thus simulating the conditions obtaining in nature.
Without such processing, the sound accompanying a television picture
would not only lose valuable contributing effects, but at times might
give an almost ludicrous effect, for the human eye and ear have been
tráined to expect a certain correlation between sight and sound perspective, and violations of their normal relationships are not acceptable.
Another problem peculiar to television sound is the result of a
demand for realism in its dynamic range. In motion pictures, the
www.americanradiohistory.com
10
TELEVISION, Volume IV
acceptable range of loudness is a strictly measurable and controllable
quantity. The lowest modulation permitted is a function of track hiss,
frequency response, audience noise, etc., and the highest sound output
is determined by the 100 per cent modulation point and reproducer
power. No one in the audience is able to change the volume reaching
him, nor does he expect to hear sound which is not dimensioned to fit
the picture on view. In radio broadcasting, almost the exact opposite
is true. With no visual program, the listener demands the maintenance
of a relatively constant level, and often writes to his station complaining that he has to adjust his receiver volume control during the progress of a show. Television encounters portions of both of these troubles,
and has had to evolve its own operating procedures to combat them.
Since television is broadcast, a reasonably high average modulation
should be maintained, in order that receiver noise levels may be low.
Maximum deviation is determined, of course, by channel width. Within
these two extremes must be confined sound to suit anything from the
scraping of a pen across paper, to the crashing thunder of a modern
blitzkreig. Such matching of sound and sight is necessary, for if the
eye sees what would in nature produce a loud sound, but the ear hears
only a small, muted version of what is expected, the mind rejects both
sight and sound as being counterfeit. Thus a dynamic range is required
of television sound which is greater than absolutely necessary in sound
broadcasting. Here the home receiver enters the problem. If dialogue,
and other relatively quiet sounds, ate broadcast at their proper level
over a period of time, it is likely that the volume of the home receiver
will be increased by the listener to match what has come to be expected
of broadcast sound. Then, if full dynamic range is employed, the
louder passages will exceed reasonable living -room power, or perhaps
overload the receiver. Hence, some compression must be employed,
yet without producing the above -mentioned unconvincing mis-match.
Treatment of this problem has evolved into a skillful handling of audio
levels in such a way as to produce changes in apparent loudness which
are greater than those actually broadcast. If it is known in advance
that some particular point in a performance will require a large
increase in volume, the loudness of the passages preceding the expected
increase in level is gradually-lowered, the process sometimes extending
over several minutes. This decrease in loudness is accomplished so
slowly that it does not come to the attention of the listener and is in
some degree compensated by what appears to be an increase in the
listeners aural sensitivity. Then, when the large amplitude is required,
an increase to maximum deviation is sufficient to produce an admirable
www.americanradiohistory.com
TELEVISION SOUND
11
effect. Of course, such a loud period causes thh listener's hearing again
to be reduced, and care must be exercised in returning to a medium
or low level of modulation.
The television sound problems which have been discussed are a few
of those that have already been encountered in television broadcast
operation. They have increased in complexity as television program
production has advanced its techniques. In an art developing as rapidly
as television, no one can be certain that indicated trends will be followed or that present methods and materials will be adequate, or even
useful, in the future. It is only by continuing the present close cooperation between the studio and the development laboratory that television's
sound problems can be solved.
www.americanradiohistory.com
THE FOCUSING VIEW- FINDER PROBLEM IN
TELEVISION CAMERA *t
BY
G. L. BEERS$
RCA
Manufacturing Company, Camden, N..1.
Summary-The technical excellence of a television program may frequently depend on the characteristics of the view finder used in the television
camera. Conditions peculiar to television make it desirable that television camera view finders be of the focusing type. The requirements of an ideal
view finder of this type are discussed. During the past ten years a number
of view -finder arrangements have been investigated in connection with the
development of television cameras. Several of these are described and their
relative merits indicated.
INTRODUCTION
0
NE of the most essential elements in a television camera is the
view finder. On its characteristics may depend the technical
excellence of the television program. The desirability of minioperating
personnel and the necessity for keeping a camera in
mizing
practically continuous operation during television programs of one or
two hours make it necessary that the view finder be of the focusing
type. Such a view finder not only provides a view of the scene which
is included in the field of the camera but also indicates when the lens
is properly focused on the desired scene.
During the past ten years a number of focusing view finders were
investigated to determine their suitability for use in television cameras.
Brief mention of some of these arrangements has already been made in
the technical literature on television equipment. Practical operating
experience with several view finders both in the studio and outdoors has
established certain requirements which an ideal view finder should
meet. It is the purpose of this paper to discuss these requirements; to
describe briefly several of the view -finder arrangements which have
been investigated, and to indicate their relative merits.
*
Decimal classification: R583.3.
f Presented at the Summer I.R.E. Convention, Cleveland, Ohio, on
June 30, 1942. Reprinted from Proc. I.R.E., March, 1943.
$ Now Asst. Director of Engineering in Charge of Advanced Development, RCA Victor Division, Camden, N. J.
12
www.americanradiohistory.com
VIEW FINDER
13
REQUIREMENTS OF AN IDEAL FOCUSING VIEW FINDER FOR
TELEVISION CAMERAS
The requirements of an ideal view finder may be stated as follows:
1. It should at all times accurately indicate when the camera is in
focus on the desired scene or object.
2. It should not only define that portion of the scene which is being
converted into the television image but also should reproduce a sufficient portion of the scene outside the camera field so that the cameraman will know in advance what the television picture will include if he
pans the camera in any direction.
3. It should provide an erect image which is correct left to right
and of sufficient size and brightness to minimize eyestrain.
4. It should not unduly complicate the procedure of interchanging
camera lenses or pickup tubes.
5. For portable pickup work the view finder should not contribute
substantially to the size and weight of the camera.
It will be noted that the first three of these requirements deal with
performance whereas the last two are concerned primarily with operating convenience.
In order to appreciate the significance of these requirements it is of
interest to discuss them in connection with the two general groups of
view finders into which the several individual view finders are subsequently classified. For the purpose of this discussion the first group
will consist of those view finders which derive the view -finder image
either directly or indirectly from the camera lens. The second group
includes those which make use of a separate optical system for producing the view -finder image.
REQUIREMENT NUMBER 1
Requirement 1 specifies that the view finder should at all times
accurately indicate when the camera is "in focus" on the desired
scene or object. Practical operating experience has shown that in
respect to this requirement it is desirable that the cameraman be
aware of a degradation in picture detail due to improper focus before
the loss in resolution is apparent to the television audience. The
view finders in group 1 have several limitations with respect to this
requirement. When the scene which is being televised is sufficiently
illuminated so that the camera lens can be stopped down to provide
greater depth of focus the view finders in this group do not provide
an accurate focus indication since the view -finder image has the same
depth of focus as the camera image. In other words, no apparent
www.americanradiohistory.com
14
TELEVISION, Volume IV
change in detail is observed by the cameraman as the lens is moved
back and forth through an appreciable range. This limitation may
not be particularly apparent to the television audience from the
standpoint of picture detail but is likely to be disturbing for another
reason. Under this condition the cameraman has a tendency to move
the camera lens back and forth to determine by approximation the
center of the range over which no effect on picture detail is observed
and thus establish the "in- focus" position of the lens. As the lens is
moved back and forth, the area included in the television image
changes in such a manner that the sides of the picture appear to move
in and out; an effect which is disturbing to most observers.
Another result of this inaccurate focus indication is encountered
when the camera is used under conditions where the illumination may
vary suddenly through a fairly wide range. Such conditions are frequently encountered in outdoor pickup of sporting events or spot news,
etc. If the lens is stopped down and the camera is inaccurately focused
on a scene in bright sunlight and the sun subsequently goes behind
a cloud, making it necessary to increase the lens aperture, the camera
will be out of focus. The focusing readjustment which is then required would have been avoided if the view finder had met requirement 1.
The view finders in group number 2 can all be made to meet requirement 1 provided they are constructed with sufficient mechanical rigidity to maintain, at all times, the proper alignment between the optical
systems for the view finder and pickup tube.
REQUIREMENT NUMBER 2
Requirement 2 states that the view finder should always provide
an image of and accurately define that portion of the scene which is
being converted into the television picture and should also provide a
view of at least a small part of the scene on each side of the television picture area. Unless the first part of this requirement is fulfilled the
cameraman may not know, for example, whether or not an individual's
head is in the picture. The second half of this requirement gives the
cameraman an indication of what will be included in the picture if
he pans the camera in any direction. The need for this information
may depend to some extent on whether the camera is being used in
the studio or outdoors. From one standpoint, there is less need for this
additional view-finder -image area in the studio because studio programs are usually rehearsed several times. On the other hand, in
studios several sets are frequently used in a limited space so that a
www.americanradiohistory.com
VIEW FINDER
15
camera can be changed quickly from one scene to another. This
makes it necessary for the cameraman to know what is included in a
small area outside the field of his camera so that he does not inadvertently include an edge of an undesired set in the picture. If the view
finder does not provide an image of this additional area it is frequently necessary for the cameraman to move his head sufficiently so
that he can look along one side of the camera to determine the effect
of panning the camera in a desired direction. Not only is this inconvenient but when the cameraman looks around the camera at the
brighter scene and then again looks at the image in the view finder
it is necessary for his eyes to readjust themselves to the difference in
the light intensity. In outdoor pickup work such as sporting events,
where the action is unpredictable, if the cameraman looks around one
side of the camera he may lose the action altogether before he has
time to again look into the view finder.
In general, the view finders in group 1 do not meet requirement 2
since the view-finder image which they provide is obtained from the
camera lens and covers the same picture area as the television image.
The view finders in group 2 make use of a separate optical system
and, therefore, can be made to provide a view of some of the scene
around the area which is converted into the television image. Such view
finders are, of course, provided with hairlines on the viewing screen or
some other expedient which indicates the actual area of the scene which
is included in the television picture. It is essential that the view finders
in this group be provided with some means which will correct for
parallax between the two optical systems.
REQUIREMENT NUMBER 3
The ideal view -finder requirement 3 is met if the view finder provides an erect image which is correct left to right and of sufficient size
and brightness to minimize eyestrain. A difference of opinion may
exist as to the necessity of having the view -finder image erect and correct left to right. If the cameraman has received considerable training
with cameras providing images which are inverted and reversed left to
right, such a view finder is undoubtedly satisfactory. He will then have
developed the proper co- ordination between the image he sees in the
view finder and the direction in which he must move the camera to produce a desired effect. On the other hand, in a new field, such as television, where it will be necessary to start with relatively untrained
personnel, it is felt that the corrected view -finder image will be more
satisfactory.
www.americanradiohistory.com
16
TELEVISION, Volume IV
With respect to the other stipulations in requirement 3, a viewfinder image at least 3 by 4 inches at a viewing distance of 12 inches
has been considered to be satisfactory. The image should be as bright
as possible. No difficulty has yet been encountered from having the
view -finder image too bright. The ability of a specific view -finder
arrangement to meet requirement 3 is basically determined by the
amount of light which it supplies to produce the optical image since,
if sufficient light is available, an optical system can be used to increase
the image size and reverse it in either or both directions.
The problem of providing sufficient light to produce a satisfactory
view -finder image is becoming more difficult as the sensitivity of
camera pickup tubes is increased. This limitation may ultimately make
it necessary to resort to a highly complicated view -finder arrangement
which will be described later.
REQUIREMENT NUMBER 4
This requirement is concerned with the effect of the view finder on
the ease of interchanging either pickup tubes or lenses. Since emergencies may arise which make it necessary to change pickup tubes and
since it is frequently desirable to change to a different focal -length
camera lens, it is essential that these changes be made in the shortest
time and with the least inconvenience.
This requirement is met to the greatest extent by the view finders
in group 1 since they derive the view-finder image from the camera
lens. The view finders in group 2, which use a separate optical system
for producing the view -finder image, all contain some element which
must be adjusted to provide satisfactory optical alignment between the
two optical systems when pickup tubes are changed. Up to the present
time it has been impracticable to manufacture pickup tubes with sufficiently close tolerance on the position of the mosaics and other elements
of the tubes to make them optically interchangeable. Some adjustment,
therefore, is necessary so that the view -finder image and the image on
the pickup tube are "in focus" simultaneously. It is possible to shift
the position of the pickup tube in a camera to obtain satisfactory optical alignment between the two optical systems. The size and weight of
the pickup tube with its deflecting yoke, however, make it much more
practical to move a ground -glass screen or some other element in the
view finder to provide the necessary alignment between the two optical
systems.
No serious complications are encountered in interchanging lenses of
different focal lengths in cameras employing the view finders in group 1
since only the camera lens is changed.
www.americanradiohistory.com
VIEW FINDER
17
The additional lens required with the dual -lens view -finder arrangements in group 2 make the problem of interchanging lenses somewhat
more difficult. This is particularly true where the lenses are large and
heavy such as those having focal lengths of 20 inches or more and
apertures of the order of f/4.5.
REQUIREMENT NUMBER 5
Requirement 5 is based on the desirability of keeping the size and
weight of television cameras for portable pickup work at a minimum.
Studio cameras are usually semipermanently mounted on large dollies
similar to those used in motion -picture work and the size and weight of
the television camera for studio work is, therefore, not a primary consideration. Portable television cameras, however, are used on conventional tripods and are set up and subsequently taken down at each
pickup location. It is, therefore, desirable to keep the size and weight
of cameras for portable pickup work at a minimum. In some cases a
sacrifice in view -finder performance has been made to permit a reduction in the size and weight of the camera. Some portable cameras
which employ one of the more complicated view finders are constructed
so that the camera can be separated into two units. This construction
not only makes the camera more portable but also makes it possible to
mount the two parts separately on the tripod.
Since the view finders in group 1 require less parts, occupy less
space, and contribute less weight, they are more acceptable from the
standpoint of requirement 5 than those in group 2.
DESCRIPTION OF INDIVIDUAL VIEW FINDERS
The following is a list of the view finders which will be described.
1. Mirror arrangement for observing the optical image on the
mosaic of the pickup tube.
2. Semisilvered mirror arrangement for utilizing the camera lens
to produce an optical image on a ground -glass viewing screen.
3. Kinescope or electronic view finder.
4. Kinescope or electronic view finder with remote focusing control.
5. Split -image view finder as used in the Contax and similar
cameras.
6. Duplicate -lens view finder as used in the Rolliflex camera.
7. Combination duplicate lens and kinescope view- finder.
The first four view finders in this list derive the view -finder image
either directly or indirectly from the camera lens and are those which
www.americanradiohistory.com
TELEVISION, Volume IV
18
were previously classified as the group 1 view finders. View finders 5,
6, and 7 are the group 2 view finders and obtain the view -finder image
from a separate optical system. For the sake of simplicity, the diagrams which will be used to illustrate the several view finders will not
show any means either for magnifying the optical image or for correcting it in the vertical and horizontal directions. It is apparent, that if
sufficient light is available, lens and mirror arrangements can be used
to accomplish any of these results. The means for correcting for
parallax is likewise omitted from the diagrams of the group 2 view
finders. Although the iconoscope is shown as the pickup tube in each
of the diagrams it is obvious that the orthicon or any other type of
pickup tube may be used.
-aA
-
VIEWING
APERTURE
LENS -C
ICONOSCOPE
FOCUS
CONTROL
Fig.
1-Mirror
arrangement for viewing the optical image on the mosaic
of the pickup tube.
MIRROR ARRANGEMENT FOR VIEWING THE OPTICAL IMAGE ON THE
MOSAIC OF THE PICKUP TUBE
The original iconoscope camera view -finder arrangement is illustrated by the diagram in Figure 1. With this view finder the cameraman, through the use of mirror A, observes on the mosaic B the optical
image which is produced by the camera lens C. The shape of the glass
envelope of the pickup tube is usually such that only a portion of the
image on the mosaic can be observed through the use of this system.
The mosaics of the more recent pickup tubes have very poor light reflecting properties and the optical image produced on the mosaic is,
therefore, unsatisfactory from the brightness standpoint. The chief
advantages of this arrangement is its simplicity. It requires a minimum of equipment since it makes use of only the camera lens and does
www.americanradiohistory.com
VIEW FINDER
19
not employ a separate viewing screen. No special adjustments are
necessary when changing either the camera lens or the pickup tube.
It has, however, all the limitations previously mentioned in connection
with the group 1 view finders.
SEMISILVERED MIRROR ARRANGEMENT UTILIZING THE CAMERA LENS TO
PRODUCE AN OPTICAL IMAGE ON A GROUND-GLASS VIEWING SCREEN
The diagram in Figure 2 illustrates the view finder system, which
makes use of a semisilvered mirror A to reflect some of the light transmitted by the lens B. This light is again reflected by the mirror C to
produce an optical image on the ground -glass viewing screen D. In the
experimental work on this arrangement, mirrors were used in which
the reflected light varied from 15 to 40 per cent. Since the total light
MIRROR
-.-GROUNDD°GLASS
11_
vIEwING
APERTURE
FOCUS
CONTROL
Fig.
2- Semisilvered- mirror
view -finder arrangement.
reflected from the front -surfaced mirror A is a comparatively small
percentage of the light passing through the mirror, the light reflected
from the back surface of the mirror may be a fairly large percentage
of the total reflected light. It is, therefore, necessary to use either a
very thin mirror or else have the back surface of the mirror coated
with a nonreflecting film; otherwise the light reflected from the back
surface produces an image which is sufficiently displaced from the
front -surface image to reduce the effective resolution of the view finder
to a point where it is definitely unsatisfactory.
The chief advantage to be found in this view finder likewise lies in
its relative simplicity. With respect to the arrangement shown in
Figure 1, it has the advantages of giving a somewhat brighter image
www.americanradiohistory.com
TELEVISION, Volume IV
20
and also will provide a view of the scene whose area is greater than
that included in the field of the camera.
The most serious disadvantage of this view -finder arrangement is
that it robs light from the mosaic of the pickup tube and therefore
decreases the effective light sensitivity of the system. Although it
meets requirement 2 it has the other limitations of the group 1 view
finders. Since a separate ground -glass viewing screen is used with this
arrangement it is necessary to adjust the position of the viewing screen
when changing pickup tubes so that the viewing screen is the same distance from the optical center of the lens as the mosaic of the pickup
tube. This view -finder arrangement also imposes a limitation on the
shortness of the focal length of the camera lens which can be used.
KINESCOPE
VIEWING
APERTURE
LENS
FOCUS
CONTROL
Fig.
3- Kinescope
or electronic view finder.
KINESCOPE OR ELECTRONIC VIEW FINDER
This view -finder arrangement is obtained by incorporating in the
camera a kinescope on which is reproduced the television image. It is
illustrated by the diagram in Figure 3.
The chief advantage of this view-finder system is that the relative
brightness of the view -finder image does not diminish as the sf-nsitivity of the pickup tube is increased. The brightness of the Kinescope
view -finder image is determined primarily by the characteristics cf th'.
kinescope which is used and the operating voltages which are employed.
It, like the view -finder arrangements illustrated in Figures 1 ?rid 2,
does not necessitate any view -finder adjustments when either pickup
tubes or camera lens are interc fanged and no correction for parallax is
required.
In addition to the several limitations discussed in connection with
the group I. yiew finders the kinescope type of view finder has the
www.americanradiohistory.com
VIEW FINDER
21
further restriction that the sharpness of the view -finder image is
dependent on the resolution of that portion of the television system
which it includes. It is, therefore, necessary that satisfactory electrical focus of the kinescope be maintained for this view finder to function satisfactorily. The space required in a camera to house this type
of view finder is relatively large. The several thousand volts which are
used as anode supply for the kinescope present a problem in providing
a satisfactory camera cable. If this camera -cable problem is avoided
by incorporating a voltage -supply unit in the camera a corresponding
increase in the size and weight of the camera results.
I
LENS
ICONOSCOPE
MOTOR-P'T
VIDEO
AMPLIFIER
L
IEWING
APERTURE
KINESCOPE
I
IC
Fig.
4- Kinescope or elect/onic
" coFocu
ti
view finder with remote focusing control.
KINESCOPE OR ELECTRONIC VIEW FINDER WITH REMOTE -FOCUSING
CONTROL
In the kinescope view -finder arrangement just described, a television monitoring unit with its kinescope is in effect moved from its
normal location so that it can be associated directly with the focusing
control in the camera. In the remote- control form of the kinescope
view finder the physical location of the parts is reversed and a remote
camera -focusing control is provided that can be used at the normal
location of the television monitoring unit. The diagram in Figure 4
www.americanradiohistory.com
TELEVISION, Volume IV
22
illustrates this arrangement. As indicated in the diagram, the remote
control of focus is accomplished through the use of Selsyn motors.
The chief advantage of this view-finder system lies in the fact that
it permits a camera which is small in size and light in weight. This is
especially desirable in portable pickup work. It makes possible a
camera which is particularly suitable for locations which are inaccessible to a cameraman. It also provides the advantages which have been
discussed in connection with the previous kinescope view finder. With
the remote -focusing arrangement the only view -finder equipment which
must be housed in the camera is the small Selsyn motor. A wire -frame
view finder mounted on the side of the camera is used by the cameraman to keep the camera trained on the desired scene. The focusing is
done by a control operator at the monitoring unit.
SPLIT- IMAGE
VIEWING
APERTURE
RANGE FINDER
LENS - COUPLING
LENS,
U
Fig.
5
-Split -image
FOCUS
CONTROL
view finder.
In addition to the deficiencies of the kinescope view finder illus-
trated by Figure 3, this arrangement has the further limitation that a
fairly high degree of co- ordination is required between the man who
is panning the camera and the man at the remote point who is operating the focusing control. When the focusing and panning are done by
the same individual he subconsciously starts to adjust the focusing control in the proper direction to correct for any change in distance
between the camera and the desired scene.
SPLIT -IMAGE VIEW FINDER AS USED IN THE CONTAX AND
SIMILAR CAMERAS
The diagram in Figure
5
illustrates this type of view finder. It
www.americanradiohistory.com
VIEW FINDER
23
utilizes an optical system which is actuated by the focusing control
simultaneously with the camera lens and produces two optical images
which are accurately superimposed when the camera lens is in focus on
a desired object or scene. The two images are displaced with respect to
each other when the focusing control has not been properly adjusted.
In a view finder of this type which was investigated the condition of
focus could be accurately determined only in a small area in the center
of the picture. Another limitation of this particular view finder was
that when using long focal -length lenses the actual size of an object in
the view finder remained the same as when a short focal -length lens
was used. A hairline indicator was provided to indicate the smaller
field covered by the longer focal-length lens. An adjustment is required
with this type of view finder when interchanging pickup tubes so that
--
VIEW- FINDING
LENS -A
n
GLASS - C
-
-~ ÁPERTURE
VIEWING
lFOCU 5
coNTgoL
-r
CAMERA
LENS -ß
1CONOSCOPE
MOSAIC -D
F.g.
6-Duplicate -lens view finder.
the optical system of the view finder is adjusted to compensate for
variations in the position of the mosaic in different pickup tubes.
DUPLICATE -LENS VIEW FINDER AS USED IN THE ROLLIFLEX CAMERA
As shown in Figure 6, an auxiliary lens A, which has the same focal
length as the camera lens B, is used to produce on the ground glass C
an optical image which corresponds to the optical image on the mosaic
D of the pickup tube. The position of the ground glass C, with respect
to the optical center of the lens A, must always correspond to the position of the mosaic D with respect to the lens B. The two lenses must be
matched accurately for focal length. To facilitate interchanging lenses
of different focal lengths each pair of lenses are usually assembled on a
single mounting plate. This view -finder system provides an image of
www.americanradiohistory.com
24
TELEVISION, Volume IV
a portion of the area outside that covered by the field of the camera.
A view finder of this type provides a very accurate indication of focus
under all conditions since the view -finder lens can be kept wide open
when the camera lens is stopped down. Since a fast lens is normally
used to provide the view -finder image the brightness of this image has
been relatively satisfactory. The increased sensitivity of pickup tubes,
however, is causing the image brightness obtained from this viewfinder system to decrease to the point where it no longer will be satisfactory. Some system for parallax correction is required with this type
of view finder. The amount of correction which is necessary is generally determined by the maximum diameter of the lenses supplied with
the camera.
The inability of this view finder to meet the ideal view -finder requirements is found in connection with requirements 4 and 5. Since a
separate lens is used to produce an optical image on a ground -glass
screen, the position of this screen must be adjusted to correspond to
that of the pickup -tube mosaic whenever pickup tubes are interchanged.
Since the longer focal-length lenses (20 -inch, f/4.5 lenses are frequently used) are large and heavy, the additional lens required for this
view finder not only makes the problem of interchanging lenses more
difficult but materially increases the over-all size and weight of the
camera.
COMBINATION DUPLICATE -LENS AND KINESCOPE VIEW FINDER
It has previously been pointed out that as the sensitivity of the television pickup tube is increased a corresponding decrease occurs in the
relative brightness of the image in an optical view finder. At present,
when the maximum sensitivity of the orthicon pickup tube is utilized,
the image brightness obtained from an optical view finder, such as the
duplicate-lens arrangement previously described, is on the verge of
being unsatisfactory. With the kinescope type of view finder any increase in the sensitivity of the pickup tube is automatically compensated insofar as the brightness of the view -finder image is concerned.
The types of kinescope view finders which have been described, however, do not meet performance requirements 1 and 2. If a further
improvement is made in the sensitivity of television pickup tubes, it
may be necessary to use a view finder of the type illustrated in Figure 7.
In this diagram it will be noted that two pickup tubes are used with
a pair of duplicate lenses. Associated with the camera pickup tube are
the normal television amplifier and deflection circuits. The amplifier
used with the view -finder pickup tube is designed to pass a wider frequency band than is normally required by the television system. The
www.americanradiohistory.com
VIEW FINDER
%
26
increase in resolution, which the wider frequency band permits, enables
this view finder to provide a more accurate indication of focus than
could be obtained from the previous kinescope view finders. Since a
separate view -finder lens is employed it can always be used at its maximum aperture even though the camera lens is stopped down, and thus
provide at all times an accurate indication of the proper focus adjustment. The deflection circuits for the view -finder pickup tube are so
arranged that a slightly greater area of the scene is scanned than is
the case with the camera pickup tube.
The deficiencies of this view finder from the standpoint of the ideal
view finder are in respect to the requirements 4 and 5 which deal primarily with operating convenience. With reference to requirement 4,
when pickup tubes are interchanged, the position of one of the pickup
tubes must be adjusted so that the mosaics of the two tubes are the
VIEW - FINDING
LENS
WIDE- BAND
VIDEO
AMPLI FIER KINESCOPE
VIEW -FINDINß
ICONOSCOP
CAMERA ICONOSCOPE
VIEW ING
APERTURE
FOCUS
CONTROL
CAMERA
LENS
Fig. 7-Combination duplicate-lens and kinescope view finder.
same distance from their respective lenses. The electrical focus of both
the view -finder pickup tube and kinescope must be kept in proper
adjustment for this view finder to function satisfactorily. The extra
equipment required for this type of view finder materially increases
the size and weight of a television camera.
COMPARISON OF THE INDIVIDUAL VIEW FINDERS
Table I shows the ideal view -finder requirements that are met by
the several view finders which have been described. The wording used
in the table for each of the requirements is such that a "yes" in the
column beneath a given view finder indicates that it meets the specified
requirements.
CONCLUSIONS
It is apparent that none of the view finders which have been de-
www.americanradiohistory.com
H
TELEVISION, Volume
`
I
2æ
);
k.
,--
}})
4
°
);
/
)
.
ª
ª
}/
ª
ª
,
&
/
>u,
:1.'
_
0
§
k
;0\
,.
ª
§(-
Q;
ª
ª
ª
/
/
o
ª
/
&
ª
ª
/
\
.
°
/j
\(
]®>
&..
,}
&
)§;2
)}\Ñ
ª/|Ja/2
al" 9
- \|§],
kEi
Ça
jI)22w
ª
ª
ª
§ §®
b
)
t
;
\/ 22
5) ªg
/
77\ 2}
<ÿ}
ª|§ ª7
ª!]
/
t
))
î\
§
))
ª
]
I\d
ª
| e
/
/§
\\
ªE
www.americanradiohistory.com
/\
VIEW FINDER
27
scribed meet all the requirements of an ideal view finder. The relative
importance of some of the requirements is determined to a considerable
extent by whether the camera is intended for studio or outdoor pickup
work. In general, the duplicate -lens type of view finder has given the
most satisfactory results. If it is desired to keep the size and weight
of the camera as near the minimum as possible, the kinescope view
finder with remote -focusing control is a practical arrangement. A substantial increase in the sensitivity of television pickup tubes will result
in more consideration being given to the several types of kinescope
view finders.
In this discussion no reference has been made to the relative cost
of the various view-finder arrangements. For the time being, at least,
the cost of television pickup equipment has been considered to be of
secondary importance to performance and operating convenience.
ACKNOWLEDGMENT
The writer wishes to acknowledge the individual and cooperative
efforts of numerous Radio Corporation of America and National Broadcasting Company engineers who have contributed to the solution of the
view-finder problem.
www.americanradiohistory.com
ELECTRON BOMBARDMENT IN
TELEVISION TUBES *t
BY
I. G. MALOFF
RCA
Victor Division, Camden.
N. J.
Summary-A detailed analysis of actions occurring in an Iconoscope
when an elemental area of the mosaic is bombarded by the scanning electron
beam under conditions varying from dark to light. The sticking effect,
important in projection kinescopes, is explained.
N AN ordinary vacuum tube there are two main effects-the controlled unidirectional flow of electrons from cathode to plate, and
the bombardment of the plate by these electrons. The controlled
unidirectional conduction is in these tubes the desired effect, and the
electron bombardment is generally undesirable because it heats the
collecting electrode and results in energy losses.
While modern all- electronic television makes use of a great num-
ber of ordinary radio vacuum tubes, its actual functioning depends
mainly on cathode -ray tubes at the transmitting and receiving ends
of the television system. In television cathode -ray tubes the same two
main effects exist, but with an important difference-the electron
bombardment is utilized, while the unidirectional conduction is incidental to the operation of the tubes. Electron bombardment of targets
and the resulting secondary emission make possible the operation of
both the Kinescope (receiving tube) and the Iconoscope (camera pickup tube).
There is another distinction between the ordinary vacuum tube and
the television cathode -ray tube. In ordinary tubes the plate acts as
both the target and the collector, whereas in television tubes the target
is either an insulator or an insulated conductor, with the collection of
electrons being done by another electrode. This collecting electrode
is usually the second anode.
When a surface is bombarded by electrons of considerable velocity,
the incident electrons may impart to the electrons near the surface
sufficient energy to escape from the surface. The incident electrons
are called the primary electrons, while the electrons leaving the sur*
Decimal Classification: R583.6 x R138.3
t Reprinted
from Electronics, January, 1944.
2F
www.americanradiohistory.com
ELECTRON BOMBARDMENT
29
face are called the secondary electrons. The number of secondary electrons emitted for each primary electron, and the velocity of the secondaries, vary with the velocity of the primary electrons and with the
chemical nature and physical condition of the surface.
BOMBARDMENT RESEARCH TUBE
a typical tube for studying electron bombardment of a
given surface is shown. The target is located at the center of a
metallic sphere and is kept at a desired potential with respect to the
In Figure
1
electron gun
e
Target
terminal
--II II -11111111111
Fig.
1- Essential
features of special cathode -ray tube developed for
studying electron bombardment of target surfaces.
cathode, thereby assuring a definite velocity of the bombarding primary electrons. The electron beam is produced in a conventional electron gun and enters the sphere through a small hole in its wall.
Provisions are made for varying the potential of the sphere with
respect to the target, as well as for reading the currents to the target,
the sphere and the beam current.
For a conducting target, such as pure nickel, the target current It
is equal to the beam (primary) current It, minus the collector current
I, (It = Ib Ia) A negative value for target current It indicates that
the ratio of secondary current to primary current (I, /Ip) is greater
-
.
www.americanradiohistory.com
TELEVISION, Volume IV
30
1.5
k
LO
Tr;
0.5
,..
400 800 1200 I600 2000 2400
Speed of Primary Electrons
in
Equivalent Volts
i
Fig. 2 -Ratio of secondary electrons to primary electrons as plotted against
speed of primary electrons when using a pure nickel target in the tube of
Fig. 1.
than unity. When It is positive, the ratio is less than unity, and when
It is zero the ratio is unity. A curve of variation of this ratio for pure
nickel target as a function of velocity of primary electrons is shown
in Figure 2. (For a contaminated metal surface the secondary emission ratio is generally greater than for the clean surface shown.)
EQUILIBRIUM POTENTIAL OF TARGET
At a given voltage difference between the target and the cathode
one may apply increasing negative or retarding voltages on the collector with respect to the target. When this is done, the collector current will gradually drop to zero.
For pure nickel and a 500 -volt beam, the curve of the ratio of collector current to beam current as a function of the retarding voltage
is shown in Figure 3. At point P, where the curve goes through a ratio
1.5
Ic
Tb
P
LO
0.5
O
0
4
8
12
16
Retarding Potential in Volts
Fig. 3 -Ratio of collector current to beam current as plotted against negative retarding potential (collector negative with respect to target) when
using a pure nickel target and a fixed primary electron speed of 500 equivalent volts in the cathode -ray tube of Fig. 1,
www.americanradiohistory.com
ELECTRON BOMBARDMENT
31
value of unity, the target current It is equal to zero. This is an important point on the curve. Since there is no current flowing to the target,
the lead to the target may be cut under the conditions at P without
disturbing either the electrode potentials or the currents to them.
Since the potential of the target is not changed when the target lead
is cut under the conditions at P, the target potential is still equivalent
volts
to the beam velocity. The collecting sphere, however, is at
with respect to the target. A conclusion follows: an insulated nickel
target will assume a potential of 3 volts higher than the collector for
500 -volt primary electrons. (The collector has to be at 497 volts.) Or,
generally speaking, when the secondary emission ratio is higher than
unity, an insulated metal target will assume a potential of a few volts
positive with respect to the collector. This potential is called the
equilibrium potential of an insulated target under electron bombardment. The velocity of the primary electrons will of course be equivalent to the sum of the collector voltage plus the voltage between the
insulated target and the collector.
-3
STICKING EFFECT
Now let the nickel target float and increase the collector potential
to 1700 volts. The secondary emission ratio of pure nickel at this
voltage is unity and the target will float to the same potential as the
collector, so that the difference of potential between the two is zero.
If the collector potential is further increased, the insulated target
will stay at the same potential with respect to the cathode. In other
words, it will be getting more and more negative with respect to the
collector. This phenomena is called the "sticking effect" in television
vernacular. To observe how it happens, make a metallic connection
between the target and the collector and raise both to 2000 volts with
respect to the cathode. The secondary emission ratio at 2000 volts is
nine -tenths, and while meter Ib will measure the current flowing to the
target and collector together, nine -tenths of it will flow to the collector
and one -tenth to the target.
If the target is now disconnected and left floating, more electrons
will be arriving at it than departing from it, charging it negatively
with respect to the collector. At the same time the arriving primary
electrons slow down to the velocity equivalent to the actual voltage
between the target and the cathode.
With primary electrons slower than 2000 volts, the secondary emission ratio increases, and finally, when the target is at exactly 1700
volts positive with respect to the cathode, and 300 volts negative with
www.americanradiohistory.com
TELEVISION, Volume IV
32
respect to the collector, the ratio becomes unity. In other words, the
target here "sticks" at an equilibrium potential of 1700 volts, and any
increase in the voltage on the collector will not increase the potential
of the floating target.
The two cases just described play a very important part in the performance of television cathode -ray tubes and have to be clearly understood before an analysis of their performance can be undertaken.
Sticking is especially important in projection kinescopes when it is
desired to use high voltages on the second anode to get more light.
Before using these high voltages, such as 20 to 70 thousand volts, one
Fig.
4- Essential features
of a standard Iconoscope for
electronic television cameras.
must make certain that the luminescent material does not "stick"
below the value chosen.
MOSAIC
IS
TARGET IN ICONOSCOPE
The action of electron bombardment in a standard Iconoscope is
somewhat similar to the case of bombarding an insulated metal target
with electrons having a velocity at which the secondary emission ratio
of the target is greater than unity.
Figure 4 shows the arrangement of the essential parts in a stand-
www.americanradiohistory.com
ELECTRON BOMBARDMENT
33
and Iconoscope. Generated by a conventional electron gun, an electron
beam of approximately 1000 -volt velocity enters a nearly equipotential
space in the bulb portion of the tube, where it strikes the photosensitive
mosaic. The mosaic consists of a multiplicity of minute silver globules,
oxidized and caesiated (in other words, photo- sensitized), uniformly
distributed on a 9 X 12 -cm sheet of mica only 0.0025 cm thick. The
back side of the mica sheet is platinized to form a capacitor between
the globules and the platinum coating, having a value of 122 if per
sq cm.
The secondary emission characteristics of oxidized and caesiated
silver vary greatly, depending on the condition and method of preparaThe Iconoscope
-a television pickup
tube.
tion of the surface. These variations however are restricted to the
maximum value of the secondary emission, while the relative velocity
or energy distribution of secondaries changes little.
A typical energy distribution curve of secondary emission of photosensitized silver for 1000 -volt primaries is shown in Figure 5. As may
be seen from the curve, the secondary emission characteristics of a
complex surface differ from those of pure metals. In the case under
consideration two important characteristic features attract attention
at once. The first is the fact that the secondary emission ratio reaches
a very high value of 5.1, compared with slightly more than one for
pure nickel.
www.americanradiohistory.com
TELEVISION, Volume IV
34
The second feature is that, with the collector at zero potential with
respect to the target, not all the electrons from the target are collected. Apparently either there are some electrons hidden in the
crevices of the surface, or some electrons are emitted with "insufficient" velocities to reach the collector. They may be drawn to the
collector by applying positive potentials to the collector with respect to
the target. The second feature, while interesting, is of little importance to us since the collector is seldom positive with respect to the
mosaic. Our interest lies, therefore, in the portion of the curve to the
right of the line of zero collector potential.
SCANNING ACTION
In actual Iconoscopes, the primary beam may be used with electron
_
,
,.-:
O
+2
Poteniiat of`i'tìrge with
Fig.
5-Secondary emission
f
characteristic of photosensitized silver.
velocity between 500 and 2000 volts. Usually however, the velocity is
1000 equivalent volts. The beam is focused at the mosaic to an area
of approximately one picture element. Therefore, at any one instant
the area under electron bombardment is that of one picture element.
The scanning spot, however, moves along the mosaic at a very high
speed, bombarding a point on the mosaic for only 1.28 X 10 -7 sec. for
every frame of a 441 -line picture.
If light falls on a portion of the mosaic, photoelectric emission
takes place, and by losing some electrons that portion of the mosaic
acquires a positive charge. Suppose a portion of the mosaic is charged
to one volt positive, and consider what happens when this mosaic is
scanned, first the part of it having no charge, then the boundary
between the dark, the uncharged mosaic and the positively charged
lighted mosaic, and finally when the lighted area is scanned.
The scanning spot may be considered as a square brush having an
www.americanradiohistory.com
ELECTRON BOMBARDMENT
35
area of one picture element. The charges on the globules or sub elementary capacitances under the spot are instantaneously equalized,
so that one may talk of the potential of the spot and the charge on the
spot. The charge on the spot is affected by the charging current, which
is equal to the difference between the secondary current and the
primary or beam current.
The charging current characteristic shown in Figure 6 is easily
derived by subtracting the beam current (unity) from the secondary
emission characteristic. In the interval from 0 to 2 volts the charging
characteristic is represented very closely by the straight line:
1V) In, where V is the potential of the
V) Ib =(3
1cm= (a
-ß
-
RIM
Mill
MIME
MAIM
MI
H°(ceQV)Ib
MN
ME=
Fig.
6- Charging current characteristic of an
Iconoscope mosaic.
target with respect to the collector.
SCANNING
A
DARK AREA
Now suppose the spot is moving in a normal scanning manner along
the uncharged portion of the mosaic. If a steady state has been reached,
the current to the signal plate, the potential of the moving spot, and
the potential of elements already scanned are all steady and constant.
The charging current is therefore
ICH=
(a-ßV)
Ib
www.americanradiohistory.com
(1)
TELEVISION, Volume IV
36
The mosaic elements after scanning are all charged to the same
potential V. Since the capacitance charged per second by the spot is
nNC, the charge left by the spot on the mosaic is nNCV coulombs per
second, where n is the number of elements per frame, N is frames per
second and C is the capacitance of a picture element. Coulombs per
second is current in amperes; therefore, the two currents must be
equal:
I
(a- ßV)
Ib=nNCV
(2)
Solving this for V gives
V=
alb
nNC
- ßl
b
a
1
ß
nNC
+1
f31n
Fig.
7-Equivalent circuit of a dark mosaic undergoing scanning.
Now letting
gives
1
/nNC = R, letting 1/ß lb = r and letting a/ß = E8
V
= E8
R1
r -R1
(3)
The solution in Equation (3) is an exact equivalent of a simple
charge of capacitors by a commutating brush in the arrangement
shown in Figure 7. The brush, covering a number of capacitors of a
total capacitance C and commutating nNC farads per second, charges
them through a resistance 1,1/3 Ib from a battery of a/ß volts.
SCANNING A BOUNDARY
If at time t= 0 the leading edge of a rectangular scanning spot or
brush reaches the boundary between dark and lighted portions of the
www.americanradiohistory.com
ELECTRON BOMBARDMENT
37
mosaic, a transient condition will prevail until some later time, when a
new and different steady state will be reached. As shown in Figure 8,
before the leading edge of the spot reached the lighted portion, the
current to the mosaic and voltage to which its elements are charged
0, V = EXR1/ (r
are both steady and of values given. Thus, before
+ R1) and to = E8 R1/ (r + R1) r.
After t = 0 the charge coming under the spot is nNCE0 coulombs
per second, the charge left over on the mosaic after scanning is nNCV
coulombs per second, and the secondary emission charging current is
(E8
V) /r. These quantities should satisfy the following equation
t=
-
at all times:
i
VC =
Fig.
f
?INGE() -l- E8
8- Scanning from
-V
nNCV
r
)
dt
( 4)
a dark area into a lighted area.
Dquation (4) reduces to a linear differential equation which when
solved by conventional methods gives
dV
dt
V= E8
At t =
co
Rl
r+Rl
+
r + R1
rR1C
+ Eo
V=
Eor + E8R1
R1rC
r
r+R1 C\
1- e
rRiC
t/
(5)
a new steady state is reached, expressed by
V
= E,
R1
r+R1
r
-}-
Eo
www.americanradiohistory.com
(6)
TELEVISION, Volume IV
38
,
SCANNING CURRENTS AND POTENTIALS
The transient charging current while scanning the boundary is
given by
lo=loll=
(«---,(3V)
1
=10
r
}
R
lb= (E8-V)/r
[Es_Eo(i_e
rRiC
(7)
For values of constants encountered in practice, a plot of the charging current I is given in Figure 9. A spot which instantaneously
equalized all the charges under it has been assumed. That such a con-
Fig.
9-Transient charging current I and mosaic
during scanning.
potential V
d:tion actually occurs, there is little doubt. However, authorities disagree as to the values of electron densities at which the surface under
bombardment begins to be thoroughly conductive.
If there is no surface conductivity under the electron spot, the
nonconductive charging current curve in Figure 9 applies. It may be
seen that it is of small importance whether there is, or is not, conductivity under the spot. Except for a small difference in the duration of
the transient, the two discharge processes differ very little one from
another. The magnitudes of current changes are between 8 and 9
hundredths of a microampere.
The potentials V left over on the mosaic by the scanning spot are
also almost equal in the two cases, as is shown by the lower curves of
Figure 9. A thing to note, however, is that it is hard to expect a
potential change on the globules from before to after scanning, greater
than 6 tenths of a volt. For a most efficiently activated mosaic this
value may rise to one volt.
www.americanradiohistory.com
ELECTRON BOMBARDMENT
39
MOSAIC CAPACITANCE
So far in this study of charging action of the electron bombardment
some broad simplifying assumptions were used. This action was
studied under the assumption that the mosaic plate is infinite in area,
but has constants per unit area of the actual mosaic. As a result expressions were obtained for the video -frequency currents, time constants, potentials, of the mosaic before and after scanning, etc. An
infinite area, however, means an infinite capacitance capable of absorbing any charge without a change in voltage. Actually, the iconoscope
mosaic has large but finite capacitance, which may be considered
infinite only for the upper region of the video frequencies.
At low video frequencies the mosaic capacitance plays a very important part, as is substantiated by the experimental evidence shown
in Figure 10. When scanning a mosaic in the dark, the iconoscope
output current is not zero. In fact, its peak -to-peak value is of the
order of 5 x 10 -8 ampere. In the reproduced picture this spurious or
dark spot current would produce a great distortion, if it were not compensated for and balanced out.
Now, if a light pattern of horizontal bars is thrown on the mosaic,
a square wave of current is generated by the bombardment. The square
wave is superimposed on the dark-spot signal. The oscillograms show
the wave shape of one field of television signal with its darkspot signal
(vertical dark spot). Besides the vertical dark spot there is a horizontal dark spot in the signal, but in these oscillograms, the line frequency and all higher frequencies were filtered out.
SPURIOUS OUTPUT SIGNALS
The spurious signal of the iconoscope is a result of scanning of the
mosaic by the bombarding beam. Essentially, the Iconoscope is an a -c
device, since its output current flows to its signal plate which is a
terminal of a 0.013 -4 capacitor. The secondary emission current at a
steady -state condition therefore has to average out to a value equal to
the beam current, while the bombarding electrons are knocking out
five times their number from the mosaic. The excess electrons return
to other parts of the mosaic and charge it in their turn. They charge it
in a nonuniform manner, contributing to the vertical and horizontal
spurious signals.
In general, with the mosaic in darkness at the start of the scan of
a frame (or a field, rather) more electrons flow to the collector or
www.americanradiohistory.com
TELEVISION, Volume IV
40
E
v
>
50
30
Period of One F 1Pfd+1
Sac.
0
Time
(b)
10- Oscillograms
and curve showing effect of mosaic capacitance. (a)
mosaic in darkness; (b)- signal plate potential
with mosaic in darkness; (c)- signal plate current with mosaic illuminated
by 120-cycle square wave pattern.
Fig.
-Signal plate current with
www.americanradiohistory.com
ELECTRON BOMBARDMENT
41
second anode than are supplied by the beam to the mosaic. The current
flows down from the signal plate, charging it negatively. Somewhere
in the middle of the scan the flow of electrons to the mosaic becomes
equal to the flow to the collector, and the current to the signal plate
becomes zero. Then it reverses direction and charges the signal plate
in a positive direction, while the flow of electrons to the mosaic is larger
than to the collector.
EQUIVALENT GENERATED
EMF
Since the Iconoscope is a generator of electrical signals, one may
inquire whether it can be represented as a source of an emf generated
in some sort of a passive network. Since its output occurs as a current
between an output terminal and ground one may inquire as to the value
of equivalent emf generated and the magnitude and sense of its equivalent impedance, reactance and resistance. These have been investigated
by the old reliable experimental method used in determining the emf
and the internal resistance of a battery. The device is loaded with
resistances of various values and the output current is read by a suitable meter. A set of simple simultaneous algebraic equations results
which when solved yields the desired values.
EFFECT OF BAR PATTERN
Assume that a pattern of alternating black and white bars is thrown
onto a mosaic of a normally operating Iconoscope. The resultant signal
is then composed of a 180 -cycle square wave plus a spurious signal.
The voltage output across a normal coupling resistance is observed,
then resistors of different values are inserted in series with the normal
coupling resistor. The resultant oscillograms are shown in Figure 11.
Apparently more than the capacitance of a single picture element is
active in the Iconoscope at low frequencies -about one -tenth of the
mosaic area in fact, while the emf generated by the Iconoscope is 0.3
to 0.5 v.
CONCLUSIONS
As we have just shown, the fact that the Iconoscope is utilizing
electron bombardment and secondary emission does not mean that it is
in a class of devices which are foreign to communications engineers.
It is a generator of electrical signals, having an internal impedance
and a definite electromotive force. Its characteristics are readily measured and used in the design of television systems; while certain of its
www.americanradiohistory.com
TELEVISION, Volume IV
42
310,000
,4
1300
(e)
Fig. 11- Internal impedance measurements on an Iconoscope having a beam
current of 0.1 µa and a 180 -cycle square -wave output signal. Impedance
values R in Iconoscope circuit (c) for the four oscillograms are as follows:
(a)-o
-6
ohms; (b)
megohms; (c) -12 megohms; (d) 18 megohms. The
equivalent circuit of 180 cycles with no backlighting is shown at (f).
actions do not make it an ideal generator of a video signal, it has made
possible modern high- definition television.*
* For a list of references on the subject the reader is referred to the
extensive bibliography in "Television," by Zworykin and Morton, John Wiley
& Sons, New York, 1940.
www.americanradiohistory.com
IMAGE ORTHICON CAMERA *t
BY
R. D. KELL AND G. C. SZIKLAI
Research Department, RCA Laboratories Division,
Princeton, N. J.
Summary -One of a series of developmental television cameras using
the image orthicon is described. The complete camera weighs less than
forty pounds. The input power required by the camera is 300 watts. This
power may be supplied by a non-regulated power supply or generator. A
unique regulated high voltage supply was developed for the electron multiplier and image section of the camera tube. The camera circuits include the
deflection system, voltage regulators, black -level setting, blanking circuits,
and video amplifiers. A total of seventeen tubes is used in the camera. An
extremely high -sensitivity version of the camera, using reflective optics, is
also described.
I. INTRODUCTION
THE development of the image orthicon' provided a camera tube
for an extremely sensitive television camera. In addition, due to
its high output signal level, it permitted a substantial reduction
in the number of tubes used in the video amplifier, and thus permitted
the incorporation of other circuits within the camera that were built
into auxiliary equipments in previous types of cameras. While certain
operating features of the image orthicon provide simple blanking and
black level setting, the photo-cathode image section and multiplier electrodes require potentials of such values and stability that new circuits
had to be designed to permit the incorporation of these supplies in the
camera.
II. RESOLUTION
During the early part of the development the major effort was
applied to improving the resolution of the image orthicon. In the
course of the investigation, it was observed that by scanning only a
portion of the target considerably better resolution was obtained than
when the whole target area was scanned. In order to determine whether
the lack of resolution was due to limitations in the scanning or in the
image section of the tube, or possibly in the coupling section between
the camera tube and the amplifier, a variable frequency signal was
applied to the target. It was found that a 5- megacycle signal was satisfactorily passed by the scanning section and the amplifier, indicating
Decimal Classification: R583.12.
f Reprinted from RCA REVIEW, March, 1946.
Paper on the image orthicon was presented by A. Rose, H. B. Law,
and P. K. Weimer, at the I.R.E. Winter Technical Meeting, on January 24,
1
1946.
43
www.americanradiohistory.com
TELEVISION, Volume IV
44
that the limitation was caused by the image section of the tube. The
fact, however, that scanning a small portion of the target provided a
well- resolved picture indicated that the image section itself formed a
picture of satisfactory resolution. These experimental results tended to
show that an interaction between the scanning and the image section
was degrading the picture.
Upon the assumption that the horizontal scanning field was vibrating the electron image on the target, and thereby blurring the picture,
a portion of the horizontal deflecting current was applied to an auxiliary coil locatea over the image section. This current had a direction
opposite to that in the deflecting coil in order to cancel the variable
component of the magnetic field. The experiment resulted in considerable improvement in resolution.
Fig.
1-Focusing,
Deflection, and Alignment Coil Assemblies.
As another approach to the problem, this crosstalk effect was reduced by careful shielding. The problem was pursued further, since it
was known that a reduction of the deflection power would proportionally reduce the effect. A simple reduction of the focusing field intensity
allowed a reduction in deflection power, but it degraded the resolution
around the edges of the picture, and hence was not permissible. However, it was found that by reducing the focusing field over the deflection coils and reinforcing it over the gun and the target, better resolution in the corners was obtained because of better electron landings.
It was further found that the desired field distribution could be
obtained with a uniformly -wound focusing coil and a magnetic shield
www.americanradiohistory.com
IMAGE ORTHICON CAMERA
45
(of iron wire) over the focusing coil. This method, with the addition
of electrostatic shielding, was finally adopted. The arrangement considerably reduced the required deflection power and provided a resolution in excess of 450 lines under high light conditions. In cameras
where a maximum resolution is required, the image section bucking
coil was also provided.
Figure 1 shows the focusing coil assembly, the deflection coil assembly, and the alignment coil in the usual order with the shields. The
shield which extends from the deflection coil over the gun end is to prevent pickup from the deflection coil by the signal lead.
III. THE CIRCUITS
diagram of the camera is shown in Figure
A block
"\ ¡¡'
TAR4kT
FOCU61Np
^
V
11`1
I
(
The video
ALI4NMkRT
can.
COIL
OkI.IKTIOw
2.
I
OIL
OUT
.CAT, Ref,
6 RkOULsOR
Ohl
6AC7
Fig.
2- Camera
Block Diagram.
output is taken from the last, or fifth, multiplier of the orthicon across
a 33,000 -ohm resistor, through which a high potential of approximately
1500 volts is fed to the multiplier. This load resistance is about one tenth of the conventional value used with iconoscopes. It is permitted
by the higher signal current output of the image orthicon. The lower
signal output resistance also permitted the use of a correspondingly
reduced amount of equalization in the high peaker circuit' in the second video amplifier plate circuit. With the five stage multiplier image
orthicon, substantially all the noise generated is due to the scanning
beam, and with the reduced equalization in the high -peaking circuit
there were no noticeable microphonics due to the amplifier system.
2
U. S. Pat. No. 2,151,072 -A. V. Bedford, March, 1939.
www.americanradiohistory.com
TELEVISION, Volume IV
46
By using a clamping circuit at the fourth video amplifier stage to
reinsert the low video frequencies, further assurance was taken to
keep the camera free from microphonics generated in the amplifier.
The clamping circuit is shown in Figure 3, and it functions as follows:3 At the input to the amplifier the video signal is given a reference
level, such as black, during the horizontal return time. This reference
level is readily obtained by applying pulses to the target of the image
orthicon during the horizontal blanking interval. These pulses cause
all of the scanning beam to return to the multipliers. This is a signal
which is equivalent to black level. After this reference level is inserted
in the signal, the low frequency response of the video amplifier can be
reduced to the point where it will just pass a square wave corresponding to line frequency. At a high signal level, where all danger of
microphonic disturbance in the amplifier tubes is passed, the signal at
+300
Fig.
«300
3-Direct Current Setting Circuit.
the time of the black reference (which has become variable in level due
to the presence of picture signal) is again established at a fixed value.
With black level representing a fixed bias on the amplifier stage, it
follows that the low frequency and direct current component of the signal are again present. Referring to Figure 3, the video signal which
has lost the direct current and all low frequency components, passes
from the plate circuit of tube A to the grid tube B through the small
coupling condenser C. The grid leak on tube B is replaced by the two
diodes of the 6H6 type. The push -pull pulses obtained from the tube E
are applied to the diodes. The pulses cause both diodes to conduct. This
is equivalent to connecting the grid of tube B to the battery through a
switch. This makes the potential of the grid corresponding to black
equal to the battery voltage.
U. S. Pat. No. 2,299,945 -K. R. Wendt, October, 1942.
www.americanradiohistory.com
IMAGE ORTHICON CAMERA
47
The reconstructed signal is mixed with a blanking signal in the
plate circuit of the fourth amplifier stage, then fed to a cathode follower
output stage, which provides a complete video signal of approximately
one volt peak-to -peak value.
The deflection circuit consists of the horizontal and vertical oscillators, the two discharge tubes in one envelope, and class Al type deflection output stages for both the vertical and horizontal deflection.
The high voltages for the image orthicon were obtained by rectifying the return sweep voltage of the horizontal output stage. Any change
in the deflection voltage then tended to upset the operating conditions
of the image orthicon tube. Since the photo- cathode voltage, in particular, is very critical, a simple voltage regulator was devised.
Owing to the fact that the current required was exceedingly small,
the constant current property of a pentode was considered the simplest
method of providing a constant voltage. A further improvement in
regulation was obtained by applying a portion of the rectified potential
Fig.
4-High
Voltage Power Supply and Regulator
to the control grid of the pentode rectifier and degenerating any change
that might occur.
The circuit is shown in Figure 4. A portion of the high alternating
current pulse voltage across the horizontal deflecting output transformer is rectified by the pentode V3. The useful direct current voltage supply then occurs at the negative terminal shown and is regulated by suitably controlling the grid voltage of the pentode. A portion of the output
of the power supply G is regulated by the glow discharge tube V2 and is
used for the screen supply to V3. This regulated voltage also serves as a
reference potential for the control action, in that a portion of the rectified output voltage is subtracted from it and applied to the control grid.
This arrangement will produce a large potential change of the grid voltage with small percentage change of the output voltage. When the negative potential tends to increase across the load resistance R, the grid
www.americanradiohistory.com
TELEVISION, Volume IV
48
becomes more negative and the resistance of the circuit increases,
thereby reducing the potential across the load. The high voltage for the
multipliers is supplied by the rectifier V4. The wall coating and persuader voltages are obtained from the voltage regulator V2 which is
actually two VR -150 tubes in series.
Fig.
5
-Top
View of Camera Chassis
Figure 5 shows the top of the camera chassis. The high voltage
signal coupling capacitor may be seen in the left side of the picture.
The video amplifier is located in the bottom row. The voltage regu
lators, high voltage supplies, and deflecting circuits occupy the top row.
Fig.
6-Bottom View of Camera
Chassis
The bucking coil to eliminate the image jiggling is on top of the focusing coil. A bottom view of the chassis, showing the circuit components,
is given in Figure 6. The voltage divider for the electron multipliers
www.americanradiohistory.com
49
IMAGE ORTHICON CAMERA
at the top, and the deflection
transformers at the left side of the picture. Four controls, namely, the
scanning beam bias, the scanning section focusing control, the image
section focusing control, and the amplifier gain control are readily accessible by the opening of a hinged lid. The other controls are normally
is at the left side, the potentiometers
covered with a plate fastened with screws.
Fig.
7- External View of
IV.
Camera Assembly with Lens
THE CAMERA ASSEMBLIES
Figure 7 shows an external view of the camera assembled with a
f 2.7 lens. Figure 8 shows an image orthicon camera assembled
with a reflecting Schmidt optical system. The photo- cathode surface of
12 cm.
Fig.
8- Camera Assembly with Reflective Optical
www.americanradiohistory.com
System
TELEVISION, Volume IV
50
the image orthicon used in this camera was properly curved in order
to secure proper focus of the optical image of the entire field of view,
and it was placed approximately in line with the spherical mirror. The
ASPHERICAL
MIRROR
SPHERICAL
MIRROR
IMAGE
ORTHICON
-PHOTO
SURFACE
FOCUS
ADJUSTING
SCREW
Fig.
PLANE
MIRROR
9- Construction of the Reflective Optical
Fig.
10- Camera
Demonstration Setup
www.americanradiohistory.com
System
IMAGE ORTHICON CAMERA
51
design of the optical system is shown in Figure 9. A brass barrel provides a rigid structure for the system. The focusing is adjusted by the
plane mirror which reflects the image on the photo -cathode of the image
orthicon. The system has an (f) power of .7 and an aperture of 10
inches. The completed optical unit has a resolution of better than 1000
lines at the image surface.
Fig. 11- Television Picture Taken
with the Subject Illuminated by
3 Kilowatt Incandescent Light.
Fig.
Fig. 12- Picture with the Subject
Illuminated by a 25 Watt Desk
Lamp.
13- Picture with the Subject
Illuminated by One Candle.
V. PERFORMANCE
Figure 10 shows a typical demonstration setup with the image
orthicon camera using the f 2.7 lens. Lighting can be provided by the
two one-and -a -half kilowatt reflectors, a 25 watt lamp, or by one to four
candles. Figure 11 shows a picture taken from a 12-inch direct viewing monitor when the subject was illuminated by the two one- and-ahalf kilowatt lights. Figure 12 shows the same subject illuminated
with the 25 -watt lamp, and Figure 13 shows the same subject with a
single candle at a distance of three feet as the only source of illumina-
www.americanradiohistory.com
52
TELEVISION, Volume IV
tion. The main difference between the last two pictures is in the noise
present, which can not be seen in the photographs due to the inherent
integration of the exposure.
The sensitivity of the Schmidt camera was found to be adequate to
detect the presence of a test pattern in an incident illumination of 150
microfoot candles. For 200 line resolution of the test pattern, however,
1.5 millifoot candles were required.
ACKNOWLEDGMENT
The authors wish to acknowledge the help and suggestions given by
the members of the television section of RCA Laboratories Division,
particularly that of Mr. R. R. Thalner. Part of the work described in
this article was carried out under contract between the Office of Scientific Research and Development and the Radio Corporation of America.
www.americanradiohistory.com
FIELD TELEVISION t
BY
R. E. SHELBY AND H. P. SEE
Engineering Department, National Broadcasting Company, Inc.,
New York, N. Y.
Summary-A resume is given of the history of NBC Field Television
Operations. The four periods of this history, corresponding to four major
types of pickup equipment, are outlined and the scope of activities possible
during each period is described. Special attention is paid to the fourth
period, just now beginning, which is characterized by a greatly widened
scope of potential field programs made possible by the new Image Orthicon
camera. Some of the characteristics of this new camera, as they affect field
operation, are discussed and experience in its use is described.
A
recapitulation of NBC television programs reveals that 40%
of the program hours broadcast between the opening date
of the public service, April 30, 1939, and December 31, 1945
were originated by remote pickup. A total of 1167 program hours
were devoted to field events during that period although, because of
the war, no activity was recorded for the 161/2 months between the
middle of May 1942 and the first of October 1943. Television programs
originating outside the studio have always been considered to hold an
important place in a well- rounded program service, and in television
development work early attention was given to providing facilities for originating such programs. Progress in this line of work has
been reported from time to time in the technical literature#; therefore
the present report will not give the history of these developments in
detail. It is intended rather to give an overall picture of the present
status with some review of the past work which has lead up to the
present state of the art.
Field operations for NBC television broadcasting is divided into four
rather distinct periods, determined principally by the characteristics
and capabilities of the pickup equipment available during each period.
It is true that much progress was continuously being made at all
stages in the development of this service on matters of technique and
operating procedures, but it is, nevertheless, also true that the characteristics of the pickup equipment were the major factors in determining
the scope of field television operation. The four periods referred to
above may be characterized as follows in terms of the pickup equipment
available in each:
Decimal Classification: R583.17.
f Reprinted from RCA REVIEW, March, 1946.
# See Various Footnotes.
53
www.americanradiohistory.com
TELEVISION, Volume IV
54
Iconoscope Studio Type Equipment permanently mounted in
large vehicle.
2. Orthicon Pickup Equipment permanently mounted in large
vehicle.
3. Transportable Suitcase Type Pickup Equipment.
1.
Equipment Employing the Image Orthicon.
The fourth period is the one which we are just entering and it
gives promise of surpassing by far all previous records with regard
to the wide variety of events which will become available for television
broadcasting.
In addition to the pickup equipment employed in field operation,
it is, of course, necessary to provide a suitable means for transmitting
the television signal back to the main studios or to the broadcast transmitter. This may be done by either radio relay circuits or wire lines.
Both means have been used successfully in the past and there is every
indication that we will see the continued use of both means for at least
some time to come. Progress made in microwave radio frequency equipment for a variety of uses during the war will undoubtedly lead to substantially improved radio frequency links for this television relay application in the near future. The present paper will be concerned
principally with the pickup equipment and the program limitations
imposed by it rather than with the problem of the relay link. Each of
the four periods of NBC field television operation will now be considered
in more detail.
4.
ICONOSCOPE STUDIO TYPE EQUIPMENT PERMANENTLY
MOUNTED IN LARGE VEHICLE
Some indication of the rapid development that has taken place in
television pickup equipment may be obtained by examining the facilities available for field operation at the beginning of the NBC television
public service on April 30, 1939. A large van type of vehicle was necessary to house and transport the studio type equipment- cameras, camera
and power cable, microphone cables, interconnection cables, and the
large variety of accessories required to do a proper job on a field set-up.
Despite the size of these vehicles (a second vehicle of about the same
size housed the permanently mounted radio relay transmitter) , they
were relatively efficient and mobile. (See Fig. 1) Their size and
weight did prevent their reaching marshy or sandy locations, and some
areas where parking was at a premium were difficult to reach. They
were, however, capable of moving normally in city traffic and on the open
road. The great disadvantage from an operating standpoint lay in the
fact that the equipment was, of necessity, permanently mounted in the
www.americanradiohistory.com
FIELD TELEVISION
55
vehicles. The cameras were equipped with 250 feet of camera cable
and this accordingly was the radius of action from the vehicle housing
the control equipment.
The original equipment installed in this large vehicle employed
an iconoscope camera and studio type rack -mounted amplifiers, control
equipment and synchronizing signal generator. At the time this equipment was built, the iconoscope was (and in some respects still is) the
most satisfactory type of direct television pickup tube available. With
medium and high levels of illumination, it produced highly satisfactory
pictures and for day -time out -door field television operation where the
incident illumination did not fall below several hundred foot candles,
Fig.
1- Telemobile Units in Rockefeller Plaza,
New York City, 1939.
the results obtained were generally good. In addition to the fact that
the pictures obtained under conditions of very low light levels were
degraded by excessive amounts of "dark spot" signal, edge flare, and
other defects, it was soon found that the iconoscope had additional
limitations in field operation which were not serious in its use in the
studio. As regards the use of iconoscope cameras, studio operation has,
in effect, a threefold advantage over field operation. In the studio, the
lighting is under control of the operator and may be modified to suit
the needs of the scene being televised; scenery, back -drops and drapes
may be employed having light reflection characteristics suitable for
www.americanradiohistory.com
TELEVISION, Volume IV
56
the camera; and finally, in studio operation rehearsal will usually have
given the operator a knowledge of the shading problem to be encountered as the program progresses. On the other hand, in field operation
-especially outdoors -the lighting is generally not subject to any
control, and may fluctuate over a wide range; the scene being televised
in many instances may have an unfavorable background or direction
of the lighting may be unfavorable; and usually the event being televised is spontaneous and unrehearsed so that the operator has little
warning of the changes to be encountered in shading. During this
period of the NBC field television operation, the fact that peak field
program hours occurred between the months of June and September
Fig.
2-Ceremonies
at Opening of New York World's Fair, April 30, 1939.
(Television Camera on Platform at Extreme Right.)
was not because most field events suitable for television pickup occurred
during those months, but rather because most of the outdoor events
of this kind occurred at that time. The field program curve tapered off
in November during the football season and took a sharp drop as the
gridiron season ended. During the winter months, an occasional outdoor program was originated and a few indoor programs were attempted using added illumination for the benefit of the iconoscope
camera. Many events which would have made good television programs
could not be transmitted because of the amount of light required by
the iconoscope camera.
www.americanradiohistory.com
FIELD TELEVISION
57
Despite the limitations imposed by the equipment in use, several
hundred interesting and timely television programs were presented
originating at locations as far as 28 miles from the main studios.
Although the iconoscope suffers from relatively low sensitivity in comparison with other types of pickup tubes, when an adequate light level
is available, the overall quality of the picture obtained with an iconoscope camera is probably superior to that from any other type of pickup
device so far used in the program service.
The outstanding program originated with this first field pickup
equipment was the inauguration of the television public service on
April 30, 1939 when the late President Roosevelt was televised during
the opening ceremonies of the New York World's Fair. (See Fig. 2)
A program more typical of field operation with this equipment, however, was the pickups of tennis matches at the Westchester Country
Club at Rye, New York during the summer of 1939.
ORTHICON PICKUP EQUIPMENT
A major revision in the field pickup equipment was made during
the month of September 1939 when a new camera employing the
orthicon type of pickup tube was substituted for one iconoscope
camera in the mobile unit. For some time thereafter, the unit was
operated with a combination of one orthicon camera and one iconoscope camera. The importance of this change in equipment lay in the
relatively greater sensitivity of the orthicon tube. Although a direct
comparison in sensitivity between the orthicon and the iconoscope is
difficult because of their differing contrast characteristic, the orthicon
has an effective sensitivity between 3 and 10 times that of the iconoscope for scenes of low incident illumination. In addition, it is essentially free from spurious signals, such as the "dark spot" in the iconoscope. With the orthicon camera, the scope of television field activity
was immensely widened. Many events previously unavailable because
of the lighting problem now could be satisfactorily televised. The
second halves of football games played in the late Fall were now made
entertaining as television program fare whereas with the iconoscope,
particularly on overcast days, the pickup had been very unsatisfactory.
Even with the orthicon camera, in late November and early December,
the light condition near the end of the games sometimes was such that
the picture quality was seriously degraded. In some instances, the light
level dropped so low that floodlights were turned on for the benefit of
the players and spectators and under these conditions the orthicon
gave very satisfactory results. Another important advantage of the
www.americanradiohistory.com
TELEVISION, Volume IV
58
orthicon, in comparison with the iconoscope, is the smaller mosaic
size. The area of the mosaic of the orthicon is only approximately
one-fourth that of the iconoscope, and this permits the use of substantially smaller lenses to obtain the same angle of view. This is a fairly
important item in field operation where a large assortment of lenses
must be provided to obtain various viewing angles depending upon the
available camera location and the type of coverage contemplated.
Perhaps the most important class of programs made available for
the first time by use of the orthicon camera was the large class of
indoor sporting events such as boxing, wrestling, ice hockey, basketball,
indoor track, etc. Nearly all of these events are presented under conditions of lighting which are satisfactory for the orthicon camera but
Fig.
3- Orthicon Camera in Use in Madison
Square Garden, New York.
too low for useable pictures with the iconoscope camera. The impor-
tance to television broadcasting of the availability of events of this
kind can hardly be over-emphasized. They have formed an important
part of the total television program service ever since the first orthicon
camera was placed in service. Figure 3 shows the orthicon camera in
use for program pickups that are typical of the enlarged scope afforded
television field operation by the use of this important development.
Although the orthicon has substantially greater sensitivity than
the iconoscope and is essentially free from the spurious "dark spot"
signal which plagues the iconoscope, it does have certain disadvantages
compared to the earlier, less sensitive type of pickup tube. One of the
principal disadvantages is its tendency to "charge up" when subjected
to light intensity exceeding a certain threshold value. This phenome-
www.americanradiohistory.com
FIELD TELEVISION
59
non is due to the inability of the low velocity electron scanning beam
to completely discharge areas of the mosaic subjected to high light
intensities exceeding a certain threshold value determined by the beam
current. This weakness produces especially annoying effects on indoor
pickups when photographers' flashbulbs are set off in the field of vision
of the camera. This invariably occurs during climactic episodes of the
event being televised and frequently portions of the most exciting
action are lost to the view of the television audience before stable operating conditions can be re- established in the orthicon pickup tube.
Trouble is also encountered in televising outdoor events in bright sunlight when a portion of the scene is in shadow and the rest in sunlight.
Changes in the contest being televised which necessitate sudden and
rapid panning of the camera from dark shadow to bright sunlight
frequently lead to this "charging up" or blocking effect. It is minimized by operation of a special control that temporarily raises the value
of the scanning beam current to an abnormally high value in order to
dissipate the excessive charge on the mosaic.
The orthicon has a contrast characteristic (frequently referred to
loosely as "gamma" characteristic), which is linear over its entire
useful operating range. The iconoscope, on the other hand, possesses a
contrast characteristic which exhibits a substantial amount of saturation in the higher light ranges and is, therefore, roughly equivalent
to a "gamma" of less than unity. This contrast characteristic of the
iconoscope complements rather well the corresponding characteristic
of the kinescope tube used in most receivers so that the overall contrast
characteristic of the system is generally satisfactory. When using the
orthicon, however, it is necessary to provide in the video amplifier
chains associated with the camera a controllable amount of saturation
which will reduce the equivalent "gamma" characteristic to a value
more suitable for the kinescope. Generally speaking, the combination
of orthicon and its "gamma" correction circuit does not produce quite
as satisfactory an overall contrast characteristic as the iconoscope
possesses, but it is hoped that future work will improve the "gamma"
correction circuits. Despite the previously mentioned fact that the
orthicon requires smaller size lenses than the iconoscope for a given
angle of view, the sizes of lenses necessary to provide the desired range
of camera angles are still substantial. The Type 1850 iconoscope requires a lens, of 16" focal length, when the camera is located 70 feet
from a boxing ring, to create images of the contestants which will be
large enough when reproduced on the average receiver to provide
optimum coverage of a prizefight. A lens of longer focal length and
smaller viewing angle makes it difficult for the camera operator to
www.americanradiohistory.com
TELEVISION, Volume IV
60
follow the fast -moving contestants about the ring and often results in
one of the two fighters being out of the picture. The smaller mosaic
in the orthicon tube allows the use of lenses which are approximately
one -half the focal length of those used with the iconoscope camera for
the same viewing angle. With the orthicon camera, boxing at Madison
Square Garden is generally covered with a lens of 8" focal length, the
distance from the camera tó the ring being approximately 70 feet.
Even with the reduction in lens size made possible by use of the orthicon pickup tube, it has not been considered feasible to attempt the use
of lens turrets on these cameras.
TRANSPORTABLE SUITCASE TYPE PICKUP EQUIPMENT
The wider field of television activity permitted by the orthicon
camera was still restricted by the limitation due to the location of
equipment permanently mounted in a vehicle. The camera cables contained four flexible coaxial cables and thirty -two other electrical conductors and were 11/2" in diameter. Storage and transportation difficulties, plus the need for compensating electrical networks to correct
for pulse delay in very long cables were the factors which restricted
the length of the camera cables to a practical value of approximately
250 feet. Shorter lengths of 50 feet each were carried for use where
the longer lengths were not necessary. The 250 foot radius of activity
of cameras was sufficient for most outside pickups when only two
cameras were employed. Additional cameras, which would be those
located at greater distances, could not readily be accommodated because
of lack of space for additional control equipment in the vehicle. Field
pickup service was restricted to coverage of events taking place below
the fourth floors of buildings. Banquets, important meetings, interesting exhibits and panoramic scenes are among the potential programs
which were unreachable with this equipment. The need for portable
equipment which could be carried closer to the pickup scene became
apparent early in field operations.
The development of the small iconoscope (Type 1848) transportable
equipments which was contained in eight boxes approximating the size
of suitcases again greatly expanded the scope of the field service. The
boxes, weighing approximately 65 lbs. each, were inter -connected on the
scene of action to form a complete operating chain of television pickup
equipment. This equipment could be used for television pickup either
in a vehicle in which it was transported or removed from the vehicle
and carried to the upper floors of a building or to other locations inaccessible to the vehicle. (See Fig. 4)
G. L. Beers, O. H. Schade and R. E. Shelby, "The RCA Portable Television Pickup Equipment ", Proc. I.R.E., Vol. 28, pp. 450 -458, October, 1940.
1
www.americanradiohistory.com
FIELD TELEVISION
61
The availability of this transportable type of television pickup
equipment again opened up new sources of television programs pre ' viously unavailable, and also increased the ease of operation in some
cases where television pickups had previously been made with equipment permanently mounted in a vehicle. This type of equipment was
used in the Rainbow Room atop the RCA Building in Radio City to
televise the New Year's Eve festivities there in 1940. When the
Republican National Convention of 1940 at Philadelphia, Pennsylvania
was televised and the signals transmitted via coaxial cable to New York
for broadcasting, both the transportable type of pickup equipment and
Fig.
4- Complete
Single Camera Iconoscope Type Transportable Pickup
Equipment.
the equipment permanently mounted in the mobile unit truck were
employed, thus providing a four -camera pickup. The four cameras
were used to pick up both wide angle and close -up scenes inside the
main Convention Hall, for studio type pickup in a small improvised
interview studio and for pickups on the sidewalk in front of the Convention Hall. All sessions of the five -day convention were televised.2.3
2 O. B. Hanson, "Televising a Political
Convention ", RCA REVIEW,
Vol. V, No. 3, pp. 267-282, January, 1941.
3
H. P. See, "Televising the National Political Conventions of 1940,
Jour. Soc. Mot. Pic. Eng., Vol. XXXVI, pp. 82 -100, January -June, 1941.
www.americanradiohistory.com
TELEVISICN,
62
Fig.
5-Transportable
Fig.
6- Televising
Volarme
IV
Iconoscope Camera in Temporary Studio Set -up at
GOP Convention, Philadelphia, 1940.
the GOP Convention, Philadelphia, 1940. (Orthicon
Camera in Foreground, iconoscope Camera on Right.)
www.americanradiohistory.com
FIELI) TELEi'ISION
The photographs of Figures
Convention pickup.
5
and
6
63
show scenes at the Philadelphia
Transportable equipment employing orthicon cameras' s became
available shortly before the war and a two -camera system of this type
has been used for the majority of all NBC field pickups since 1944.'.7
Figure 7 shows the suitcase type equipment for the orthicon cameras
situated on a movable table in a small control room at Madison Square
Garden. This is the normal location for the control equipment during
the televising of events in Madison Square Garden and its use here
7-
Fig.
Television Control Set -up at Madison Square Garden, 1944. (Audio
Amplifier on Left, Transportable Orthicon Control Units on Right.)
illustrates one of the advantages of the transportable type of equipment. Prior to its availability, when televising events in Madison
Square Garden, the large mobile unit vehicle was parked at the curb
outside the Garden and camera cables and power and communication
cables had to be strung in place for each program.
Although the transportable type of equipment has several distinct
advantages, as already indicated, there are some offsetting disa'dvan4 Albert Rose and Harley Iams, "The Orthicon, A Television Pickup
Tube", RCA REVIEW, Vol. IV, No. 2, pp. 186 -199, October, 1939.
5 M.
A. Trainer, "Orthicon Portable Television Equipment ", Proc.
I. R. E., Vol. 30, pp. 15 -19, January, 1942.
° R. E. Shelby, H. P. See, "NBC's Experience with Portable Television
Broadcast Equipment ", Broadcast News, No. 39, pp. 14-21, August, 1944.
7 R. E. Shelby, H. P. See, "NBC and Madison Square Garden ", Television, pp. 2 -3, April, 1945.
www.americanradiohistory.com
64
TELEVISION, Volume IV
tages. In comparison with the older style of rack mounted equipment,
the suitcase type of equipment is highly condensed and the components
crowded rather closely together. This means that maintenance work
and trouble- shooting are sometimes more difficult than in the rack
mounted type of equipment. The large number of interconnecting
cables with their sockets and plugs increases the chance for contact
failure and delays occasioned by loss or damage. The smaller size of the
monitoring kinescopes and cathode-ray oscilloscopes is somewhat of a
handicap in operations compared to the larger size screens used in the
older equipment. Convenient accessibility of operating controls has had
to be sacrified somewhat in the interest of portability. In spite of these
disadvantages, however, there can be no doubt that this type of equipment provides a substantial net gain in television field operation. It is
interesting to note in passing that the extensive amount of work done
during the war on highly compact military type of television pickup
equipment which had to operate under very rigorous conditions will
undoubtedly lead to significant improvements in future designs of
transportable pickup equipment for the television broadcasting service.
IMAGE ORTHICON EQUIPMENT
There has recently been announced a new type of television pickup
tube known as the Image Orthicon which in comparison with any
previously available pickup tube possesses rather startling characteristics, particularly as regards operation at extremely low levels of
illumination. Development work on this tube had been started prior
to the war but it received its greatest impetus as a part of the war -time
research on military television. Technical design details and characteristics of the device have been given in a recent paper' by Dr. Albert
Rose and these will not be repeated here. Very briefly, however, it
consists essentially of an orthicon scanning tube with an image electron
amplifier section in front of the mosaic and a multiple stage electron
multiplier at the output. It is still smaller in size than the Type 1840
orthicon tube, having a photo- cathode area approximately one-quarter
that of the Type 1840 orthicon and one -sixteenth that of the larger
iconoscopes. The sensitivity of the tube is such that it will produce
satisfactory pictures at illumination levels lower by a factor of about
100 than those required for the iconoscope. In addition, it is much
less subject to the "charging up" effect which is so troublesome in the
case of the Type 1840 orthicon.
The importance of this new tube in television field pickups is selfs Albert Rose, P. K. Weiner and H. B. Law, "The Image Orthicon. A
Sensitive Television Pickup Tube ", presented at the I.R.E. Winter Technical
meeting, on Jan. 24, 1946.
www.americanradiohistory.com
FIELD TELEVISION
65
evident. It removes all practical barriers with respect to operation at
low light levels. It seems safe to state that any event which has
illumination adequate for direct viewing by an audience can be televised satisfactorily with the Image Orthicon camera.
The smaller length and diameter of the new tube plus its increased
sensitivity and small photo- cathode area result in a camera substantially smaller than the standard orthicon field camera. The standard
iconoscope and orthicon cameras in use today are considered too large
and heavy for efficient handling in the field. The comparatively large
mosaic areas of the iconoscope and orthicon tubes require optical
systems of appreciable proportions. In contrast, the optical systems
used with the Image Orthicon are about the same as those used on
standard thirty -five millimeter motion picture cameras. Whether
future camera models incorporate the twin lens optical viewfinder, the
kinescope viewfinder or some other view -finding device, the reduction
in the size of lens with the introduction of this tube makes practical
the use of a lens turret. The sizes and weights of lenses included in
the complement for a Type 1840 tube are generally considered to be
approximately half those necessary for a Type 1850 iconoscope. The
Image Orthicon reduces this by approximately one -half again for the
same viewing angles. The lack of a lens turret in some cases in the
past has imposed a limitation upon the latitude of operations and
programming.
There are characteristics of the Image Orthicon which, at least for
the present, partially offset some of its advantages. Its signal -to -noise
ratio is not as good as that of the iconoscope under conditions of strong
illumination, although at low levels of illumination it continues to
produce satisfactory pictures far below the levels at which the signals
from the orthicon and iconoscope are completely submerged in noise.
The lower signal -to -noise ratio of the new tube is generally not noticeable except on scenes which include relatively large dark areas. Test chart resolution in excess of 500 lines has been obtained with the Image
Orthicon, but it has not yet quite equalled the performance of the
better iconoscopes in this respect. Models of the tube produced for
military use and early samples available for tests in television broadcasting possess rather high infra -red sensitivity. This necessitates
the use of optical filters to attenuate the infra -red light when televising
most outdoor scenes in daylight-particularly scenes which include
appreciable amounts of living green foliage.
The Image Orthicon is more sensitive to ambient temperature than
other types of pickup tubes. It does not give maximum performance
until it has warmed up to approximately 100° F., and when used out-
www.americanradiohistory.com
66
TELEVISION, Volume IV
doors in cold weather auxiliary heating units are sometimes needed.
When the tube is too cold, its resolution may be impaired and retentivity of the electrical "image" on the target will be abnormal, producing excessive smearing in the reproduced picture whenever the
camera is panned or when rapid motion occurs in the scene.
One important advantage which the Image Orthicon has over the
Type 1840 orthicon is its ability to handle a very wide range of light
values. If the various electrode potentials are properly set, a change
in the scene from deep shadow to brilliant sunshine in outdoor pickups
is readily accommodated without serious degradation in the transmtted picture and without the necessity for instantaneously coordinated
readjustment of controls, as in the case of the Type 1840 orthicon. In
one outdoor test, the Image Orthicon was adjusted for optimum performance with the iris on the pickup lens set for an opening of
f 32. Without changing any other control, the iris setting was then
changed to f 8, thus increasing the amount of light on the photocathode
by a factor of 16. To the casual observer, at normal viewing distance,
there was no appreciable change in the transmitted picture. In an
indoor test at Madison Square Garden, it was found that a single suitably chosen setting of all controls would give acceptable results when
the Image Orthicon camera was panned from the dimly -lighted outer
fringes of the audience to the brilliantly -illuminated boxing ring at
the center of the arena.
While it is true, as indicated above, that the Image Orthicon possesses great practical flexibility under varying conditions of illumination, it is also true that peak performance will be obtained only when
the settings of the iris and other controls are proper for the brightness
of the scene being televised. Close inspection of the transmitted picture
shows appreciable loss of information in the highlights if the light
image on the photocathode is excessively bright, and signal -to -noise
ratio will be lowered by operating the beam current at a value greatly
above that required to discharge the target. In preparing for a television pickup, after the Image Orthicon has warmed up to normal operating temperature, the entire scene to be televised should be explored
with the camera to determine, to the extent possible, the upper and
lower limits of reflected light to be expected from the scene during the
program. Settings of the iris, beam current, and other controls should
be made in the light of this test, and if wide fluctuations in brightness
are to be encountered, plans should be made for readjusting beam
current and iris opening at proper times during the program.
Television broadcasting experience with the Image Orthicon camera
has included the following pickups:
www.americanradiohistory.com
FIELI) TELEVISION
67
Herald -Tribune Forum, Grand Ballroom of Waldorf- Astoria
Hotel. Moderate illumination as normally used for audience.
Navy League Dinner, Grand Ballroom of Waldorf- Astoria Hotel.
Moderate illumination as normally used for audience.
1.
2.
Army -Navy Football Game. The day was an unusually bright
one for the season, and offered no test of sensitivity, but did
afford a good comparison with orthicon cameras under conditions of high -level illumination. (See Fig. 8)
3.
Fig.
8
4.
5.
6.
-Image
Orthicon Camera Televising Army -Navy Football Game at
Philadelphia Municipal Stadium, December first, 1945.
Mayor-Elect O'Dwyer of New York City, from his campaign
headquarters in the Commodore Hotel on Election Day. Illumination provided by two 100 watt lamps in a frosted glass shade,
and a 40 watt lamp for relieving shadow areas on faces.
New Year's Eve Celebration in Times Square, New York City.
Illumination as normally present in Times Square at night
from electric signs, street lamps, theatre marquees, show windows, and automobile lights.
-
Memorial Service at Lincoln Monument, Washington, D. C. on
February 12, 1946. A program originated jointly by stations
WABD, WCBW and WNBT and transmitted to New York over
the coaxial cable of A. T. and T. to inaugurate television service
over this cable.
www.americanradiohistory.com
TELEVISION, Volume IV
68
In addition to the foregoing on- the -air programs originated with
the Image Orthicon camera, a number of successful test pickups have
been made. These include the following:
1. Baseball game at Polo Grounds, New York City. The day was
a bright one, and it was found that the minimum lens stop available
f 32-gave more than the optimum amount of light. Depth of focus
was, naturally, no problem. A filter had to be used to reduce the infrared light reflected from the grass of the playing field. Most observers
felt that the Image Orthicon camera had a net advantage over the Type
1840 orthicon camera, which was set up for comparison.
-
Fig.
-Image Orthicon Camera (Foreground), in Comparative Test with
Orthicon Camera, Televising Rodeo at Madison Square Garden.
9
2. Strollers on the Mall at night in Central Park, New York City.
The only illumination was that provided by the normal light fixtures
in the park.
3. Scenes from the Rodeo in Madison Square Garden, New York
City. The dark tan -bark on the floor, and the relatively low light level
employed for many events to avoid blinding the contestants, had made
this a relatively unsatisfactory program for orthicon cameras. The
Image Orthicon camera was able to do an excellent job on all events.
(See Fig. 9)
4. Television Pickup of Standard Sound Broadcast. No special
lighting of any kind was employed, and those in the studio did not
even know that the test was being made.
www.americanradiohistory.com
FIELD TELEVISION
69
5.
"Stunt" Pickups in the Studio. Successful pickups were made
using only the light of a pocket flashlight, or a single candle, or one
match, and also in total darkness, using invisible infra -red illumination.
The importance of the Image Orthicon development to field television broadcasting is emphasized by statistics of operation which show
that in the past more than one -half of the seven hundred odd field
programs have originated indoors under artificial illumination-despite
the fact that many potential indoor programs had to be passed up
because cameras were not sensitive enough to give acceptable results
with the illumination available. In the future, it is probable that the
percentage of indoor programs will go even higher, now that there
are virtually no technical limitations on the televising of such events.
The economic advantages of the new camera are by no means insignificant, since the cost of providing special illumination for some events
televised in the past has been a major item of expense. On the basis
of experience to date, it seems safe to say that the Image Orthicon
represents the greatest single advancement so far made in field
television.
www.americanradiohistory.com
THE IMAGE ORTHICON -A SENSITIVE
TELEVISION PICKUP TUBE*t
BY
ALBERT ROSE, PAUL K. WEIMER AND HAROLD B. LAW
Research Department, RCA Laboratories Division,
Princeton, N. J.
Summary -The image orthicon is a television pickup tube incorporating
the principles of low -velocity- electron -beam scanning, electron image multiplication, and signal multiplication. It closely approaches the theoretics'
limit of pickup tube sensitivity and is actually 100 to 1000 times as sensitive
as the iconoscope (1850) or orthicon (1840). It can transmit pictures with
a limiting resolution of over 500 lines and, if properly processed, is relatively free from spurious signals. At low lights, the signal output increases
linearly with light input; at high lights, the signal output is substantially
independent of light input. The tube is completely stable at all light levels.
The signal output is sufficiently high to make the operation of the tube
insensitive to many of the preamplifier characteristics that are normally
considered significant. The construction, operation, electron optics, and
performance of the tube are discussed.
I. INTRODUCTION
THE importance of sensitive pickup tubes to the success of a
well- rounded television service needs little emphasis. One has
only to be reminded that, insofar as the television pickup tube
is called upon to replace the human observer, the sensitivity of the
pickup tube should match that of the human eye. The demands on a
television service are often more stringent than on news photography,
for example. The latter can, within wider limits, select the times and
conditions under which it will record pictures. The pickup tube, once
committed to transmitting an event, such as a football game, must
steadily transmit pictures under the whole gamut of lighting conditions. It is, accordingly, highly desirable to have a pickup tube which
can transmit pictures both at very low and at very high light levels.
The iconoscope' has transmitted excellent pictures at high light
levels; the orthicon' has operated best at medium light levels. The
Decimal classification: R583.6.
Presented at the 1946 Winter Technical Meeting of the I.R.E. in New
York, N. Y., on January 24, 1946. Reprinted from Proc. I.R.E., July, 1946.
V. K. Zworykin, G. A. Morton, and L. E. F;ory, "Theory and Performance of the Iconoscope," Proc. I.R.E., vol. 25, pp. 1071 -1092; August, 1937.
2 A. Rose and H. A. Iams, "The Orthicon," RCA REVIEW, vol. 4, pp.
186 -199; October, 1939.
*
1
70
www.americanradiohistory.com
IMAGE ORTHICON
71
image orthicon extends the range still further toward lower illuminations by a factor of approximately 100. At the same time, the image
orthicon can operate stably at medium and high light levels. Unlike the
orthicon, it is not subject to transient loss of operation caused by sudden bursts of illumination. The use of the image orthicon in the higher
light ranges is not, however, emphasized relative to the iconoscope or
orthicon. The additional complexity of the tube needed to provide its
increased sensitivity has not yet permitted pictures whose quality
equals the best that the iconoscope or orthicon can transmit.
The present paper describes the construction, operation, and performance of the image orthicon. It is hoped to treat some of the
electron -optical and constructional problems in more detail in separate
papers.
BECTON
GUN
1
1
W.FL«raN
"U.S
Fig.
II.
1-Typical
parts of storage type of pickup tube.
GENERAL DESCRIPTION OF THE IMAGE ORTHICON
The usual storage type of pickup tube (Figure 1) has an electron
gun, a photosensitive insulated surface, referred to as the target, and
a means for deflecting the electron -scanning beam. The scene to be
transmitted is focused on the target on which it builds up by photo emission a charge pattern corresponding to the light and shade in
the original scene. The beam of electrons, generated by the electron
gun, is made to scan the charge image in a series of parallel lines.
While a constant stream of electrons approaches the target, the stream
which leaves is modulated by the charge pattern. A signal plate located
close to the target surface picks up the modulation by capacitance and
feeds it into the grid of the first amplifier tube. The same video signal,
however, appears in the modulated stream of electrons leaving the
target, and if these electrons could be collected on a single electrode,
the signal could be fed through it into an amplifier.
The image orthicon (Figures 2 and 3) has, in addition to the usual
gun, deflection means, and target, three parts that contribute to its
www.americanradiohistory.com
TELEVISION, Volume IV
72
CATHODE (ZERO)
DECELERATING RING
(ZERO)
SECONDARY
ELECTRONS
SECONDARY
ELECTRONS
DEFLECTION
"SIGNAL OUTPUT\
"S
ELECTRODE \
'
(+1500 V)
ALIGNMENT COIL
Fig.
ELECTRON IMAGE
YOKE
/
\
IN-
PHOTO -CATHODE
(O
C
V
?
TARGET SCREEN
CZ ERO)
TARGET
2- Diagram of the image orthicon.
sensitivity and stability. An electron multiplier, built into the tube
near the gun, multiplies the modulated stream of electrons returning
from the target before it is fed into an amplifier. Sensitivity gains of
10 to 100 are thereby made possible. The charge pattern on the target,
instead of being generated by photoemission, is formed by secondary
emission from an electron image focused on the target. The electron
image is released by light from the scene to be transmitted falling on
a conducting semitransparent photocathode and is focused on the
target by a uniform magnetic field. The combination of the higher
photo- sensitivities that can be obtained for a conducting surface than
for an insulated surface, together with the secondary- emission gain
of the electron image at the target, provides another factor of about
fivefold increase in sensitivity. The use of a separate conducting
photo- cathode is made possible by a two -sided target in place of the
usual one -sided target. The two -sided target allows the charge pattern
to be formed on one side and the scanning to take place on the opposite
side. Further, it permits the tube to operate stably over a large range
of scene brightnesses.
Fig.
3
-The
image orthicon.
www.americanradiohistory.com
IMAGE ORTHICON
73
The electron multiplier, two-sided target, and electron -image section will be recognized as elements whose virtues and incorporation
into a pickup tube have been discussed frequently in the literature.1
The image orthicon represents one way of including all three elements
in a useful, sensitive, and stable pickup tube.
3
III.
-2
TYPICAL OPERATING CYCLE
The scene to be transmitted is focused on the semi -transparent
photocathode (Figure 2). Photoelectrons are released in direct proportion to the brightnesses of the various parts of the scene. The photoelectrons are accelerated from the photocathode toward the target by
a uniform electric field and are focused on the target by a uniform
magnetic field parallel to the axis of the tube. The paths of the electrons from photocathode to target are, except for emission velocities,
substantially straight lines parallel to the axis. The electron image,
accordingly, has unity magnification.
The photoelectrons strike the target at about 300 volts, at which
potential the secondary- emission ratio is greater than unity. Because
more secondary electrons are emitted than there are incident photoelectrons, a positive charge pattern is formed on the target, the high
lights corresponding to the more positive areas. The secondary electrons are collected by the fine -mesh target screen.
At the same time that a charge pattern is being formed on one
side of the target, a beam of electrons scans the opposite side. The
scanning beam is of the low- velocity type already described for the
orthicon.2 It starts at the thermionic cathode of the electron gun at
zero potential and is accelerated by the gun to about 100 volts. From
the gun to the target the beam is in an approximately uniform magnetic focusing field. As the beam electrons approach the target they
are decelerated again to zero volts. If there is no positive charge on
the target, all the electrons are reflected and start to return toward
the gun along their initial paths. If there is a positive charge pattern
on the target, the beam electrons are deposited in sufficient numbers
to neutralize the positive charges. The remaining electrons are re3 H. A. Iams and A. Rose, "Television Pickup Tubes with Cathode -Ray
Beam Scanning," Proc. I.R.E., vol. 25, pp. 1048 -1070; August, 1937.
H. A. Iams, G. A. Morton, and V. K. Zworykin, "The Image Iconoscope," Proc. I.R.E., vol. 27, pp. 541 -547; September, 1939.
5 A.
Rose, "The Relative Sensitivities of Television Pickup Tubes,
Photographic Film, and the Human Eye," Proc. I.R.E., vol. 29, pp. 293-300;
June, 1942.
6 P.
T. Farnsworth, "Television by Electron Image Scanning," Jour.
Frank. Inst., vol. 218, pp. 411 -444; October, 1934.
4
www.americanradiohistory.com
74
TELEVISION, Volume IV
fleeted. In this way a stream of electrons, amplitude -modulated by
the charge pattern, is started on its way toward the gun.
The return beam not only starts back toward the gun, but it actually arrives at the gun very near the defining aperture through which
it emerged. An electron beam will follow closely the lines of a magnetic field under the following conditions: (1) that the beam is initially
directed along the magnetic lines; (2) that the beam velocity in volts
does not greatly exceed the magnetic field strength in gausses; (3)
that electric fields transverse to the magnetic field are small or absent;
and (4) that the magnetic lines do not bend sharply. These conditions
are approximately fulfilled in the image orthicon. The beam is shot
into the magnetic field parallel to its lines. The beam velocity in volts
and magnetic field strength in gausses are each in the neighborhood of
100. The only prominent electric field is near the target and parallel
to the magnetic field. The bends in the magnetic field caused by the
transverse fields of the deflecting coils are well tapered.
The return beam accordingly strikes the gun in an area around the
defining aperture which is small compared with the defining aperture
disk, but large compared with the defining aperture itself. Also, the
return beam strikes this surface at about 200 volts and generates a
larger number of secondary electrons than there were incident primary electrons. In short,. the defining aperture disk is also the first
stage of an electron multiplier. Succeeding stages of the multiplier
are arranged symmetrically around and back of the first stage. More
will be said of the multiplier in a following section. Meantime, the
secondary electrons are drawn from the first stage by suitable electric
fields into the succeeding stages. The number of stages, as will be
explained, need not be large to exhaust the useful gain of the multiplier. In its present form, the image orthicon uses five stages of
electron multiplication.
The output current from the final stage of the multiplier is fed into
a wide -band television amplifier in the usual manner. Because this
output current is already at a high level, the required gain of the
amplifier is small compared with that for an iconoscope or orthicon.
The high -level output has other advantages. The performance of the
tube, for example, is not critically dependent upon the noise characteristics and input -circuit parameters of the preamplifier, as is the
case for the iconoscope and orthicon.
The above operating cycle, while somewhat elaborate, is nevertheless easily traceable. On the other hand, the detailed operation of the
parts of the tube does include some interesting and less obvious problems. These will be discussed below.
www.americanradiohistory.com
IMAGE ORTHICON
75
IV. ELECTRON -IMAGE SECTION
The semitransparent conducting photocathode is a well -known
structure for getting photoemission from the side opposite to that
from which the light enters. Photosensitivities several times higher
than those for insulating mosaic surfaces can be obtained.
The use of a uniform magnetic field to focus the electron image is
not only well known but is also one of the simplest methods of electron image formation. Unity magnification, erect image, and good definition at low anode voltage are its characteristics.
V. CONSTRUCTION
OF THE TWO -SIDED TARGET
The two-sided target is perhaps one of the oldest and most frequently proposed structures for improving the sensitivity of a television pickup tube. It makes possible the separation of charging and
discharging processes so that the sensitizing procedures and electric
fields appropriate to each may be incorporated in the tube without
mutual interference. The two -sided target must conduct charges
between its two surfaces but not along either surface. It should have a
conducting element nearby to act as the common capacitor plate for the
separate picture elements.
Most of the attempts to fabricate two -sided targets have centered
on a structure which had discrete conducting elements or "plugs"
embedded in an insulating medium. These have been satisfactory for
testing the properties of a two-sided target but have failed thus far
to provide the uniformity necessary for a commercial tube.
The two -sided target used in the image orthicon is exceedingly
simple and capable of a high degree of uniformity. It is a thin sheet
of low-resistivity glass. The resistivity is chosen low enough so that
charges deposited on opposite sides of the glass are neutralized by
conduction in a frame time (1/30 second). It is chosen thin enough
so that these same charges do not spread laterally in a frame time
sufficiently to impair the resolution of the charge pattern. Thicknesses of five to ten wavelengths of light have been found to be satisfactory.
The thin sheet of glass, about 11/2 inches in diameter, is mounted
flat to within a few thousandths of an inch and spaced about two
thousandths of an inch from a similarly flat fine -mesh screen. The
mounting techniques to achieve these tolerances have been the subject of a considerable amount of work. The problem is especially accentuated when it is realized that the assembled structure must go through
www.americanradiohistory.com
76
TELEVISION, Volume IV
a standard bake -out schedule at about 400 degrees centigrade. Satisfactory assemblies were obtained only after the glass and screen were
each mounted under tension on flat metal rings. The metal ring for
the glass had to be carefully chosen so that the 400 -degree- centigrade
bake -out did not cause the glass either to break or to wrinkle on
cooling.
The fine -mesh screen mounted near the glass target to collect secondary electrons and to act as the common capacitive member for all
of the picture elements has been, itself, a problem of appreciable magnitude. Because the electron image passes through the screen and
impresses the shadow of its wires on the picture, the screen had to be
of extremely fine mesh and highly uniform. In addition, for efficient
operation, it was desirable to have the percentage open area of the
screen 50 per cent or greater. The finest commercial screen available
during the early development of this tube which had even reasonable
uniformity was a 230 -mesh per linear inch, woven -wire, stainless -steel
screen. It had 47 per cent open area and could be etched to about 60
per cent open area. The 230 -mesh screen was, however, readily resolved
in the transmitted picture and limited the resolution objectionably.
In contrast to this screen, a technique was developed for making
fine -mesh screens with 500 to 1000 meshes per linear inch, an open
area of 50 to 75 per cent, and an accuracy of spacing comparable with
that of a ruled optical grating. These screens have made possible the
transmission of pictures with high definition and substantial freedom
from spurious signals.
VI. OPERATION
OF THE TWO -SIDED TARGET
Figure 4 shows the potentials' of the two sides of the glass target
during a typical charge- discharge cycle. In Figure 4 (a) the tube has
been in the dark. The scanned side of the target has been brought to
zero volts by the scanning beam. The picture side also is at zero volts
as a result of leakage to the scanned side. The fine -mesh screen for
collecting secondary electrons is held at + 1 volt. Figure 4(b) shows
the target potentials after exposure to light for a frame time. The
picture side of the glass has been charged to + 1 volt by the electron
image. The scanned side of the target also has been brought up to
+ 1 volt by capacitive coupling to the picture side. In Figure 4(c),
7 For simplicity, the emission velocities of the thermionic and secondary
electrons are taken to be zero and the contact potentials of all surfaces are
taken to be the same. Including finite emission velocities and contact potential differences would merely shift the values of the potentials shown in
Figure 4 without affecting the argument.
www.americanradiohistory.com
IMAGE ORTHICON
77
the beam has just scanned the target, bringing the scanned side down
to zero volts and the picture side down almost to zero volts by its
capacitive coupling to the scanned side. The "almost" results from the
fact that there is a positive charge on one side of the glass and a
negative charge on the other, constituting a charged capacitor. If,
therefore, the scanned side is brought to zero volts, the picture side
must be positive by an amount equal to the picture charge divided by
the capacitance between the two sides of the glass. This turns out to
be small compared with the + 1 volt to which the target as a whole
has been charged. In particular, it is shown to be 0.01 volts in the
illustration chosen. During the next frame time the charges on the
)CREEN
i(;"
TUE
SIDO
}
I
SCANNED
3iDE
;
A. DE /ORE SCANNING AND EXPOSURE
S.
EE /ORE SCANNING AND
EXPOSURE
PATEE
rl
D. Y30 SECOND
EATER AND JUST
WORE NEWT EYPOSUIIE
Fig.
4- Target potentials
during a typical scanning cycle.
two sides of the glass unite by conduction to wipe out the potential
difference between the two sides. Figure 4 (d) shows the potentials at
this time, and by comparison with Figure 4 (a) the target has returned
to its initial state ready for another cycle.
In the above cycle, the charging by the picture, discharging by the
beam, and leakage between the two sides of the glass were described
as events in series. Actually, of course, all three events occur simul-
taneously and steadily.
It may be remarked, in passing, that the choice of a glass with too
high a resistivity (that is, a leakage time constant greater than a
frame time) tends to allow charge to accumulate on the picture side.
www.americanradiohistory.com
78
TELEVISION, Volume IV
For sufficiently high resistivities, an objectionable loss of signal, as
well as spurious after -images, are encountered.
VII. AN
ELECTRON-OPTICAL PROBLEM
It has been found that, for good operation over a large range of
scene brightnesses, the fine -mesh screen potential should be kept low,
about + 1 volt. This means that the glass target potential can swing
only between the narrow limits of zero volts, to which the scanning
beam charges it, and + 1 volt, to which the picture can charge it as
limited by the potential of the fine -mesh screen. The maximum signal
output is proportional to the maximum potential swing of the target
(e.g., + 1 volt as above). It is important; therefore, in order to insure
uniform signal output at all points on the target, to have the limits
constant over the target. The upper limit, + 1 volt, as set by the fine mesh screen, is obviously thé same at all points on the target. The
lower limit, however, is set by the lowest potential to which the beam
can charge the target. If the beam approached the target at all points
with normal incidence, the lower limit would be constant over the
target and equal to zero volts.e The attainment of this "if" is not,
in general, a simple task. The ease with which the beam can depart
from normal incidence is, perhaps, more suggestive. A few possibilities
will be mentioned.
When the beam is shot into the magnetic field by the short electron
gun, it is usually not quite parallel with the magnetic lines. The component of the beam's velocity transverse to the magnetic field lines goes
into helical motion of the beam. The energy of this helical motion is
subtracted from the energy of the beam directed along the magnetic
lines. The latter energy, however, determines the potential to which
the beam can charge the target. Thus if 1/2 volt of energy is absorbed
in helical motion, the beam can charge the target to only + 1/2 volt
instead of to zero volts. This permits the target to swing only between
the limits of + 1/2 volt and + 1 volt. In other words, the maximum
signal output is reduced by half.
Another contribution to the helical motion of the beam may come
from the deflection fields. The electron beam, in the process of negotiating a bend in the magnetic field lines, redistributes some of its
energy into helical motion.° The amount of this energy increases in
general for larger angles of deflection, weaker magnetic fields, and
e Again for simplicity, the thermionicemission energies of the beam
electrons are taken to be zero.
o A. Rose, "Electron Optics
of Cylindrical Electric and Magnetic Fields,"
Proc. 1.R.E., vol. 28, pp. 30-39; January, 1940.
www.americanradiohistory.com
IMAGE ORTHICON
79
higher beam voltages. Here one expects, and finds, the helical energy,
and correspondingly the loss of signal, increasing from the center of
the picture out to the edges.
Helical motion introduced into the beam is fortunately a removable
defect. One has only to introduce a second source of helical motion of
equal amplitude and opposite phase. To correct for helical motion
resulting from misalignment of gun and magnetic field, an adjustable,
small (in magnitude and physical extent) transverse magnetic field is
introduced at the exit end of the gun. To correct for helical motion
resulting from the deflection fields, a second source, whose contribution also increases from the center of the picture to the edges, is introduced near the target. This source is the component of the electric
field of the decelerating ring transverse to the axis of the tube. The
relative phases of the helical motions resulting from the deflection coil
and decelerating ring can be adjusted for cancellation by sliding the
coil along the axis of the tube. In practice, once a design of the tube
and coil has been decided upon, this can be fixed.
What is of particular interest in this problem is the delicacy of
adjustment necessary for good performance. A 100 -volt beam must be
generated, deflected, and corrected in such manner that it approaches
all points on the target with not more than a tenth of a volt energy
"squandered" in helical motion.
VIII.
ELECTRON MULTIPLIER
In spite of the variety of electron multipliers offered by the literature, it was thought desirable to add still another to the list one
which was more nearly suited to the requirements of the image orthicon. A brief consideration of the diffuse spray of secondary electrons
emerging from the first multiplier stage (defining- aperture disk) suggests immediately the difficulties of getting all of them to enter the
relatively narrow mouth of the more conventional electron multipliers.
This is particularly true because it was desirable, for other reasons,
to retain the axial symmetry of the electric field in front of the first
stage. To focus the secondary electrons into a narrow -mouth multiplier might very well require objectionably strong assymetric electric
fields. Once committed to the symmetry of fields, one is also committed to a relatively large entrance opening for the second stage of
the multiplier because the secondary electrons spray out symmetrically
or "fountain- wise" from the first stage.
It was found to be relatively easy to arrange for substantially all
of the secondary electrons from the first stage to strike the large
annular -disk second stage shown in Figure 2. The arrangement con-
-
www.americanradiohistory.com
80
TELEVISION, Volume IV
sisted of surrounding the first stage with electrodes all at lower potential than the first stage, with the one exception of the second stage.
In this way the electrons were offered two alternatives : to return to
their place of origin, the first stage, or to land on the second stage.'°
Energetically the electrons could return to the first stage, since they
were emitted from it with a few volts of spare energy. But to return
to the first stage, the electrons must approach it at nearly normal
incidence or, more accurately, with all but their emission energy
directed normal to the surface. The brief excursion of the electrons
into the strong dispersing field provided by the more positive second
stage makes the probability of such return small. The secondary electrons from the first stage accordingly quickly find their way to the
second stage.
Here the problem is to multiply the electrons again and send them
on to a third stage, and so on through a number of stages to the final
collector. The use of a series of parallel -screen multipliers is well
suited geometrically to the problem, but the efficiency of the screen type multiplier is low. That is, for a secondary -emission ratio of four,
the gain per stage is only about two. The "pinwheel" type of multiplier
shown schematically in Figure 2, on the other hand, has an efficiency
of 80 to 90 per cent. By inspection it is evident that the electrons
incident on a "pinwheel" see an almost opaque surface. There are no
holes, as there are in the screen -type multiplier, through which electrons are lost. The secondary electrons, however, readily pass through
the blades toward the succeeding stage. They are helped in their path
by the coarse -mesh guard screen which shields them from the suppressing action of the negative potential of the preceding stage. Succeeding stages have their blades opposed to accentuate their opacity.
The operation of the multiplier was found to be uncritical to electrical
adjustment and mechanical alignment. Both these features are highly
desirable to simplify the construction and operation of an otherwise
complex tube.
Total gains of 200 to 500 are readily obtained for the five-stage
multiplier. These gains are usually more than sufficient to exhaust the
sensitivity possibilities of electron multiplication. The "useful" gain
obtainable with electron multiplication is discussed in the following
section.
IX. SENSITIVITY AND SIGNAL -TO -NOISE RATIO
It was pointed out in the introduction that the image orthicon
1° The third possibility, that of retaining their freedom in space, is
usually of negligibly short duration.
www.americanradiohistory.com
IMAGE ORTHICON
81
derives its increased sensitivity over the iconoscope and orthicon from
(1) the higher photosensitivity of a conducting photocathode relative
to that of an insulating mosaic; (2) the multiplication by secondary
emission of the electron image at the target; and (3) the use of an
electron multiplier for the signal current. The gain from (1) and
(2) is about a factor of five. It must be remembered that this factor
reflects more the state of the art of making photosensitive surfaces
than any intrinsic limitations. The gain from (3) is a function of
the signal -to -noise ratio in the transmitted picture. The term "noise"
as used here refers to the more or less fundamental current fluctuations associated with amplifiers or generated in the pickup tube.
These fluctuations give rise to a masking effect, often referred to as
"snow ", in the transmitted picture. The video signal current must
exceed the noise current before a picture can be seen. The noise currents, therefore, set the threshold scene brightness that a pickup tube
can transmit; they also define the scene brightness required for the
transmission of good pictures, that is, pictures with high signal -tonoise ratios.
The performance of the iconoscope and orthicon is limited by the
noise currents in the first tube of the television preamplifier. The
performance of the image orthicon is limited by the much smaller
noise in the scanning beam. The multiplier, accordingly, provides a
useful gain in sensitivity up to the point at which the shot noise in
the scanning beam is made equal to, or slightly greater than, the
noise current in the preamplifier. The usual preamplifier noise currents' is 2 X 10 -9 ampere for a 5- megacycle bandwidth. The shot
noise in the scanning beam is (2eJf f )1 /2 = /1/2 X 10-e ampere for the
same bandwidth, where I is the scanning -beam current in amperes.
The "useful" multiplier gain is, therefore,
2X10 -9
11/2 X
10-e
2X10 -3
I1/2
A more convenient way of expressing this gain is to make use of the
relation between the scanning -beam current and the maximum signal to -noise ratio that can be obtained when the beam is fully modulated.
Under these conditions, the maximum signal is the beam current itself ;
the noise associated with this signal is the shot noise in the beam;
and the signal -to -noise ratio R is given by
11 H. B. DeVore and H. A. Iams, "Some Factors Affecting the Choice
of Lenses for Television Cameras," Proc. I.R.E., vol. 28, pp. 369 -374;
August, 1940.
www.americanradiohistory.com
TELEVISION, Volume IV
82
1
R
I1/2 X
10-6
-Il/2X10".
With this relation, the useful gain of the multiplier may be written as
2000 /R. Some comments and caution are needed in the application of
this gain expression.
The useful gain was computed for 100 per cent modulation of the
scanning beam. In practice, for medium- and high -light pictures,
modulations in the neighborhood of 50 per cent are realized. The
lowered modulation results, for the most part, from the fact that all of
the electrons that strike the target do not stick some are reflected
or scattered back. Further, for low -light pictures, near threshold, the
modulation is still lower because the potential swing of the target is
smaller than the emission velocities of the electrons in the scanning
beam
only the higher -velocity electrons can land. Whatever the
source of lower modulation, the useful gain is reduced in proportion
to the modulation.
With the above limitations, the useful gain of the multiplier is of
the order of 20 for a high -light picture and of the order of 200 for
a low -light picture. The combined gain of the electron -image section
and the multiplier make the image orthicon from 100 to 1000 times as
sensitive as the iconoscope or orthicon.
The sensitivity of the image orthicon is high enough to make
comparisons with the performance of the eye both significant and
interesting. The image orthicon has approximately the same intrinsic
sensitivity5 as the eye. This means that, for scene brightnesses near the
threshold for the tube, both tube and eye can transmit the same pictures. On the other hand, the greater flexibility of the eye relative
to a television system enables it still to "see" scenes whose brightness
is as little as one thousandth of the threshold scene brightness for
the pickup tube. The eye attains this low threshold by sacrificing
resolution for operating sensitivity.
-
-
X. SIGNAL VERSUS LIGHT CHARACTERISTICS
A representative curve for the video signal as a function of light
is shown in Figure 5. Three equivalent abscissa scales are shown for
convenience in referring to scene brightness, image brightness, or
photocathode current. Also, the video signal is given in microamperes
of modulated signal at the target. It is this current which determines
the signal -to-noise ratio. The final output signal is the product of the
www.americanradiohistory.com
IMAGE ORTHICON
83
video signal at the target and the gain of the electron multiplier,
usually several hundred. The multiplier is an almost noiseless device.
The curve is divided, for purposes of discussion, into four parts by
the letters A, B, C, D, and E. These will be considered in order, start-
ing from the left.
The low -light range A -B is particularly simple. Here the signal
out is proportional to the light in, just as it is for the orthicon. At
the lowest point on the curve, the video signal is equal to the shot noise
in the scanning beam. The beam current is adjusted in this range just
to discharge the picture. As point B is approached, higher signals
and signal -to -noise ratios are obtained. At B, the light is just sufficient to cause the target to be fully charged (i.e., to the potential of
.
E
B
aim
t_,
.
,
AI
p-.
,
Fig.
.d
_
5- Signal
p
....
L[..
versus light characteristic.
the fine -mesh screen) in a frame time of 1/30 of a second. One would
ordinarily expect that increasing the light level beyond B would tend
to saturate the transmitted picture. The high lights would remain
constant in amplitude in this range; the low lights would continue to
increase and tend to make the entire picture white. This is what one
ordinarily would interpret from Figure 3. Actually, pictures transmitted by the image orthicon in the range B -C have, except for large
black areas, the same or improved contrast. The explanation follows.
Figure 6 (a) shows the transmitted picture of a single spot of light
whose brightness is located at B. The picture is normal. Figure 6 (b)
shows the transmitted picture of the same spot illuminated to ten
times the previous brightness. One sees in this figure that the signal
output did not change for a tenfold increase in original picture
www.americanradiohistory.com
84
TELEVISION, Volume IV
(a)
(b)
Fig.
6- Transmitter picture of light spot at low and at high spot brightness.
brightness, that the contrast of the spot is maintained in the immediate neighborhood of its boundaries, and that the rest of the background, supposedly black in the original, has begun to lighten up. The
black halo surrounding the light spot in Figure 6 (b) is the key to the
www.americanradiohistory.com
IMAGE ORTHICON
85
preservation of good picture contrast in the B -C range. This halo is
formed by low- velocity secondary electrons originating in the light
spot and scattered into the immediate neighborhood of the light spot.
Where they land, they tend to keep the target charged negatively and
to counteract the effect of stray light, tending to wash out the picture.
In brief, the brighter areas in the B-C range tend to maintain their
potential higher than neighboring less- bright areas by spraying the
less-bright areas with more low- velocity secondary electrons than they
get in return. While the "halo" effect is unnatural in Figure 6 (b) ,
it is not visible, as such, in the usual fine-detail half-tone picture (see
Figure 8), and serves only to maintain picture contrast.
The "halo" has another useful function. If the spot of light in
Figure 6 (b) is moved rapidly across the field of view, the transmitted
picture is not a continuous white streak as one would expect from an
orthicon or from an image orthicon in the low -light range A -B. The
transmitted picture is a series of relatively sharp tilted images of the
spot separated by 1/30- second intervals. In effect, the sharp tilted
image is not unlike what one obtains from a focal -plane shutter in a
photographic camera. The mechanism for generating the effect is the
discharging action of the halo electrons. When the spot of light is
displaced from an initial position, the halo electrons erase, by discharging, the initial charge pattern. The brighter the light, the more
rapid the erasing action and the more sharply resolved are pictures in
motion.
The second rise in video signal, namely, the range C -D, has an
interesting origin. An outline of the argument for its existence will
be given here. The signals in both the ranges B -C and C-D are determined by the charge accumulated on a picture element just prior to
being scanned by the electron beam. In the range B -C, this charge is
equal to the total charge that the entire target, considered as a parallel plate capacitor, can accumulate divided by the number of picture
elements. In the range C -D, the picture -element charge is the total
charge that an element can accumulate as determined by the capacitance of that element, alone, to the signal plate. If the spacing between
target glass and fine -mesh screen is small compared with the diameter
of a picture element, these two charges are equal and there is no
"second rise" in the C -D range. As the spacing between glass and
screen is increased, the capacitance of the target as a whole decreases
linearly with the reciprocal spacing, while the capacitance of a picture
element alone levels off to a constant value, independent of spacing
and equal to the capacitance of a disk, the size of a picture element,
www.americanradiohistory.com
86
TELEVISION, Volume IV
in free space. The usual spacing is such that the capacitance of a
picture element alone is two or three times the capacitance that would
be computed for the picture element by dividing the number of picture
elements into the total target capacitance.
Thus far, a basis has been established for the separate picture
elements having more capacitance and being able to store more charge
than is possible when these picture elements act together as a complete
target. It turns out, however, that the additional storage capacity
does not become effective until the light is sufficiently intense to charge
the target as a whole in a small fraction of a frame time. Hence, the
flat plateau B -C before the "second rise" C -D sets in. The end of the
second rise, point D, should and does occur when the light is sufficiently
intense to charge the target as a whole in a line time.
Beyond D the signal output curve again levels off and the transmitted picture does not change with changes in scene brightness.
To summarize: in the low light range, the image orthicon acts like
an orthicon; in the high light range, the transmitted picture is substantially independent of scene brightness, the contrast and half-tone
scale being maintained by redistributed secondary electrons on the
picture side of the target. These redistributed electrons have also the
property of tending to keep moving images in sharp focus.
XI.
RESOLUTION
Starting at one end of the tube with a well- focused image on the
photocathode, the picture undergoes three transformations before
emerging from the multiplier at the other end in the form of a modulated signal current. The transformations are, in order: optical image
to electron image, electron image to charge pattern on the target,
charge pattern to modulated stream of electrons in the scanning beam.
Each transformation has been capable separately of resolving over
1000 lines per inch; the combination has resolved well over 500 lines
per inch.
The resolution of the electron image is limited by the emission
velocities of the photoelectrons. The resolution of the charge pattern
on the target is limited, at high lights, in part by the fine -mesh screen,
and ai low lights, in part by the leakage along the glass target. The
ability of the scanning beam to resolve the charge pattern is controlled
by a number of factors, among which are defining aperture diameter,
thermionic- emission velocities, angle of approach to the target, and
magnitude of the potential differences in the charge pattern. The magnetic field strength, once adjusted for focus, has no first -order effect
on the resolution of either the scanning beam or the electron image.
www.americanradiohistory.com
IMAGE ORTHICON
87
On the other hand, the resolution of both the scanning beam and the
electron image improves with increasing electric field strength on the
scanned side of the target and in front of the photocathode, respec-
tively.
An expression has been derived12 for the limiting current density
that may be focused by an electron gun into a spot on a target. This
current density is proportional to the target potential and to the sin2
of the angle of convergence of the electrons approaching the target.
Experience with oscilloscopes and kinescopes has led to high anode
potentials, kilovolts and tens of kilovolts, for the purpose of getting
small spots. It may, accordingly, appear surprising to find even smaller
spot sizes attained in the image orthicon at a target potential of
approximately zero volts. The smaller beam -current densities used in
the pickup tube are only part of the explanation. The larger part is
the difference in the convergence angles of the electrons approaching
the pickup tube target and kinescope screen. For the orthicon type
of pickup tube the sin2 of this angle is near unity, while for the kinescope it is usually 10 -3 to 10 -i. Thus the low- velocity scanning beam
makes up for its low velocity by its large convergence angle.
XII.
PERFORMANCE
Representative pictures transmitted by the image orthicon are
shown in Figures 7, 8, and 10. Figures 7 and 8 are the transmitted
pictures of slides projected on the photo cathode. Figure 10 shows
the results of a test in which a direct comparison was made between
the operating sensitivity of an image orthicon and of a 35- millimeter
camera using Super-XX film. The experimental setup for the comparison is shown in Figure 9. The original subject was illuminated with an
ordinary 40 -watt bulb attenuated with neutral filters. The television
camera was focused on the subject alone and its picture was reproduced on a receiver located alongside the subject. The 35- millimeter
camera photographed simultaneously the original and reproduced pictures. Both cameras used f/2 lenses and an exposure time of 1/30
second. It will be seen from Figure 10 that only in the first exposure,
at 2- foot -lamberts brightness of the subject, do both original and
reproduced pictures appear. At 0.2 foot -lambert only the picture
reproduced by the television camera is present. And, in fact, the
television camera continues to transmit a picture even at 0.02 foot lambert, which is the brightness of a white surface in full moonlight.
12 D. B.
Langmuir, "Theoretical Limitations of Cathode -Ray Tubes,"
Proc. I.R.E., vol. 25, pp. 977 -991; August, 1937.
www.americanradiohistory.com
TELEVISION, Volume IV
88
Fig.
7
-Test pattern transmitted
by image orthicon.
ACKNOWLEDGMENT
The work on the image orthicon has had an extended course.
Throughout, it has profited from the experience and helpful criticism
of many of the writers' associates both in these Laboratories and in
other divisions of The Radio Corporation of America. Much of the
work was made possible by an immediate background of pickup -tube
research, largely as yet unpublished, and contributed by a number of
Fig.
8
-Half tone
transmitted by image orthicon.
www.americanradiohistory.com
IMAGE ORTHICON
KINESCOPE
89
-\ TRANSMITTED
PICTURE
BY IMAGE
ORTHICON
IMAGE ORTHICON
CAMERA
WITH //j LENS
3SMM CAMERA
WITH
SUPER BY FILM
AND % LENS
Fig.
9
LIGHT
SOURCE
-Setup for comparing the sensitivities
of image orthicon
and photographic film.
individuals. Among these are H. B. DeVore," L. E. Flory," R. B.
Janes,15 H. A. Iams," G. L. Krieger," G. A. Morton," P. A. Richards,"
J. E. Ruedy,14 and O. H. Schade.'6 The writers would particularly like
to acknowledge the encouraging direction of B. J. Thompson (now
deceased) and V. K. Zworykin, and the valuable contributions of S. V.
Forgue,14 J. Gallup," and R. R. Goodrich."
The groundwork for the image orthicon had already been laid prior
to the war. Early in the war, effort was directed under an Office of
Scientific Research and Development contract toward developing the
image orthicon in a form suitable for military purposes.
2 -foot
(a)
lamberts
(b)
0.2 foot -lambert
(e)
(d)
0.07 foot -lambert 0.02 foot -lambert
Fig. 10- Comparison of sensitivities of image orthicon and 35millimeter Super XX film. (Incandescent light source.)
13
14
15
16
Formerly, RCA Laboratories.
RCA Laboratories Division.
RCA Victor Division, Lancaster, Pa.
RCA Victor Division, Harrison, N. J.
www.americanradiohistory.com
A
UNIFIED APPROACH TO THE PERFORMANCE OF
PHOTOGRAPHIC FILM, TELEVISION PICKUP
TUBES, AND THE HUMAN EYE*t
BY
ALBERT ROSE
Research Department, RCA Laboratories Division,
Princeton, N. J.
Summary -The picture pickup devices -film, television pickup tube,
and eye -are subject ultimately to the same limitations in performance
imposed by the discrete nature of light flux. The literature built up around
each of these devices does not reflect a similar unity of terminology. The
present paper is exploratory and attempts a unified treatment of the three
devices in terms of an ideal device. The performance of the ideal device is
governed by the relation
(signal-to-noise ratio)'scene brightness = constant
picture element area X quantum efficiency
The three devices are shown to approximate this type of performance
sufficiently well to use it as a guide in treating their common problems.
Simple criteria are derived for characterizing the performance of any one
device as well as for comparing the performance of different devices. For
example, quantum efficiency is used to measure sensitivity; the signal -tonoise ratio, associated with a standard element area, is used to measure both
resolution and half -tone discrimination. The half -tone discrimination of
the eye governs the visibility of "noise" in the reproduced picture and, in
particular, requires that pictures be photographed or picked up at increased
scene brightness when the brightness of the reproduction is increased. The
observation and interpretation of visual "noise" are discussed.
INTRODUCTION
THERE are three picture pickup devices that have separately
been the subject of considerable investigation. These are the
human eye, motion picture film, and television pickup tubes. For
each of these, a large technical literature has been built up relatively
independently of the others. The language, the units, the concepts, and
the conclusions of the separate arts are not in a form that allows them
to be readily compared. This situation is understandable in the early
stages of the arts because the primary emphasis is then to get something- anything -that will transmit a usable picture. As the art progresses, however, interest shifts naturally to an examination of the
Decimal Classification: R583.11
Presented May 10, 1946, at the SMPE Technical Conference in New
York, N. Y. Reprinted from Jour. Soc. Mot. Pic. Eng., October, 1946.
*
j-
90
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
91
theoretical limits of expected improvements. Such an examination is
especially significant because all three devices are subject ultimately
to the same simple statistical limitations arising from the discrete
nature of light flux. The time is opportune for the three devices to
profit from a consideration of their problems in common terms.
Some illustrations will make the present situation clear. In films,
graininess is a familiar concept. Its origin, control, and visual effects
have been treated extensively and for a long time. In pickup tubes,
signal -to -noise ratio is an ever -present consideration for getting pictures of good quality. For human vision, interest has frequently been
centered on the minimum discernible contrast. There is good reason
now to say that graininess, signal -to -noise ratio, and minimum discernible contrast are only three different names for the same property
of a picture pickup device. Again the limiting resolution of film is a
standard and advertised characteristic; the frequency response curve
of a television pickup tube is an important specification of the tube's
performance; the minimum resolvable angle of the eye is a well -known
figure and one which, perhaps, has received more than its just share of
attention. It is obvious that in all three instances, an attempt has been
made to count the number of separate picture elements.
A third illustration concerns sensitivity. There is little need to
remind one of the variety and confusion of sensitivity scales that have
been proposed for film. On the other hand, the sensitivity of a television pickup tube can, with reasonable adequacy, be defined by its
microampere signal output per lumen input. The sensitivity of the
eye has variously, and often with deliberate dramatic emphasis, been
described in terms of the farthest distance at which one can still see a
lighted candle; the order of magnitude of the faintest visible star; the
number of lumens falling on the retina necessary for a visual sensation; and so on. Only recently have there been more fundamental
attempts to measure the sensitivity of the eye in terms of its quantum
:
efficiency.
These illustrations serve to show, first, that the basic properties of
a picture pickup device -resolution, sensitivity, and contrast discrimination -are indeed of common concern to the eye, film, and pickup
tube; and, second, that the specification of these properties has not
enjoyed an appropriately common treatment.
The purpose of the present discussion is to explore the extent to
which such a common or unified treatment is both possible and profitable.
www.americanradiohistory.com
TELEVISION, Volume IV
92
The order of the discussion will be:
(1) The development of the properties of an ideal picture pickup
device;
(2) The examination of eye, film, and pickup tube for the purpose
of finding out how well they approximate ideal performance;
(3) A re- examination of a number of current problems in the light
of (1) and (2).
It will become clear that the performance of an ideal device is completely specified by a single number, the quantum efficiency of its photo
process, taken together with some simple optical relations; that the
performance of eye, film, and some pickup tubes approach sufficiently
close to ideal performance to suggest a unified approach to many of
their current problems and that such an approach leads to simplifying
concepts.
NOISELESS
AMPLIFIER
LENS
SCENE
Fig.
1
PHOTOSENSITIVE
TARGET
-Essential parts
REPRODUCED
PICTURE
of a picture pickup system.
IDEAL PICTURE PICK -UP DEVICE
Figure 1 shows the essential parts of a system for picking up and
reproducing a picture. Attention will be centered on the target of the
pickup device, and, in particular, on one picture element of that target.
A picture element is here taken to be an element of area of arbitrary
size, not necessarily the smallest resolvable area. Let that element have
a length of side h, and absorb an average number N, of quanta in the
exposure time allowed. The absorption of each quantum will give rise
to a separate event such as the release of an external photoelectron, or
an internal photoelectron or the dissociation of a molecule. These are
uncorrelated chance events. For this reason, the average number N
has associated with it fluctuations whose root mean square magnitude
is the square root of the average number. Thtts, if N is taken to be
the measure of the signal, NV' is a measure of the smallest discernible
difference in signal. In particular, the ratio
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
N
Ni/2
93
= N'i2
is the signal -to -noise ratio. We may write, therefore,
Signal-to -noise ratio = R =
and the geometric relation
(1)
N1/2
:
Scene brightness = B
= constant
N
.
(2)
h2
Combination of Equations (1) and (2) yields:
B = Constant
R2
.
(3)
h2
Equation (3) is the characteristic equation for the performance of
the ideal picture pickup device. It must be emphasized that Equation
(3) is not concerned with the particular mechanism used to generate
a picture so long as full use is made of all the absorbed quanta. For this
reason, it is meaningful to inquire whether the performance of such
diverse mechanisms as the eye, film, and pickup tubes can all be described by the same characteristic equation.
Equation (3) defines the scene brightness B required to transmit
a picture having a signal -to -noise ratio R associated with picture
elements of linear size h. It says that the scene brightness must be
increased as the square of the signal -to -noise ratio demanded, and as
the square of the number of lines in the picture, the number of lines
being proportional to 1 /h.
The constant term on the right -hand side of Equation (3) contains,
among other parameters, the quantum efficiency of the photo process.
It is this quantum efficiency* alone which sets the performance range
of the ideal pickup device. The complete constant term will be given
later. For the moment, it will be useful to examine a plot of Equation
(3)
Figure 2 is a plot of Equation (3) for several values of scene
brightness. Figure 2 shows that the signal-to -noise ratio increases
linearly with the size of picture element considered. In particular, there
* If the term "ideal pickup device" were to receive its full emphasis,
the quantum efficiency of the photo process should, of course, be taken to
be 100 per cent. The emphasis here, however, is on the complete utilization
of all absorbed quanta rather than on the absorption of all incident quanta.
www.americanradiohistory.com
TELEVISION, Volume IV
94
is a smallest element which is determined by the smallest signal -tonoise ratio that can be observed. The smallest element would be called
the limiting resolution. The smallest observable signal -to -noise ratio
has often been taken to be unity. Actually, by virtue of its statistical
origin, the smallest observable R is a function of how often one prefers
to have his observations correct. This much is verifiable both from
analysis and from the use of physical instruments as observers. For
a human observer, tests' have been made which suggest a threshold
value of R in the neighborhood of five. Whatever this threshold is, one
may draw on Figure 2 a horizontal line whose intersections with Bl,
B2, and B3 mark the limiting resolutions for the several scene brightnesses.
2
n
0
Up
2
r
¢
W
I
Y,
0
Z
ol
-J
4
V
a
E
pqS
`1
o
E
A9Eat
Z
F
2
3
4
LINEAR ELEMENT SIZE- h
Fig.
2- Performance
X
ICI-
CMS
curves for an ideal picture pickup device.
The complete form of Equation (3) may be readily obtained' from
well -known optical relations and is
B-2.8f'
Here
R2
10-13ft-L.
(4)
th'B
f = the f value (numerical aperture) of the lens
t= exposure time
(seconds)
1 Romer, W., and Selwyn, E. N., "An Instrument for the Measurement
of Graininess ", Phot. Jour., 83, (1943), p. 17.
2
Rose, A., "The Relative Sensitivities of Television Pickup Tubes,
Photographic Film and the Human Eye ", Proc. I.R.E., 30, 6 (June 1942),
u. 295.
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
of the photo process (0 =
cent quantum efficiency)
= quantum yield
1
95
means 100 per
= length of side of element (centimeters)
1 lumen = 1.3 X 10'(' quanta per second (average for
h
white light).
If one takes the hyperfocal distance as a measure of depth of field, the
performance of the pickup device is completely specified by Equation
(4) together with the relation3
=FD
hyperfocal distance
2h
(5)
F = focal length of lens
where
1D
= diameter
of lens.
Complete specification means that one selects the desired values for
the hyperfocal distance, exposure time, signal -to-noise ratio, angle of
view, and size and number of picture elements, and from them computes the scene brightness required.
The scale factors for the curves of Figure 2 are based on Equation
1/30, 0 = 1. These curves show what
(4) with the choice of f = 2,
may be expected from an ideal device with 100 per cent quantum
t=
efficiency.
TELEVISION PICKUP TUBES
No operable pickup tube has yet been reported which completely
fulfills the properties of the ideal pickup device. The effective exposure
time of the image dissector,' or other nonstorage devices, is limited
to a picture element time and such devices are correspondingly insensitive. The performance of the iconoscope5 and orthicon(' is limited by
noise currents inherent in the primary photo process. The image
noise currents inherents in the primary photo process. The image
orthicon' (Figure 3) goes a long way toward removing this limitation
DeVore, H. B., and Iams, H. A., "Some Factors Affecting the Choice
28, 8 (Aug. 1940), p. 369.
4 Farnsworth,
P. T., "Television by Electron Image Scanning ", Jour.
Frank. Inst., 218, 4 (Oct. 1934), p. 411.
Zworykin, V. K., Morton, G. A., and Flory, L. E., "Theory and Performance of the Iconoscope ", Proc. I.R.E., 25, 8 (Aug. 1937), p. 1071.
' Rose, A., and Iams, H. A., "The Orthicon, a Television Pickup Tube ",
RCA REVIEW, 4, 2 (Oct. 1939), p. 186.
' Rose, A., Weimer, P. K., and Law, H. B., "The Image Orthicon, A
Sensitive Television Pickup Tube ", Proc. I.R.E., 34, 7 (July 1946), p. 424.
3
)f Lenses for Television Cameras ", Proc. I.R.E.,
www.americanradiohistory.com
TELEVISION, Volume IV
96
in so far as the high light signal -to-noise ratio of its output is, within
limits, determined by the signal -to -noise ratio in the primary photo
process. It is handicapped, as are the other storage -type tubes, mainly
by' having as much noise in the low light portions of a picture as in
the high lights. Equation (4) may, however, be used to describe the
performance of the image orthicon if signal -to -noise ratio is interpreted to mean the signal -to -noise ratio in the high light portions of
the picture.* The quantum yield of the primary photo process is about
0.01 and the noiseless amplifier to Ibe compared with Figure 1 is its
electron multiplier.
Ì
CATHODE
(ZERO)
°^
SECONDARY
I
"'
DEFLECTION
^
DECELERATING RING
(ZERO)
SECONDARY
ELECTRONS
I
^
YOKE
I
ELECTRON IMAGE
SIGNAL OUTPUT
PHOTO-CATHODE
ELECTRODE
(t 1500 V )
(
-600
V )
TARGET SCREEN
(ZERO)
ALIGNMENT COIL
Fig.
3
TWO -SIDED TARGET
-Image orthicon (schematic).
PHOTOGRAPHIC FILM
One does not readily think of film as having a signal -to -noise ratio.
Yet, the separate grains randomly situated in film are immediately
comparable with the separate and randomly spaced electrons in the
scanning beam of a television pickup tube. And, in fact, a number of
recent objective measurements as well as analyses of graininess have
led to the expression'
AD
- AT = constant
X
a'i'-'
(6)
T
* The beam current used to scan the target must be sufficient to discharge the high light portions of the picture. Under these conditions, the
signal -to -noise ratio inherent in the beam is approximately that of the high
lights. The same beam current, however, scans the low lights and adds
appreciable noise over and above the noise inherent in the low lights.
For summary of literature, see: Jones, L. A., and Higgins, G. C., "The
Relationship Between the Granularity and Graininess of Developed Photographic Materials", Jour. Opt. Soc. Amer., 35, 7 (July 1945), p. 435.
8
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
97
where AD and AT are the r -m -s deviations in density and transmission,
respectively, of an area a of film. ** With the notation of Equations
AT
(1) and (2) ,
=R-' and a-1/2 = h and one may write for film
-'
T
R.= constant X h.
The value of this constant is proportional to
diameter. There is good evidence that, for the
graphic grain, the film speed is proportional to
last two statements combined with Equation (7)
B= constant
(7)
the reciprocal grain
same type of photothe grain area. The
give
R2
h2
just as for the ideal device [Equation (3) ]. One can accordingly use
Equation (4) to describe the performance of film with the understanding that the ratio R2 /h2 is characteristic of film with a given average
grain diameter and changes in R2 /h2 are obtained by use of other films
with different average grain diameters. The quantum yield is the
reciprocal of the number of incident quanta required to make a grain
developable* and from published statements" is in the neighborhood
** Equation (6) obviously cannot hold for values of a in the neighborhood of and less than the grain size. Krevald and Scheffer9 and Raudenbusch" have observed such departures and more recently Jones and Higgins14 have reported them. The problem is further involved by a range of
grain sizes in any one film.
* Strictly, this use of the term "quantum yield" is in accord with its
normal definition only if a grain is made developable by the absorption of a
single quantum. If more than one quantum needs to be absorbed for this
purpose, the process still may be looked upon for noise computations as the
equivalent of the absorption of one quantum because the noise arises mostly
from the random distribution of grains rather than from the fluctuations
in rate of absorption of light quanta.
9 van Krevald, A., and Scheffer, J. C., "Graininess of Photographic
Material in Objective and Absolute Measure ", Jour. Opt. Soc. Amer., 27, 3
(Mar. 1937), p. 100.
io Raudenbusch, H., "Measurements of Graininess and Resolution of
Photographic Film ", Phys. Zeits., 42 15/16 (Aug. 1941), p. 208.
u Silberstein, L., and Trivelli, A., "Quantum Theory of Exposure Tested
Extensively on Photographic Emulsions", Jour. Opt. Soc. Amer., 35 2 (Feb.
1945), p. 93. (The writers of this paper avoid emphasizing the physical
implications of their analysis. At the same time they do interpret their
results to show that the intrinsic sensitivity of film is increased by longer
development times. So far as other measurements have shown that the
increase in speed resulting from long development time is paralleled by an
increase in graininess, the present paper would argue that the intrinsic
sensitivity is unchanged but that the developed grains are made larger by
longer development time.)
www.americanradiohistory.com
TELEVISION, Volume IV
98
of 0.001. The noiseless amplifier to be compared with Figure 1 is the
complete development of a silver grain after only a few silver atoms
have been formed by the action of the light.
HUMAN EYE
Equation (4) is not immediately applicable to the human eye
because there is no way of directly measuring the signal -to -noise ratio
that the brain perceives. It is necessary, therefore, to replace signal to -noise ratio by its equivalent in terms of minimum discernible contrast in the test object viewed. ** The signal -to-noise ratio R has
already been referred to as a measure of the minimum discernible
difference in signal. This allows one to write with reasonable assurance
Minimum discernible contrast = C
=
Const.
R
X 100
per cent.
(8)
To get a value for the constant, let C take on its maximum value,
viz., 100 per cent. This defines the constant to be equal to the minimum perceivable value of R. As mentioned earlier, the measurements of Romer and Selwyn may be interpreted to give a value of
about five. Unpublished measurements by O. H. Schade on television
pictures yield a value of three. The determination of this constant is
of considerable importance in estimating the quantum efficiency of
the human eye and deserves more experimental work. j- For the present
it will be included as an undetermined constant. Substitution of
Equation (8) in Equation (3) gives
B
= constant
1
(9)
C2h2
for the characteristic equation which the eye would satisfy if its performance were "ideal." Equation (9) may be rewritten with the minimum resolvable angle a in place of distance h to make it more readily
comparable with published data. Thus,
BL
**
Contrast is defined as
-BD
X 100 per cent, where BD is the
BL
brightness of a gray test object immersed in a white surrounding of brightness BL.
j- An interpretation of the experimental results of Jones and Higgins"
in which the blending distances and signal -to -noise ratios were measured
for the same films also leads to a value of about five.
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
B
99
1
= constant
(10)
C2 a2
How well the performance of the eye matches Equation (10) may
be seen from Figures 4 and 5. Figure 4 shows a plot'' of C versus a -1
for a large range of scene brightnesses and, as expected from Equation
(10), the data fall closely on 45- degree lines. Data in the immediate
neighborhood of a =1 minute and C = 2 per cent are omitted because
these represent limits to the performance of the eye set bye other than
IV
SCENE
110 -4TO
LUMINANCE)) FOOT LAMBERTS
2TO 100
rERFORMANCE DATA
Foe EYE IN RANGES
%CONTRAST
MINIMUM
RESOLVABLE
ANGLE-
/
100
SCENE LUMINANCE -FOOT LAMBERTS
10'4 10 -3
10 -2
1.o
10'1
10
//
30O
10
3
2'ro40'
//
/
/
-
CONNOR e GANOUNG
O O O .COBB
001
0.03
RECIPROCAL
Fig.
100
0.1
OF
0.3
1.0
E
M055
3.0 Q......rc.r1
MINIMUM RESOLVABLE
ANGLE
4- Comparison
of experimentally observed performance of
the eye with ideal performance.
statistical considerations. The
at high lights, for example, is
elements or cone structure. A
and be slightly modified by, the
smallest angle that the eye can resolve
set by the physical size of the retinal
more precise treatment would include,
shape of the eye curve near its "cutoff"
limits.
12 Connor, J. P., and Ganoung, R. E., "An Experimental Determination
of Visual Thresholds at Low Values of Illumination ", Jour. Opt. Soc. Amer.,
25, 9 (Sept. 1935), p. 287; Cobb, P. W., and Moss, F. K., "The Four Variables of Visual Threshold ", Jour. Frank. lnst., 205, 6 (June 1928), p. 831.
www.americanradiohistory.com
TELEVISION, Volume IV
100
The complete characteristic equation for the eye, from Equations
(10) and (4), is
B
= 1.4
f'k2
X 10 -2
ft -L
a2C2tO
where a is the angle in minutes of arc subtended by a picture element
at the eye and k is the undetermined constant relating C and R.
Figure 5 is .a replot of the data in Figure 4. It is a more complete test
of the characteristic Equation (11) and shows the small range* of the
PERFORMANCE DATA
FOR EYE IN RANGES
i10 -4 TO 10-$EENE
LUMINANCE J FOOT °LAMBERTS
V. CONTRAST
2 TO 100
MINIMUM
RESOLVABLE ANGLE 2 'TO 40'
CONNOR E GANOUNG
E MOSS
0 0 0 Coee
00I
0.03
0.1
A0.3
LV
3.0
lo
B y=
Fig.
5- Replot of data
in Fig. 4.
factor k -/tO from very low to very high lights as well as its actual
value. At low lights the value of k2 /t9 is 2800. If one takes the exposure time t to be 0.2 sec, k2 /0 = 560. It is known13 that at threshold
about 150 quanta (near 5300 A) must be incident on the eye to generate
a sensation. This corresponds to about 500 quanta if white light is
* If the full range of this factor is ascribed to the variation of quantum
efficiency from low to high lights, one is presented in this approach with at
most a ten -to -one variation in sensitivity of the eye from low to high lights
as opposed to the usual statement that the dark adapted eye is 103 to 104
times as sensitive as the light adapted eye.
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
101
used. Various measurements and computations13 of the number of
quanta actually used in generating the sensation vary from 1 to 50,
giving O the range from 0.002 to 0.1 and k the range from 1 to 7.
This range of k is to be compared with the independently obtained
values of five from Romer and three from Schade.
All of the above discussion has been for the purpose of showing
that the performance of the eye satisfies the same type of equation as
that obtained for the ideal pickup device. The quantum efficiency,
assuming k =5, is about 5 per cent at low lights and about 0.5 per
cent at high lights. The noiseless amplifier to be compared with Figure
1 may be some catalytic or triggering action induced by the absorption of quanta in the retina.
GENERAL DISCUSSION
The classes of picture pickup problems that have received frequent
attention are:
(1) Specification of the performance of any one pickup device;
(2) Comparisons of the performance of two pickup devices of the
same kind, or of different kinds;
(3) The setting of standards of performance for pickup devices
that would "satisfy" the human eye.
The particular problems to be discussed here are intended only to be
representative, rather than exhaustive.
SENSITIVITY
The simplest test for the relative "sensitivities" of two devices is
accomplished by observing the lowest scene brightnesses at which they
can still record a picture. This type of test is immediately subject to
the questions: Was the exposure time the same for the two devices?
What were the relative lens speeds used? What were the relative
picture sizes? While these are obvious questions, there is no essential
13 References to number of quanta used by the eye for a threshold
sensation:
1 quantum -DeVries, H. "The Quantum Character of Light and Its
Bearing on Threshold of Vision, Differential Sensitivity and Visual
Acuity of the Eye ", Physica, 10, 7 (July 1943), p. 553.
2 quanta -van der Veldon, H. A., "The Number of Quanta Necessary for a Light Sensation for the Human Eye ", Physica, 11, 3 (Mar.
1944) , p. 179.
4 quanta- Hecht, S., "Quantum Relations of Vision ", Jour. Opt.
Soc. Amer., 32, 1 (Jan. 1942), p. 42.
25 to 50 quanta-Brumberg, E. M., Vavilov, S. I., and Sverdlov,
Z. M., "Visual Measurements of Quantum Fluctuations ", Jour. Phys.
(Russian), 7, 1 (1943), p. 1.
www.americanradiohistory.com
102
TELEVISION, Volume IV
reason to pause here. The further questions of relative angles of view,
numbers of picture elements and signal -to -noise ratios are of equal
importance. In brief, the comparison of the sensitivities of two devices
is not meaningful until the devices and their transmitted pictures are
completely specified. But complete specification, as pointed out earlier,
means that the quantum efficiency of the primary photo process is the
only parameter that can vary the range of performance of an ideal
device. And accordingly, the quantum efficiency is the measure of
sensitivity. Not all devices, however, are ideal. For this reason, a
more general figure of merit, based on Equation (4) is here proposed.
The figure of merit is proportional to the reciprocal of the total light
flux required to produce a picture of specified signal -to-noise ratio and
resolution in a specified exposure time. The figure of merit is
f2
BA
where f is the numerical aperture of the lens, B is the scene brightness,
and A the area of target. If the performance of the device is ideal,
the figure of merit becomes also a measure of its quantum efficiency.
It is recognized that the signal -to -noise ratio of a given picture is
not a readily accessible parameter and that there is no general agreement on a measure of resolution. The evaluation of sensitivity, however, can be no more accurate than the knowledge of these parameters.
It is of interest to apply the figure of merit to the interpretation of
several familiar problems.
FILM SPEEDS
Consider the range of film speeds advertised. For the most part.
these are films of the same quantum efficiency but different grain size*
and, for the most part, the essential sensitivity performance of these
films is the same. A simple example will make this clear. Two films,
A and B, are rated at the relative speeds of one and four, respectively.
Their quantum efficiencies are equal and the average grain area of B
is four times that of A. Normally, one might say that B can pick up
a scene with one -fourth the light required by A. While this statement
is true, it is misleading. Suppose one wants the same resolution and
depth of focus in both pictures. This would mean a film area of B
* The relative speeds of Super XX and Eastman High Resolution plates
are in th ratio of about 104 to 1. The relative grain areas are in the ratio
of about 103 to 1.
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
103
four times as large as A to match resolutions and consequently a lens
for B stopped to twice the numerical aperture (f /number) of the lens
for A to match depth of focus. The result is that both films require
result which
the same scene brightness to transmit the same picture
or
efficiencies
quantum
equal
their
from
anticipated
been
have
could
picture
transmitted
same
the
for
merit
evaluated
from their figures of
quality.
-a
COMPARISON OF' EYE AND FILM
An interesting application of the figure of merit is to the taking
and viewing of motion pictures. For obvious reasons, the quality of
the motion picture (signal -to -noise ratio and resolution) is aimed at
equaling or exceeding the quality of picture which the eye can transmit
at the brightness of the motion picture screen. For equal quality one
can anticipate that the figure of merit for the eye would be at least
a factor of twenty better than for film based on relative quantum
efficiencies. But, in so far as film aims at better quality and attempts
to compensate for some of its limitations by projecting pictures at a
higher than unity gamma, an additional factor can be expected in
favor of the eye.
Table 1 gives approximate values for f, B, and A to be associated
with the camera that takes the pictures and the eye that views them.
The area of target used for the eye is that area of retina occupied by
the motion picture at a 4:1 viewing distance. The figure of merit for
the eye is seen to be 250 times that for film.
Table
B,
Eye
Film
2.5
2
1
Ft -L
A, In.2
10
0.03
0.5
100
f2
BA
20
0.08
COMPARISON OF FILM AND TELEVISION PICKUP TUBES
shows the setup for comparing the low light performance
of Super XX film and an image orthicon. The original subject was
illuminated with an ordinary 40 -watt bulb attenuated with neutral
filters. The television camera was focused on the subject alone and its
picture was reproduced on a receiver located alongside the subject. The
35 -mm camera photographed simultaneously the original and reproduced pictures. Both cameras used f/2 lenses and an exposure time of
Figure
6
www.americanradiohistory.com
104
TELEVISION, Volume IV
1/30 sec. Figure 7 shows the results. Only in the first exposure, at
foot -lamberts brightness of the subject, do both original and reproduced pictures appear. At 0.2 ft -L only the picture reproduced by the
television camera is present. And, in fact, the television camera continues to transmit a picture even at 0.02 ft -L which is the brightness of
a white surface in full moonlight.
One interpretation of this test is that the image orthicon is 50
times as "sensitive" as Super XX film because it can transmit a picture with 1 /50th of the light required by the film. The present paper
argues against this interpretation and sets the factor at ten. This
is based on the fact that the area of target (photo-cathode) used by
the image orthicon was five times the area of the 35 -mm film frame.
2
KINESCOPE
-
PICTURE
TRANSMITTED
BY IMAGE
ORTHICON
35 mu CAMERA
WITH
%
SUPER XX FILM
AND C/2 LENS
Fig.
ORIGINAL
SUBJECT
LIGHT
SOURCE
IMAGE ORTHICON
CAMERA
WITH f/2 LENS
-Setup for
6
comparison of low light performance of
Super XX film and an image orthicon.
If the cameras were to be set up to transmit the same picture with
the same angle of view and depth of focus, the lens on the image
orthicon would have to be stopped 51/2 times the numerical aperture
of the lens for the 35 -mm camera. The threshold scene brightnesses
would then be in the ratio of 10:1.
GRAININESS AND SIGNAL -TO -NOISE RATIO
An excellent survey of the extensive history of the problem of
specifying the graininess of film has recently appeared by Jones and
Higgins.' In this paper and in a second one14 they undertake to compare two general methods of measuring graininess. Method I, which
they describe as a psychophysical measurement, records the distance
14 Jones,
L. A., and Higgins, G. C., "Photographic Granularity and
Graininess ", Jour. Opt. Soc. Amer., 36, 4 (Apr. 1946), p. 203.
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
105
from the observer at which the grainy film appears to blend into a
uniform surface. (After introducing an observer for his special
virtues as a measuring instrument, he is ushered part way out again
by the device of normalizing his results with a standard test chart.)
Method II is an objective measurement of the transmission or density
fluctuations of the film using scanning apertures of various sizes.
Broadly, Jones and Higgins argue (1) that the objective measurements should match the "blending distance" measurements in order to
INCANDESCENT
LIGHT
SOURCE
FOOT
0.2
LAMBERTS
0.02
LAM BERTS
FOOT
®.a7
7-
Comparison of low light performance of Super XX film and an
Fig.
image orthicon. (Image orthicon picture is on the left of each frame.)
be considered valid; (2) that the two types of measurements do not
match; and (3) that a major discrepancy is that the blending distance measurements tend to decrease at large densities while the
objective measurements tend to increase.
In contrast to the above, the present paper would argue that the
two types of measurement, I (by the eye) and II (by a scanning aperture), should, so far as the eye and film satisfy the same physical
www.americanradiohistory.com
10G
TELEVISION, Volume IV
equations derived for an ideal device, show good* agreement. A
large part of the discrepancy noted above under (3) is removable
when reference is made to Figure 4. Here it is seen that in the range
of 0.1 to 10 ft-L the discrimination of the eye for small contrast differ ences varies by about five to one. This would correspond to a five -toone ratio of blending distances for the same film viewed at a
brightness of 10 and at a brightness of 0.1 ft -L. Because the visual
observations of blending distance are made with a fixed source brightness attenuated by films of varying density, the resulting blending
distance measurements are a product of the graininess properties of
the film and the contrast discrimination properties of the eye as a
function of scene brightness. When the latter term is separated out,
the graininess versus density measurements by the two methods
(observer and instrument) show relatively good agreement.
A further rough confirmation may be obtained by reference to some
"blending distance" measurements of Lowryn in which a constant
viewing brightness was preserved. These showed about a factor of
two increase in graininess for a variety of films in the range of densities from 0.2 to 1.0. This increase is in good agreement with the
objective (large aperture) measurements of Jones and Higgins14 shown
in their Figure 16.
It is worth commenting briefly on another item emphasized in the
second paper by Jones and Higgins." The concept of the "effective
scanning area" used by the eye in evaluating graininess is introduced.
This is thought to be a useful concept particularly because the results
of objective measurements, using different scanning aperture sizes,
suggest the possibility of matching visual observations with small
apertures rather than large apertures.
Arguments, similar to the above, were at one time current in evaluating the "noise" in a television picture. It was often remarked that
it was only the high- frequency noise that was objectionable. This
would correspond, for example, to selectively emphasizing the observations of graininess of film obtained with small scanning apertures
(either retinal or instrumental) It is a relatively simple experiment
in a television system to increase the effective scanning aperture
.
* A precise correlation between eye
and instrument
must,
of course, take into account the detailed performance of observations
and film near
their limiting resolution-performance which both for eye
eye and film is
determined more by the finite size of its mosaic elements than
by statistical
fluctuations. The significance attached to precise visual observations,
however, should be tempered by the known large spread of eye characteristics
from individual to individual.
15 Lowry, E. M., "An Instrument for the
Measurement of Graininess of
Photographic Materials ", Jour. Opt. Soc. Amer.,
26, 1 (Jan. 1936), p. 65.
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
107
several fold, either by reduction of pass band or by defocusing the
kinescope spot. Such aperture changes are accompanied by large
changes in total noise power as viewed on the kinescope. Yet the
effect on visibility of noise of cutting out the high- frequency noise
components is small compared with the same changes in noise power
distributed uniformly over the noise spectrum. This latter statement is borne out by the curves in Figure 9. In brief, the visibility, or
annoyance, of noise must be assessed over the full range of picture
element sizes from elements at the limiting resolution of the eye to
the largest element, which is the picture itself considered as a unit.
RESOLUTION
The most frequently used, because it is the most easily observed,
specification of resolution is the finest detail that a system can resolve.
This is true for film, pickup tubes, the eye, and optical lenses. In
general, this specification is satisfactory if it is appreciated that the
limiting resolution itself has only narrow utility and that the limiting
resolution is more an indirect measure of what detail is well resolved
by the system. The "well- resolved" detail may be two to four times
coarser than the finest detail. And in the judgment of picture quality,
the eye attaches little weight to the picture elements in the neighborhood of limiting resolution.
One illustration of the confusion caused by the use of limiting
resolution is the comparison frequently made between the resolution of
motion picture film and of a television system. The limiting resolution
of film is compared with the "cutoff" resolution of a television picture.
The picture detail at the "cutoff" resolution of a television system,
however, as limited by the amplifier pass band, has at least the possibility of being clearly resolved. It is misleading to attach the same
weight to this type of resolution as is attached to the limiting resolution of film. It would be nearer a true evaluation if the resolution of
film were specified at that number of lines at which film matched the
signal -to -noise ratio of a television system at its "cutoff." Such a
comparison would place the resolution of 35 -mm motion picture film,
normally quoted at a limiting resolution of 1000 to 2000 lines, nearer
to the resolution of a 500 -line than a 1000 -line television picture.
In general, the specification of the signal -to -noise ratio that a picture pickup device can transmit at an intermediate resolution is a more
accurate and significant specification, not only of resolution, but also
at the same time of the half -tone discrimination of the device, than
is the specification of limiting resolution.
www.americanradiohistory.com
108
TELEVISION, Volume IV
SATISFYING THE HUMAN EYE
Only one problem, that of presentation brightness, will be discussed
here. Figure 8 shows the signal -to -noise ratio curves of a picture
taken at scene brightness B1, and viewed by the eye at presentation
brightness B1'. The viewed picture is assumed to be "noise free" and
accordingly the B1 curve lies above the B1' curve. If, now, the presentation brightness is increased to B2', the original scene brightness
must also be increased, other things being constant, by the same factor to B2' in order to match the increased discrimination of the eye.
These considerations are significant because both motion pictures and
television pictures aim at higher presentation or screen brightness.
/
A
Al
P
LINEAR
Fig.
8-Dependence of
ELEMENT
B2
BZ
B,
B:
B,
B
BI
BI
SIZE - h
scene brightness on reproduction brightness.
The converse of the above operations makes an interesting test.
Given a grainy motion picture or a "noisy" television picture, the
most effective way of eliminating the fluctuations with the least cost
to picture detail is to interpose a neutral filter between the eye and
the picture. The discrimination of the eye is thereby readily reduced
below the fluctuation limits of the picture. At the same time, the
picture is shifted to a portion of the eye characteristic which shows
higher apparent contrast and thus partially compensates for the loss
of brightness. Figure 9 shows schematically the effect on picture
detail of three ways of trying to eliminate "noise": reduction of picture brightness, increase in viewing distance and reduction of bandwidth of the picture. The last -named operation is peculiar to a television
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
109
system and, while it reduces the total noise in the system, has little
effect on the visibility of noise until an extremely coarse picture is
obtained.
The curves B1 are the signal -to-noise ratio characteristic of the
picture. The curves B1', B1 ", B2' are the signal -to -noise ratio characteristics of the eye at the brightnesses B1' and B2'. In order that the
fluctuations in the picture not be observed by the eye, the signal -tonoise ratio of the picture should be above B1', B1 ", or B2'. The limits
BI
(01
R
LINEAR ELEMENT
SIZE -
h
o
I
1
1
I
i
BIQ Gt\,PE
BII
LINEAR ELEMENT
SIZE - h
BI
1
(C)
R
LINEAR ELEMENT SIZE
-h
Fig. 9 -(a) Noise reduction by lowered reproduction brightness; (b) Noise reduction by increased observer distance; (c) Noise reduction
by bandwidth reduction.
of performance of the eye are shown by the three dotted lines. They
mark out the minimum area that the eye can resolve by virtue of its
cone structure, the minimum signal -to -noise ratio that it can perceive,
and the maximum signal -to -noise ratio it can generate corresponding
to the Weber- Fechner limit of 2 per cent brightness discrimination.
The "cutoff" characteristics of the eye are shown as idealized sharp
breaks to simplify the argument.
www.americanradiohistory.com
110
TELEVISION, Volume IV
Starting with a noisy picture, that is, B1' lying above B1 as in
Figure 9a, there are several formal operations that can be performed
to get rid of the noise, that is, to insure that all parts of B1 lie above
the eye curve. Each of these operations corresponds to a physical operation and each affects the observed picture detail differently. In
Figure 9a the eye curve B1' is transformed into BZ by a change of
ordinate scale factor. This corresponds to interposing a neutral filter
at the eye. The finest detail observable is still at the "cutoff" limit of
the eye. In Figure 9b, the eye curve B' is transformed into B1" by a
change of abscissa scale factor. This corresponds to backing away
from the picture. Although the finest observable detail remains at the
"cutoff" of the eye, this "cutoff" now corresponds to coarser detail in
the picture. In Figure 9c, the pass band of the amplifier through which
the original picture is transmitted is reduced to the point where the
picture fluctuations are below the Weber-Fechner limit. This is an
expensive way to remove noise -expensive in picture detail.
A final aspect of the significance of presentation brightness arises
in the comparison of the low light performance of a man -made
device
with that of the human eye. Assume, for example, that the manmade device is as sensitive as the eye. If one picks up a scene whose
brightness is 0.1 ft -L and views the reproduction at a presentation
brightness of 10 ft -L, noise should be visible in the reproduction while
it was not visible in the original scene. The higher presentation
brightness gives the eye an unfair advantage. A more valid procedure
would match the presentation brightness of the reproduction
with
the brightness of the original scene.
VISUAL NOISE
The phrase "signal-to -noise ratio of the eye" has been used frequently in the preceding discussion. One might expect to be able
to
"see" these fluctuations just as one sees the graininess of film or the
noise in a television picture. The writer is convinced that such fluctuations are observable* particularly at low lights around 10 -4 ft -L.
A white surface then takes on a grainy appearance not unlike
that of
motion picture film. The observations in more detail are: in complete
darkness little or no fluctuations are detectable, a fact which attests
the substantial absence of local noise sources in the eye. Near threshold
brightnesses, large area, low amplitude fluctuations appear. At higher
brightnesses these fluctuations increase in amplitude and decrease in
size. In the neighborhood of 10 -2 ft -L the fluctuations tend
to dis*
See also DeVries.
www.americanradiohistory.com
FILM, PICKUP TUBES, HUMAN EYE
111
appear and a white surface takes on a "smooth" appearance and
remains so at normal brightness levels. A secondary observation is
that low -level blue light appears distinctly more grainy than low -level
red light.
The last observation, together with known data on dark adaptation, fits in well with the assumption of a gain control mechanism in
the eye. This gain control, just as the gain control in a television
receiver or the lamp brightness used for film projection, does not alter
the signal -to -noise ratio but does alter the visibility of noise by presenting the picture at a higher or lower brightness level. Thus, at
high scene brightnesses, the gain control in the eye may be turned
down to the point where the fluctuations are just not visible. (The
sensitivity of the eye is apparently high enough to afford this luxury.)
If one suddenly reduces the scene brightness, the gain control is still
momentarily set at a low value and the pictiire is dim or not visible.
As the gain control resets itself at a higher level, the picture appears
to get brighter. This corresponds with the experience of dark adaptation. At these low light levels (10 -} ft -L) one has only to assume
that the gain control is set high enough to make the fluctuations
visible.
To account for the observations that low-level blue light appears
to have more fluctuations than low -level red light, the gain control
mechanism can be assumed to be set higher for blue than for red.
This is not as "ad hoc" as it may appear. The reason is that, although
at low -light levels* blue appears brighter and grainier than red, they
both present the same resolution to the eye.16 And since the resolution is determined by signal-to -noise ratio, this is in agreement with
the assumption of a gain control that varies presentation brightness
but not signal -to-noise ratio.
ACKNOWLEDGMENTS
The writer would like to acknowledge, without committing the
acknowledged to the conclusions presented above, his indebtedness to
Dr. D. O. North and Dr. G. A. Morton of these laboratories, and
O. H. Schade of the RCA Victor Division for many profitable discussions of the subject of this paper.
The test is performed by starting with red and blue at the same
brightness at high -light levels and attenuating both by the same neutral
filter.
18 Luckiesh, M., and Taylor, A. H., "A Summary of Researches in Seeing at Low Brightness Levels ", Illum. Eng., 38, 4 (Apr. 1943), p. 189.
www.americanradiohistory.com
ANALYSIS, SYNTHESIS, AND EVALUATION OF THE
TRANSIENT RESPONSE OF TELEVISION
APPARATUS *t
BY
A. V. BEDFORDt AND G. L. FREDENDALL$
RCA Manufacturing Co., Inc.,
Camden, N. J.
Summary -The sharpness of detail in a television picture is directly
dependent upon the capability of the transmitter for the transmission of
abrupt changes in picture half tone. A suitable test signal is a square wave
of sufficiently long period.
Rules are deduced for the evaluation of the subjective sharpness to be
expected in transmitted pictures and may be applied when the square -wave
response of the transmitting apparatus is known.
Rapid chart methods have been devised for (1) the analysis of a
square -wave output into sine -wave amplitude and phase response and (2)
the synthesis of a square -wave response from a given set of amplitude and
phase characteristics. Analysis furnishes an immediate solution to the
familiar but troublesome problem of finding the sine -wave characteristics
of television apparatus.
The four aspects of the application of square waves to television, i.e.,
measurement, analysis, synthesis, and evaluation, are presented as a basis
for a unified and complete technique.
The authors hope that this paper will be a contribution to the general
problem of working out electrical specifications for television transmitters
and other television apparatus, giving information regarding the steepness
of rise and the amplitude of overswing of the square-wave response.
I. INTRODUCTION
a result of the scanning process, the sharpness of detail in a
television picture is directly dependent on the capability of the
S transmitter and receiver for the faithful
transmission of signals arising from abrupt changes in picture half tone along the scanning line. Recognition of the validity of the Heaviside unit voltage,
the electrical equivalent of an abrupt change in half tone, as a test
signal, was accorded early in the art. Notwithstanding, the preponderance of emphasis has been placed upon the sine -wave characteristics
of television apparatus, that is, upon the amplitude- and phase- versus-
d
frequency characteristics.
*
Decimal classification: R583.
f Presented at the Summer I.R.E. Convention in Detroit, Michigan,
on
June 25, 1941. Reprinted from Proc. I.R.E., October, 1942.
1 Now with the Research Department, RCA Laboratories Division,
Princeton, N. J.
112
_
www.americanradiohistory.com
TRANSIENT RESPONSE
113
It has long been known that the response of a linear signal- transmitting system to a Heaviside unit voltage contains all the information
necessary to determine both the phase-frequency and the amplitude frequency characteristics. Conversely, the two frequency characteristics determine uniquely the response to a unit voltage (and incidentally
the response to any other transient input wave). It is also known that
the response to a single abrupt rise in a repeating square wave of
sufficiently long period is essentially the same as the response to a
Heaviside unit voltage insofar as that part of the transient response
due to high frequencies is concerned.
In view of this implicit relationship between the sine -wave and the
transient -responses characteristics of electrical circuits, it is surprising
that the testing and specifying of high- frequency fidelity of both audio
and video apparatus by the response to a square -wave has not become
more common.'
Several circumstances have impeded rapid growth in the use of
square waves to determine the high- frequency response. In the first
place, the well -established sine-wave methods, which were developed for
audio work, have the advantages of precedence and well -developed
techniques for measurement, recording and plotting data, diagnosing
imperfection, and comparing performances. Second, the lack of suitable oscillographic apparatus for accurately indicating the instantaneous response as a function of time has been a large contributing factor.
A recently developed square -wave oscillograph' and square -wave generator provide a solution to this problem.
The laboriousness of classical methods for translating the results
of sine -wave measurements into transient response and vice -versa, has
tended improperly to make the two test methods seem unrelated and
competitive instead of complementary. Furthermore, there has been
no satisfactory means for evaluating numerically the fidelity of a pars In this paper, the term transient response will be used as the equivalent of the expression response to a Heaviside unit voltage.
2 The use of a 60-cycle square -wave generator and an oscillograph for
investigation of the behavior of a television system at low frequencies of
the order of the field scanning rate is well established. In these measurements, performance is judged by inspection of the "tilted" output wave
and harmonic analysis of the wave is usually not desired. In general, a
television system will have uniform response over a frequency range of
many octaves in the region between the so- called "low- frequency" end and
the "high- frequency" end, so that fidelity measurements at the two ends of
the spectrum may be considered separately. This paper will be concerned
only with the high- frequency end.
3 R. D. Bell, A. V. Bedford, and H. N. Kozanowski, "A Portable High Frequency Square -Wave Oscillograph for Television ", Proc. I.R.E., October,
1942.
www.americanradiohistory.com
TELEVISION, Volume IV
114
titular piece of apparatus or transmitting system from the transient
response. In television .applications it has been recognized that the
"mean" steepness of rise of the transient -response wave is a measure
of the sharpness of the picture which the system could transmit. Still
there is no general agreement on a method of measurement and calculation of a mean value of steepness which is a consistent and an accurate
indication of the picture sharpness for a wave having nonuniform
steepness during the time of rise. The presence of overswing in the
transient response makes evaluation even more difficult because of the
effect on the visual sharpness of the picture and because of the introduction of spurious effects in the picture.
Accordingly, for the purpose of simplifying the passage between
sine -wave response and transient response and interpreting the latter,
we present below: (1) a graphical chart method for analyzing the
response of a system to a square -wave input signal to obtain the sine wave phase and amplitude characteristics; (2) a graphical chart
method for synthesizing the response to a square wave from the sine wave phase and amplitude characteristics; (3) a method for evaluating
the mean steepness of a transient -response wave in terms of the width
of blur produced in a television image by a wave which is similar in
its visual effect and which has a linear change from one level to
another; and (4) suggestions for the supplementing of sine -wave measurements by transient measurements.
II.
ANALYSIS OF SQUARE -WAVE RESPONSE INTO PHASE -FREQUENCY AND
AMPLITUDE- FREQUENCY CHARACTERISTICS
The analysis employs several permanent charts (Figures 2 to 5),
the construction and use of which may be simply explained by reference to Figure 1.
Figure 1(a) shows the repeating square wave e applied to the
apparatus under test. Wave d is the response measured at the output
terminals. The fundamental period of e is not critical but must be
taken long enough to insure that wave d has subsided to a substantially
constant level during the latter part of each half cycle. Also the true
time relation between waves d and e under test conditions has no
significance in the present analysis and need not be known.
A basic hypothesis upon which the analysis of wave d rests is that
a stepped wave f may be drawn which approximates in harmonic content that of wave d. It is clear from inspection that the waveform of f
may be caused to approach that of e as closely as desired by taking
the steps sufficiently small. Since wave f in the regions a to b and a'
www.americanradiohistory.com
TRANSIENT RESPONSE
115
to b' is of uniform amplitude, the lengths of the time intervals from
a to b and a' to b' have no bearing on the shape of the transient
portion of the waves. Hence, the fundamental period of wave f is
unimportant provided that the value is great enough so that a part of
f is uniform after each transition. The upward transition of wave f
is the same as the downward transition except for inversion. Hence,
the rectangles g ", h ", i ", etc., may be considered as continuations of
b
d,
f
a.'
V):
h'
P
9,
9
(a)
(b)
o
Fig. 1 -(a) Approximation of a square -wave response d by a stepped wave f.
(b) Vector addition of components of frequency in waves g, h, i, etc.
(c) Vector addition of components of frequency 2f in waves g, h, i, etc.
f
rectangles g', h', i', etc. It follows that wave
f has the components
g, h, i, etc.
Each of these square -wave components is identical in shape to the
input wave e and hence contains all harmonics in the same proportion
as the input wave. Each of the square waves, however, has a different
delay and hence, the harmonics of the various square waves occur in
different phase relations. Therefore, the magnitude of each harmonic
www.americanradiohistory.com
116
TELEVISION, Volume IV
in the stepped wave, relative to the same harmonic in the input wave e,
is indicated by the sum of a group of vectors having amplitudes proportional to the steps g', h', i', etc., and angular positions corresponding to the different delays of the square waves g, h, i, etc. No cognizance need be taken of the fact that the harmonics of a square wave
vary in amplitude inversely as their frequencies. Inasmuch as the
fundamental frequency of the input wave may be allowed to approach
zero (such that wave e becomes a Heaviside unit voltage), any reference to discrete harmonics may be dropped and the response of the
apparatus approximated at any frequency to a degree of accuracy
which depends on the fineness of the steps in wave f.
Figure 1(b) shows the vector addition involved in finding the
response for the frequency
Each vector component has the same
angular position, namely, 27rf T, with respect to the preceding component since the waves g, h, i, etc. correspond to points on the real
response curve taken at equal time intervals T. Waves l and m are
negative; hence, the vectors l and m are negative. The length of the
vectorial sum OB gives the relative amplitude response of the apparatus tested and the angle AOB is the relative phase shift for the
sine wave fn. Figure 1 (c) is drawn for a frequency equal to 2f,,. The
angle between successive vectors is 272f T. As in Figure 1 (b), OB
is the amplitude response at the frequency 2f,, and the angle AOB is
f.
the relative phase shift.
A set of analysis charts, Figures 2, 3, 4, and 5, were designed in
order to reduce to a minimum the labor involved in performing the
vectorial additions in Figures 1 (b) and (c). Essentially, each chart
serves as a protractor and a linear radial scale for use in locating the
end points of the vectors. Two sets of time intervals, 1/20 and 1/30
microsecond, (referred to as 20- and 30- megacycle dots) between successive components g, h, i, etc., appear to be adequate for television
applications. The basis for a choice of one of the two sets depends
upon the degree of accuracy desired. This aspect is discussed later.
Components (i.e. readings from the transient-response wave d) are
numbered to correspond to radial lines on the charts. The angle
between consecutively numbered radial lines is the angle used in the
construction of Figures 1 (b) and (c) (e.g., on Figure 3, the angle in
scale 1 for the solution of the response at 1 megacycle is 18 degrees.)
The component vectors g, h, i, etc., lie along the radial lines and the
vectors are added as in Figure 1 (b) by manipulation of a sheet of
semitransparent paper. Detailed directions for the operation of the
charts are given below. The charts and directions are drawn or printed
preferably on cardboard or stiff paper.
www.americanradiohistory.com
TRANSIENT RESPONSE
A.
Instructions for Using Figures
2, 3, 4 and 5
117
for Analysis of Square -
Wave Response
Using a square -wave generator and an oscillograph or other means,
obtain per cent voltage readings at 1/20- microsecond intervals (corresponding to oscilloscopic readings with 20- megacycle "dots" for
timing) along the transient wave ( Figure 6) such as shown in column
s
10
-11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
2627
28
29
30
Fig.
Scale
2
-Chart
for square -wave analysis.
20 -Mc Dots
gl
0.25 Mc
0.5 Mc
02
03
1.5
Mc
30 -Mc Dots
0.375 Mc
0.75 Mc
2.25 Mc
(a) of Table I. (If more accurate analysis is desired, use readings at
1 /30- microsecond intervals)
Readings should begin at the zero -voltage level before the beginning
of the transient and end at a 100 per cent point where the voltage
becomes uniform after the transient rise has been completed and substantial rest attained. Wave d of Figure 1 is a typical plot of such a
wave but plotting is not necessary for the purpose of the analysis.
.
www.americanradiohistory.com
TELEVISION, Volume IV
118
Compile column (b) by taking the differences in adjacent readings in
column (a). These numbers represent the amplitudes of the increments in the "stairstep" wave f. Number the increments consecutively
as in column (c) .
Draw a horizontal x axis and a vertical y axis on a sheet of sufficiently transparent tissue paper. Place the tissue paper on Figure 4,
Fig.
Scale
3
-Chart for
square -wave analysis.
20 -Mc Dots
Mc
2 Mc
3 Mc
1
30 -Mc Dots
1.5
3
4.5
Mc
Mc
Mc
for example, with the origin of the x -y system at the center of the
chart. (It will be noted that Figures 2, 3, and 4 are for different frequencies.) Make a pencil dot on the tissue on the radial line marked
"1" in circular scale (1) at a distance from the center of the chart
equal to increment No. 1 in column (b). Move the tissue so that this
dot coincides with the center of the chart, keeping the x axis parallel
www.americanradiohistory.com
TRANSIENT RESPONSE
119
to the horizontal dotted lines on the chart. Make a second dot on the
radial line marked "2" in circular scale (1) at a radius equal to increment No. 2. Move the tissue until the second dot is at the center of
the chart and repeat the procedure for the remaining increments.
Positive increments are plotted in the radial direction toward the increment number in circular scale (1) while negative increments are
Fig.
Scale
4
-Chart for
square -wave analysis.
20 -Mc Dots
Mc
3.5 Mc
4.0 Mc
2.5
30 -Mc Dots
3.75
5.25
6.0
Mc
Mc
Mc
plotted in the opposite direction. Dots may be numbered to avoid
errors.
Draw a vector from the x -y origin to the final dot located. Place
the tissue on Figure 5 and read the length of the vector on the "amplitude scale." This length is the 2.5- megacycle amplitude response in
per cent. Read the angle between the y axis and the final vector by
www.americanradiohistory.com
TELEVISION, Volume IV
120
using the protractor in Figure 5. This angle4 is the phase lag of the
2.5-megacycle component of wave (d) relative to point P in Figure 1
where P is 1/2 a reading interval before the second reading in column
(a). The phase angles obtained for the various frequency components
are correct relative to one another but the absolute time delay through
the system is not obtained. The phase angles may be converted to time
.25
.20
I.
11
o
gm
W
N
o
2.05
C
.9
:
Cii=ii:i
:.C
®rawrr
:
II1iI1IIIIiII/JI
ai
I1r1IINI.N
/%l/IR
amro1lri'itlpimmm mm1
iiiiil
iiiiii.O© iiiRcMiMrEiU;
mri=inrri.ii...CoIsG
NIIJII./.N
I(/GEN111
/'L/.IIJA
II/I
Ii'/.I
I/
III
N%NIG/.IN/li
N%IIIII1LiJ
NILG/`
IIIiIrI
IJIILN//'UiEI/N/mG/
IA/1IIGi.ti
i.
inriIiIi YS/iG/I/E
íi
úliiiÿí ."iiiiiiiiiiiiiiiiii=
+
I
200
190
180
170
160
150
140
130
umrlr.
VIYSí
rainure'/.'SI%G"
IIIIIIIX0INI
%!`
00
30
60
90
120
150
180
o
210
270
300
240
DEGREES SHIFT
330
120
110
.i00
90
80
360
-
70
60
50
40
30
20
90
270
10
0
2
14
180
Fig.
5
-Chart for
square -wave analysis.
delay by means of the delay graph of Figure 5. In the example given
above, the amplitude response at 2.5 megacycles is 85 per cent. The
relative phase angle is 103 degrees which corresponds to a time delay
4 The angle read between the y axis and the vector is the phase angle
for the 2.5- megacycle response of the stepped wave f with respect to the
time of the second reading of column (a). Wave f lags the true -response
wave d by one half a reading interval.
.
www.americanradiohistory.com
TRANSIENT RESPONSE
121
of 0.114 microsecond. Obtain the response at other frequencies in a
similar manner by using the increments in conjunction with the other
scales of Figure 2 and the scales of other charts.
The charts may be used for other frequency ranges such as the
audio range. In such applications, the time interval between successive readings on the transient -response wave will not usually correspond to 20- or 30- megacycle dots. However, the analysis charts shown
in Figures 2, 3, 4, and 5 may be employed as in the example given in
Table I, when the appropriate multiplying factor for frequency is
computed. This factor is equal to To /TN where To is 1/30 X 10-6 or
1/20 X 10 -6 second depending upon the original dot frequency for
which the charts were designed. TN is the new time interval in seconds.
120
Ho
loo
90
ea
r"-LUMINOUS DOTS
70
60
50
40
30
20
10
0
6-Illustrative
example of a square -wave response reproduced by an
oscillograph showing readings at 0.05- microsecond intervals
(20- megacycle dots).
Fig.
B. Examples of Accuracy of Chart Analysis in Specific Applications
In the instance of a compensated resistance -coupled amplifier,
mathematical formulas are available for the exact calculations of the
amplitude and delay characteristics and the response to a unit function.' Hence, some conception of the accuracy of chart analysis may
be gained by observing the agreement of data thus determined with
the exact characteristics. Figure 7 contains such data based on the
analysis of the unit -function -response wave shown in Figure 8. Figure
9 contains similar data based on the response wave of Figure 10. Since
the delay characteristics determined by graphical analysis have only
5
Appendix II.
www.americanradiohistory.com
TELEVISION, Volume IV
122
TABLE I
(a)
0
2
22
56
87
108
108
(b)
-
(c)
20
34
2
-
2
1
31
3
4
21
5
0
6
(a)
(b)
95
93
101
102
99
100
100
-13
-
(c)
7
2
8
8
1
9
10
3
11
1
12
13
0
relative significance, the data points corresponding to a 20- megacycledot readings were shifted a constant amount (0.026 microsecond) in
Figures 7 and 9 so that the best correspondence with the mathematically determined delay curves was reached in order that the results of
graphical and mathematical methods may be compared directly. A
similar shift of 0.017 microsecond was made in the case of the 30megacycle -dot data. The deviations of the data points in Figures 7
and 9 from the exact curves represent the degree of approximation of
analysis as applied to two specific cases. For the general case, no
definite limits can be set up for the error in sine -wave characteristics
as determined by chart analysis. As in Fourier analysis,' the error
1
,u
AMPLITUDE
0
W
io
0-MC DOT5
osEL..
111111M10-MC
0
0
1111"
)
5
4
FREQUENCY -MC
2
DO
3
6
.10
T5
09
08
r-
ó
07
06
7-
Fig.
Curves show exact amplitude and delay characteristics of a 2 -stage
compensated resistance- capacitance amplifier in which K = V 2, f, = 3 megacycles. Amplitude and delay characteristics determined by chart analysis of
Figure 8 are shown by data points.
6 The amplitude characteristic of a circuit, as determined
graphically
from the square -wave response, may be converted into the amplitude characteristic corresponding to Fourier analysis of the response of the same circuit
to a square pulse. This conversion is discussed in Appendix III.
www.americanradiohistory.com
TRANSIENT RESPONSE
123
depends upon the specific curve which is analyzed and upon the length
of the time interval between readings, or in other terms, upon how
well the curve is defined by the series of dots. A better approximation
to the exact amplitude or delay characteristic at the higher frequencies is afforded by a solution based on 30-megacycle -dot readings rather
than on 20- megacycle -dot readings.
It is conceivable that as a consequence of strong components of
very high frequency the transient -response wave could make violent
excursions between adjacent readings such that a plot of the readings
would not clearly define the wave even for the lower- frequency components. In such a case, inspection would show that analysis would be
12
_
u
10
1
)
2
.
.2
4
.3
TIME
-
A.1
5
u
7
SEC
Fig. 8-Exact response to a unit function of a 2 -stage compensated resistance- capacitance amplifier where K = V 2 and fo =3 megacycles. Points
shown were obtained by synthesis from theoretical amplitude and delay
characteristics in Figure 7.
inaccurate. In general, it has been found that if the 20- megacycle
timing dots trace out the square -wave response unmistakably, the
20- megacycle dots are sufficiently accurate for analysis out to 3.5
megacycles. Under the same conditions, 30- megacycle dots are adequate out to 5.25 megacycles.
III.
SYNTHESIS OF SQUARE -WAVE RESPONSE FROM THE SINE -WAVE
CHARACTERISTICS
A closed
mathematical formula' is available for the calculation of
A. V. Bedford and G. L. Fredendall, "Transient Response of Multistage Video -Frequency Amplifiers ", Proc. I.R.E., Vol. 27, pp. 277-285;
April, 1939.
7
www.americanradiohistory.com
i
TELEVISION, Volume IV
124
160
140
1zoz
100
LJ
80
,
MUDELA
uviu
ram III
AMPLITUDE
ci
X
2
2030-
MCMC
,
07
06
W
05
04 Q
J
3
02
0
DD
/0
6
1
tl)
FREQUENCY -MC
9-
Fig.
Theoretical amplitude and delay characteristics of a 2 -stage compensated resistance -capacitance amplifier in which K = 1.1 and fo = 4 megacycles. Amplitude and delay characteristics determined by chart analysis
of Figure 10 are shown by data points.
140
120
W
100
N
ó
a.
e0
CC
40
N
40
20
.3
TIME -
J..1
.4
MC
Fig. 10 -Exact response to a unit function. of a 2 -stage compensated resistance- capacitance amplifier in which K = 1.1 and = 4 megacycles.
f
www.americanradiohistory.com
TRANSIENT RESPONSE
125
the response of a linear electrical system to a unit function from the
sine -wave amplitude and phase characteristics. When a numerical
answer is sought, however, the formula is rarely practicable. As
pointed out earlier, the solution for the response to a square wave may
usually be substituted when the period required does not necessitate
the computation of many terms of the Fourier series. The period 1 /f,,
chosen must be long enough to permit the transient response of the
circuit under consideration to assume a substantially constant value
between consecutive abrupt changes in the square wave.
If the square wave applied to the circuit is
E(t) =
-+- (sin
1
2
2
7r
27rfnt
+
1/3
sin 671,,,t +
sin 107rfnt
1/5
++1
sin 27rnfDt+
(1)
n
then the response is
e(t)
=-+1
2
2
(A1 sin
27rf(t-D1) +
A3
sin 67rfD(t-D3)
3
7r
++
An
n
sin 27rnfv(t-Dn)
+
(2)
where An and Dn are the amplitude response and delay of the circuit
for the nth harmonic. Dn = phase shift in radians /27rnff.
Synthesis charts, Figures 11, 12, and 13, were developed for the
summing of the significant terms of the series (2) above. The principle of the charts is based on the vector representation of a sine function. For example, the term An /n sin 27rnf, (t Dn) is represented in
Figure 14(a) by the length of the perpendicular AC to the vector OD.
The value of the term may be found rapidly, for specific values of t
which have been determined in advance, by dividing the circumference
of a circle into a number of equal parts.
For example, if the circle in Figure 14 (b) is divided into N equal
segments, then the value of the sine term may be found every 1/Nnf
second starting at t = O. It is convenient to designate a specific value
of time t by one of the whole numbers between 0 and N.
A straightedge may be placed along the radial line which makes
the required angle 27rnf,Dn with the x axis. (See Figure 14(b) ). The
quantity 27rnf1,Dn will usually be expressed in terms of the unit angular
segment used in marking off the circumference of the circle. If a
-
www.americanradiohistory.com
126
TELEVISION, Volume IV
draftsman's triangle is moved along OD, successive values of the sine
term corresponding to regular intervals of time may be read on a
calibrated scale AC (marked on one leg of the triangle) by noting the
points of intersection of the radial lines and an imaginary circle having a diameter OA equal to An /n. The multiplying factor 2 /7r in (2)
may be taken into account by suitable calibration of the scale AC. A
Fig. 11 -Chart for square -wave synthesis.
0.05- microsecond intervals
Scale
0.25 Mc
0.75 Mc
1.25 Mc
1.75 Mc
number of concentric circles permanently drawn on the chart, so as to
divide the diameter of the largest circle into a convenient number of
equal parts, will aid in finding the length OA as the triangle is shifted.
Readings along the calibrated edge are positive above the line OD and
negative below OD.
www.americanradiohistory.com
TRANSIENT RESPONSE
127
The same chart may be used for a number of different harmonic
terms by simply numbering the radial lines appropriately for each
term. Thus, in Figure 14(b), if the circular scale (1) represents the
fundamental term sin 27rfrt, then circular scale (3) will represent the
3rd -harmonic term sin 6irfrt in which the unit angle is 3 times the
unit angle in scale (1) etc. The steps involved in finding and sum-
Fig.
12
-Chart for
square -wave synthesis.
0.05 -microsecond intervals
0.25 Mc
2.25 Mc
2.75 Mc
3.75 Mc
ming the sine terms in (2) have been systematized in a set of synthesis charts, Figures 11, 12, and 13. The following directions have
been drawn up to facilitate rapid use of the synthesis charts.
A. Directions for Use of Synthesis Charts
Figures 11 , 12, and 13 are used in the compilation of the response
www.americanradiohistory.com
TELEVISION, Volume IV
128
of an electric circuit to a square wave as represented by the first 10
odd harmonics (1st, 3rd,
19th) when the sine -wave, phase- delay,
and amplitude characteristics of the circuit are known.
The particular charts illustrated were arbitrarily designed for a
fundamental frequency of 0.25 megacycle which has been found satisfactory for television purposes. In this instance the square -wave
...,
Fig.
Scale
13
-Chart for
square -wave synthesis.
0.05- microsecond intervals
0.25
Mc
15
3.75
Mc
17
4.25
19
4.75
Mc
Mc
response is found at intervals of 0.05- microsecond (i.e., 1/80 of the
period). The result is a close approximation to the response of a circuit to a unit function when (1) the square -wave response attains
substantially a steady value within 2 microseconds (i.e., one half the
www.americanradiohistory.com
TRANSIENT RESPONSE
(d)
/`F
LINEAR S CA L E
ON TR/ANGLE
(b)
Fig. 14
(a) Basic principle of construction of synthesis charts.
(b) Use of synthesis charts.
www.americanradiohistory.com
129
TELEVISION, Volume IV
130
period of the fundamental frequency) and (2) the contribution of
harmonics beyond the 19th (4.75 megacycles) may be neglected.
The rapid use of the charts is expedited by observation of the following procedure in the preparing of Table II. A synthesis of the
amplifier square -wave response shown in Figure 8 from the theoretical
amplitude and phase characteristics in Figure 7 is used as an example.
(1) Record in column (c) the amplitude response in per cent corresponding to each harmonic divided by the order of the harmonic.
(2) Shift the real delay curve up or down but parallel to itself to
any new position for which the minimum delay in the essential
frequency band is equal to, or near, zero. Shifting (i.e., subtracting a constant delay) is a convenience in manipulating
TABLE II
SYNTHESIS OF 32 -STAGE AMPLIFIER
(a)
(b)
Contributions of Harmonics Read
(d)
Amplitude
ude
from Charts
Order
FreResponse Delay
of Har- quency of Har- of Har (0)
(1)
(27)
(28)
(2)
monic, of Har- monic =n monic
t=0
n
monk
(in per
X20n micro- 1=0.05 1=0.1 1=1.35 t=1.4
(Mc)
cent)
seconds
1
3
5
7
9
11
13
I5
17
19
0.25
0.75
1.25
1.75
2.25
2.75
3.25
3.75
4.25
4.75
100
34
21
15
11
7.5
5
3.5
2.5
2
Add +50
Instantaneous Iesponse
1.5
4.7
8.5
12.9
17.6
22.2
26.0
29.0
31.6
33.4
-7.5
-7.8
-8.4
-8.1
-6.6
-4.7
-3.0
-1.7
-0.9
-0.5
-2.5
-2.9
-3.6
-4.3
-4.3
-3.7
-2.8
-2.0
-1.4
-0.9
+2.5
+2.3
+1.6
+0.8
+0.2
+50
+50
+50
+
+22
+59
1
-0.1
o
+0.2
+0.7
+0.4
the charts but is not a necessity.
(3) Record in column (d) the real delay (in microseconds) (or
shifted delay, if performed) of each harmonic multiplied by
20 times the order of the harmonic.
(4) Proceed as follows to enter the contributions of the various
harmonics in columns (0) to (28) :
(A) Pass a straightedge through the center of Figure 11 along
the radial line corresponding to a number on the circular scale (1) equal to the delay of the fundamental
frequency recorded in column (d). Circular scale (1)
refers to the series of radial lines numbered 0 to 80 in
ring (1). The straightedge remains fixed until steps (B)
and (C) below have been completed.
www.americanradiohistory.com
TRANSIENT RESPONSE
131
(B) On the radial line number 0 on scale (1) find a point
which is at a distance from the center of the chart equal
to the number in column (c) . The length of the perpendicular from this point to the straightedge is the contribution of the fundamental frequency at t = O. If the point
referred to is below the axis, the contribution is entered
in column (0) of the table with a negative sign. If above,
the contribution is entered with a positive sign. The perpendicular is measured with a calibrated triangle illustrated in Figure 15. One leg slides along the straightedge
while the calibrated leg forms the perpendicular.
Fig.
15- Calibrated
triangles for reading contributions of harmonics from
synthesis charts.
(C) The contribution of the fundamental at t = 0.05 microsecond, to be entered in column (1) opposite 0.25 megacycle, is found by locating the point 1 on scale (1) and
proceeding as directed for point 0. Point 2 corresponds
to 2 X 0.05 microsecond. In a similar manner, the first
row of entries in Table II are made.
(D) In order to find contributions of the 3rd harmonic, pass
the straightedge through the number on scale (1) corresponding to the delay factor from column (d). Use scale
8 The 100 per cent point on the radial
scales of Figures 11, 12, and 13
should coincide with the 63.6 (= (7r2) 100) per cent point on the calibrated
triangle when the triangle is drawn to the same scale as the figures.
www.americanradiohistory.com
TELEVISION, Volume IV
132
numbers on scale (3) in order to locate the perpendicular
and proceed with the measurement of its length as
directed above for the fundamental frequency.
(E) Complete the table for the other harmonics in a manner
similar to the above.
(F) The instantaneous value of the square -wave response at
t= 0 microseconds is found
by summing column (0) and
adding 50 per cent. Similarly, the response is found at
other times by summing the appropriate column and adding 50 per cent.
B. Examples of Synthesis of Square -Wave Response
Several points on the square -wave response of a 2 -stage amplifier
190
IZA
¢
60
90
10
0
2
.3
TIME
-»SEC
.6
.7
Fig. 16- Synthesis of the square -wave response of a 32 -stage compensated
resistance- capacitance amplifier in which K = 1.51 and fo=_- 7.7 megacycles.
are synthesized in Figure 8 from the amplitude and phase characteristics in Figure 7. Since the theoretical response is shown, a direct
indication of the accuracy of the synthesis is available. The discrepancy is seen to be only a few per cent in this example.
The chart method greatly simplifies and shortens the synthesis of
square -wave response in complex cases in which rigorous mathematical
formulas for response are impracticable. A typical example is the
response of a 32 -stage amplifier which was synthesized in Figure 16
from the theoretical amplitude and phase characteristics of Figure 17.
In this particular case, an absolute measure of the error of the synthesis is not known. The deviation of any synthesized square -wave
response from the exact square -wave response is due only to errors in
www.americanradiohistory.com
TRANSIENT RESPONSE
133
manipulation of the charts and the neglect of contributions of harmonics beyond the limits of the charts (4.25 megacycles).
IV. EVALUATION
OF THE SQUARE -WAVE RESPONSE
Assume that a subject containing an extensive dark area and an
extensive white area with a sharp vertical junction between the two is
used as a test subject for a determination of the fidelity of transmission
of a television system.' An ideal scanning device in crossing the junction would generate a unit voltage which becomes the input signal for
a television transmission system under test. The output signal or
response of the system may be symbolized in Figure 18. The wave3
100
Z
w
U
AMPLITUDE
ao
ñ 60
W
° 40
DELAY
J
2
20
0
Fig.
.
17
4
FREQUENCY -MC
2
3
5
6
0
-Exact
amplitude and delay characteristics of a 32 -stage compensated resistance- capacitance amplifier in which K = 1.51 and
fo = 7.7 megacycles.
shape may be determined experimentally by a square -wave oscillograph
or by synthesis from known sine -wave characteristics of the system.
If the test subject were reproduced by an ideal scanning device (i.e.,
one having negligible aperture losses), actuated by the output signal,
9 The use of a single abrupt transition from one brightness to another
brightness as a test subject for measuring "resolution" in a television picture was developed by R. D. Kell, A. V. Bedford, and G. L. Fredendall in
"Determination of Optimum Number of Lines in Television System ", RCA
REVIEW, Vol. V, pp. 7 -30; July, 1940. A test pattern consisting of several converging bars is much more commonly used in evaluating an entire
television system on account of convenience. The results obtained, however,
are less significant because the resolution of the individual bars is not
affected by the phase fidelity, whereas it is known that phase fidelity affects
the sharpness and utility of most television pictures.
www.americanradiohistory.com
TELEVISION, Volume IV
134
the variation of light intensity on the screen along the scanning lines
would be as shown in the figure when using for the ordinates and
abscissas the light intensity and distance, respectively. The distance
required for the complete change from black to white is greater than
zero due to the finite rate of rise of the response curve. Upon close
observation of the screen, the junction would appear blurred. Furthermore, several alternate light and gray striations following the junction
would be observed as a consequence of the damped oscillation in the
response. The overswing shown in Figure 18 would not be entirely
objectionable for television purposes because at the optimum viewing
distance most of the overswing is not distinguishable from the transi-
WH/TE
REAL
TRANS /T /ON
LINEAR
TRANSITION
ur-
8
BL AC/C
(
DISTANCE ÄLONG SCREEN)
Fig. 18-Nonlinear response of a hypothetical complete television system
and the equivalent linear response in terms of light along the screen surface
or electrical response of the system.
tion and the net effect would be substantially a single transition. In
fact, the visual sharpness of the transition is enhanced by the overswing in the response wave. If a quantitative measure of the sharpness were found, the response wave in Figure 18, for example, could
be compared directly with other waves having different shapes, and
the relative merit of television apparatus with reference to picture
sharpness could be determined.
With the object of finding such a measure of sharpness, the writers
constructed several different synthetic black-white transitions with ink
on cardboard using fine shading lines of variable widths to reproduce
accurate half -tone values. When the transitions were viewed at a distance for which the "blur" was just discernible, that is, the optimum
www.americanradiohistory.com
TRANSIENT RESPONSE
135
viewing distance it was observed that a simple linear transition similar to that shown dotted in Figure 18 could be found for each nonlinear transition such that the observer was unable to distinguish
between the nonlinear transition and its linear equivalent.
The width of the visually equivalent linear transition was termed
the equivalent "blur" of the nonlinear transition. The blur may be
specified in units of distance along the picture screen or in the corresponding units of time. Complex transition curves containing damped
oscillations of sufficiently long duration are properly represented by a
linear transition followed by that part of the oscillation which the eye
does not include with the transition.
The comparison method of determining blur is objectionable
because the labor involved is great and the evaluation of each transition depends somewhat upon the observer's judgment. A method not
subject to these objections has been devised whereby the blur may be
found directly from a plot of the light intensity along a transition.
The steps involved in applying the method are given below in a "Generalized Statement" which defines the conditions under which a linear
transition is visually equivalent to a nonlinear transition such as
Figure 18. For clarity, this "Generalized Statement" is presented in
the form of a law or theorem, but we do not presume to use these
terms until the statement has been proved more adequately by theory
or experiment than is done in this paper.
Generalized Statement
A linear transition having a uniform rate of change of intensity
along the surface from a first mean brightness to a second mean brightness is visually equivalent at the optimum viewing distance to any
nonlinear transition from the first mean brightness to the second mean
brightness when conditions 1 and 2 below are fulfilled.
Condition 1 -The summation of the weighted differences of the
light intensities of the linear transition and of the nonlinear transition
is zero, where the weighted differences are the real differences of light
intensity along the transitions multiplied by a weighting factor. The
The weighting factor varies linearly with distance from a value of
at the first inflection point of the linear transition to + 1 at the second
inflection point; also linearly with distance from a value of zero at a
point preceding the first inflection point by half the, distance between
value at the first inflection point; also
inflection points, to the
linearly with distance from the + 1 value at the second inflection point
to zero at a point following the second inflection point by half the
distance between the inflection points; and is zero for all other points.
-1
-1
www.americanradiohistory.com
TELEVISION, Volume IV
136
Condition 2 -The summation of the differences of the light intensities along the transitions is zero, over the range where the weighting
factor is not zero.
According to the generalized statement, the linear transition E in
19 is visually equivalent to the transition A. In this and other
cases, it is convenient to consider the vertical dimensions J of the
difference areas C, D, F, etc., as positive when A is above E and negative when A is below E. Then, if these dimensions are multiplied by
the corresponding ordinates of the weighting curve N and plotted as
at V, the positive and negative areas C', D', E', etc., are formed. Condition 1 is fulfilled if the algebraic sum of these areas C', D', F', etc.,
Figure
100
SO
I
W
I
WEIGHTING
CURVE
O
N
Fig. 19- Linear transition E is equivalent to A. By condition 2, the algebraic sum of areas C, D, F, G, H, and I is zero. By condition 1, the sum of
weighted areas C, D, F, etc., shown at C', D', F', etc., is zero.
equals zero. Condition 2 simply requires that the algebraic sum of the
positive and negative areas C, D, F, etc., equals zero.
The location of a line E that fulfills the two conditions for a particular nonuniform transition is obtained by a trial- and -error method.
The equivalent blur is indicated as the distance B. Each trial curve E
requires a different weighting curve N since the weighting curve is
itself defined by the equivalent linear- transition curve. Such a procedure would appear to be laborious since the line E has two variable
characteristics, slope and position. However, the number of trials
required is much reduced by the knowledge that for small changes of
slope of line E only condition 1 is affected and for small changes of
position only condition 2 is affected.
www.americanradiohistory.com
TRANSIENT RESPONSE
137
Although the generalized statement for defining a linear transition
which is visually equivalent to a nonlinear transition is essentially
empirical, the conditions appear to be consistent with observed properties of the eye. It is well known that the eye responds to a surface
consisting of many tiny, uniformly distributed white dots on a black
background (or black dots on a white background) as though the surface were uniformly white but the illumination reduced so that the
total reflected light is the same. In other words, the eye responds to
the average10 brightness or to the total brightness. Condition 2 is consistent with this observation in that it requires equality of the total
amount of light of the equivalent transition and the nonlinear transition.
WHITE
'sz
INFLECTION
j-
BLACK
B
INFLECTION
-1
(d)
ADDED
LIGHT
(b)
/
ADDED
LIGHT
(C)
Fig.
20- Actual
LGHT
L
(d l
transitions are shown in solid lines and equivalent linear
transitions are shown in dotted lines.
The plausibility of condition 1 may be established by using the
simple transitions of Figure 20 to illustrate some basic factors which
affect apparent steepness. Figure 20 (a) represents a linear change
from black to white or from any half tone to another half tone. The
blur by our definition is the distance B. Now assume that an extra
amount of light is added near the second inflection point as shown in
Figure 20(b). As a result, the transition has been made effectively
steeper so that the equivalent linear transition must also be made
steeper as shown by the dotted line.
If the light were added near the first inflection point as shown in
10
The eye also responds to the average brightness of a rapidly flickering
source.
www.americanradiohistory.com
138
TELEVISION, Volume IV
Figure 20 (c) , it is apparent that the transition would be less steep as
indicated by the dotted linear transition. From these two observations,
it is plausible that the addition of light at the middle of the transition
as in Figure 20 (d) would not alter the effective steepness. The equivalent linear transition, however, would be displaced to the left as
required by condition 2.
If light were added at a point considerably ahead of the first inflection point of the linear transition or considerably after the second
inflection point, it is reasonable that the effect on the steepness would
be less than if the light were added near the inflection points.
Let us refer back to the example of Figure 19 in which curve E is
assumed to be the visual equivalent of the curve A. Obviously, E is
identical to curve A except for the difference areas C, D, G, etc. Some
of the areas such as C and H render curve A effectively steeper than
the equivalent curve but other areas such as D and G detract from the
steepness of curve A.
If curve E is to be the equivalent of curve A, the algebraic sum of
the effects of all the aiding difference areas C, H, etc., must be canceled
by the opposing areas. Condition 1 states this requirement and provides a weighting factor for determining the effectiveness of each
difference area resulting from its location, as discussed in connection
with Figure 20.
A linear weighting curve has been taken arbitrarily due to the
absence of evidence which would point to a specific form for the curve.
A linear variation is simple and may also be considered as a mean
between the various possible concave and convex forms.
It should be noted especially that in the application of the generalized statement to transitions which have appreciable irregularities in
the region in which the weighting curve is zero (beyond Y), the
irregularities are not included in the equivalent blur. As shown in
Figure 19, the equivalent curve may be continued through Z to the
actual curve A. It should be noted also that the equivalent transition
defined by the statement is intended to be equivalent only when viewed
at a distance for which the nonlinear transition is indistinguishable
from a linear equivalent transition. At such a distance, a very definite
blur may be visible without the shape of the transition being discernible.
The observations made in the several paragraphs above lend support to the concept of an equivalent linear blur but obviously do not
place the concept on a firm physical basis. Furthermore, the physical
accuracy of the particular conditions set forth in the generalized state-
www.americanradiohistory.com
TRANSIENT RESPONSE
139
ment have not been established because of inadequate knowledge of the
eye and brain. However, actual viewing tests with specific transitions
lead us to believe that the application of the generalized statement is
sufficiently accurate to serve a useful purpose for the evaluation of the
transitions found in television.
The authors feel that such a measure of blur also can have utility
in specifying the quality of many other devices related to vision, such
as lens system, photographic film and processes, facsimile transmission,
printing, duplicating processes, and paper.
V. APPLICATION OF SQUARE -WAVE METHODS IN TELEVISION
Square -wave and sine -wave measurements should be mutual aids in
the solution of many television problems. In some instances, a square wave measurement may furnish the data for analysis into amplitude
and delay characteristics. In others, the evaluation of blur is indicated.
In some applications, sine -wave measurement and synthesis of the
square -wave response may be indicated, followed perhaps by evaluation. Most applications will not require the use of all three processes
of square -wave treatment.
We know no simple satisfactory method of judging the degree of
fidelity of a television system from inspection of the sine -wave characteristics. If the square -wave response of the system cannot be secured
experimentally, resort must be had to synthesis from amplitude and
phase data. However, acceptable tolerances in terms of sine -wave performance may be more easily determined when the amplitude and phase
characteristics corresponding to a large variety of transient-response
curves have been determined and cataloged.
In design work in which the characteristics of television apparatus
must be determined by calculation based on circuit constants, synthesis
from the sine -wave characteristics probably constitutes the only
feasible means for obtaining the square -wave response. When the
apparatus is susceptible to experimental test, the most expeditious
method is direct measurement of the square -wave response by oscillographic equipment such as described in a companion paper.' An analysis for amplitude and phase characteristics may then be performed
in order to facilitate the design of equalizing networks if required or
for any other purpose. In this connection it is significant that the
experimental difficulty of phase measurements of extensive apparatus
such as a complete television system including transmitter and receiver
by sine -wave methods is usually so great that the attempt is not often
www.americanradiohistory.com
140
TELEVISION, Volume IV
made.lr, 12 Notwithstanding, a reasonably linear phase shift is conceded to rank with amplitude uniformity in importance. The extreme
ease with which square -wave response can be recorded with suitable
equipment has been pointed out in the companion paper. The analysis
of the response for delay versus frequency through the use of charts
is simple and immediate.
In particular cases, square -wave measurements may provide useful
data which are almost impossible to obtain by sine -wave measurements
and synthesis. An example is the modulation amplifier of a television
transmitter in which the output impedance may be relatively high at
low frequencies, in order to permit a high output voltage, and relatively low at high frequencies such that the frequency fidelity of the
stage alone is poor. Adequate equalization may be inserted in earlier
stages of the system but high- frequency components may be saturated
at high levels. A sine -wave characteristic of the entire amplifier taken
at a low level for which saturation is negligible would indicate high
fidelity. It would be almost impossible to synthesize from the low -level
sine -wave data the transient response for a high level corresponding to
conditions occurring in common use. Square -wave oscillographic tests,
however, would indicate the transient response corresponding to any
desired level. An evaluation of the square -wave response for the purpose of finding the blur corresponding to various levels would be
significant but a determination of the sine -wave response at levels
where saturation exists would have no meaning in the usual sense.
In general, square -wave methods have greatest usefulness (as compared with sine -wave methods) in dealing with performance of units
which are likely to contribute a substantial amount of the distortion
in the transmission characteristic of the system. Included in this
classification are entire transmitters, entire receivers, long transmission lines, pickup chains, and any single amplifier stages which may
be regarded as "bottleneck" stages. It would be rather pointless to
find the square -wave response and the equivalent blur of a single stage
of a video voltage amplifier of good fidelity in a sÿstem in which many
similar stages exist. Usually the distortion of a single stage is so
small that only the accumulated effect of several stages is clearly evident in the over-all square -wave response. The writers are not aware
of the existence of a practicable method of combining the individual
rr B. D. Loughlin, "A Phase Curve Tracer for Television ", Proc. I.R.E.,
Vol. 29, pp. 107 -115; March, 1941. Loughlin describes apparatus which
furnishes a direct plot of phase versus frequency on the screen of a cathode ray tube. The complexity of the apparatus may limit its general utility.
12 M. E. Strieby and J. F. Wentz, "Television Transmission Over Wire
Lines ", Bell Sys. Tech. Jour., Vol. 20, pp. 62-81; January, 1941.
www.americanradiohistory.com
TRANSIENT RESPONSE
141
square -wave response of two or more units for the purpose of finding
the over-all square -wave response. Even if a 'feasible method should
exist, the experimental errors involved in many individual measurements would jeopardize the dependability of the calculated over-all
response.
If units of good fidelity must be considered individually, it becomes
expedient to determine first the sine -wave characteristics and then to
combine the sine -wave characteristics so that a synthesis of the overall square -wave response may be performed.
The accuracy of square-wave methods is more than adequate to
reveal those imperfections of transmission which cause discernible
effects in a television picture. Therefore, the authors hope that this
paper will be a contribution to the general problem of working out
electrical specifications for television transmitters and other television
apparatus, giving information regarding the steepness of rise and the
amplitude of over -swing of the square -wave response.
APPENDIX I
ANALYSIS OF SQUARE -WAVE RESPONSE
Let the square -wave response be approximated by a stepped wave f
as shown in Figure 1. The kth step may be approximated by the series
---
2N1
C1
Ak
2
-sinno,
1
7r
n
(t- TK)
(3)
where N is very large.
The sum of M steps leads to the following expression for the stepped
wave
ar
Am
2
ai
m=1
2
a
m=1
[Am
N
- sin
1
n=1 n
nwp
(t
-
T,,,)
(4)
The constant term in (3) is not of interest. If the order of summation
of the second term is reversed, there results
2
N
1
7f
nL1
n
M
>
[A1 sin
(t- Tm)
(5)
1
The inner sum written for some specific value of n as N1 is the total
N1th harmonic content of the stepped wave. That is,
www.americanradiohistory.com
TELEVISION, Volume IV
142
2
EN1
=
if
1
_-2
1
7r
N1
Am
m=1
n N1
sin
N1roo
(t
-
Tm)
(6)
BN1
sin N1cil0 (t -tN1).
Since the input square -wave signal may be approximated by the form
- -7r"=1n- sin
2
1
N
N
I
moot
2
it follows that the
N1
component of the input is
(b)
Ca)
(a) Compensated resistancecapacitance amplifier.
Fig.
- - sin
2
1
7r
N1
(b) Equivalent circuit for
high frequencies.
21
N10ot.
The amplitude response of the circuit is, therefore, 13N1 and the
time delay is TN1. These two quantities may be found graphically as
indicated in Figure 1(a) and (b).
Since o , may be allowed to approach zero, it follows that reference
to a fundamental frequency is not required and the amplitude and
phase corresponding to any frequency may be found.
APPENDIX
II
THE COMPENSATED RESISTANCE- CAPACITANCE AMPLIFIER
A schematic diagram of the amplifier appears in Figure 21. When
Z (L, C, R) the response of the equivalent circuit to a unit func-
Rs»
tion is expressed by the following equation:
www.americanradiohistory.com
TRANSIENT RESPONSE
e
=
1tR
[1- e- 7f,Kt(A sin
143
+ ¢) + Bt sin
(S21t
(E2,t
+P}]
Rp
in which
= time
fo = 1/ (27r LC)
A= V1+M2
K= 2afoRC
-3+3K2-K4/2
t
M=
N=
(1- K2/4)
rfo(K2 -1)
4K
01.=
B=
27rfoN/1
VP'-'
- K2/4
+ N2
1
¢
= tan--'M
ß=
312
P=
K2V1-K2/4
tan'-NP
rio (3 -K2)
2K(1-K2/4)
.
The amplitude response is
amplitude =
-
1+ B2 /K2
B2K2 + (B2-1)
pi?
Rp
i
delay
2af
tan
in which
f
'
(1-K2- B2) B/K
- frequency
B=
K
2
f/fo
= 27rfoRC.
APPENDIX
III
RELATION BETWEEN SQUARE -WAVE ANALYSIS AND
FOURIER ANALYSIS
Assume that the response of a linear electrical network to a unit
function is represented by e1(t) in Figure 22(a). The response e2(t)
of the same circuit to a square pulse Lt seconds long is found by shifting el (t) to the right along the time axis by an amount At and subtracting the result from e1(t) that is, e2(t) =e1(t) -e1(t + At). If
the condition is imposed that el (t) = 1 for t ? t9, then e2 (t) = 0 for
t
www.americanradiohistory.com
TELEVISION, Volume IV
144
It is clear that the waveform of the response e3(t) of the same
circuit to the periodic pulse wave of Figure 22(b) is given for each
cycle by e2 (t) if To ? (tq + Lt). Hence, the coefficients of the Fourier
series written for e1(t) may be expressed in terms of the ordinate
readings of el (t) taken at successive equal time intervals. Thus
e3(t)
=ao+al cos 2afot+-f-ak cos K2rfot-}Fan /2
COS
-n
2irfot
b1
sin 2irfot+
1
l
2
+bk sin K2rrfot-{+
(n /21
-1 sin (
Y
Y
t)
1
+et.11(0)
:
;'
Y3
'
2
-
2'rfot
-At)
4(b
`
Yo
iI/
Ya
L
¡
I
NI
Y,
YQ
t, ta
t,
TIME
-
(h)
TIME
Fig. 22- Fourier analysis applied to the response of a circuit to a square
pulse compared with square -wave analysis applied to the response to a
unit function.
where
ak
= 2Atfo[ (y1
+
+
(y2
(yq
-
-yi) cos K4rrAtfo
-
yq
+
yo) cos K27rAtfo
-i) cos Kg27rptf0]
= 2,n,tf 0A
www.americanradiohistory.com
TRANSIENT RESPONSE
and
bk
= 2Atfo[ (yl
+
+
(y2
(Yq
-
145
yo) sin K27rAtfo
- sin K4TAtfo +.- Yq_i) sinkg27rAtfo]
y1)
= 2Atf0B
The Fourier series for E (t) is
E2 (t)
- sin Kr At- cos
_ -+ To
2
At
7r
1
k
27rKf ot
T
1.6
30-M0 DOTS
1.0
14
12
10
8
MC
FREQUENCY
for
obtaining Fourier analysis from square -wave
Fig. 23- Conversion factor
analysis. C times amplitude according to square -wave analysis equals
amplitude according to Fourier analysis.
o
2
4
6
Hence, the amplitude characteristic of the circuit is given by
R(K,f0) =
or
limit
fo--->0
R(K, fo)
=R(f) _
7rAtKf 0
sin 7rAtKf
7rpif
VA2 +B2
limit
f *OVA2 +
sin 7rPtf
and the delay characteristic is given by
www.americanradiohistory.com
B2
TELEVISION, Volume IV
146
limit
6
(K, fo)
f0-- 0
27rf
B
=-tan-1-.
f
A
1
27r
Now '/A2 + B2 is the same expression which is obtained for the amplitude characteristic by the graphical method of square -wave analysis.
Hence, 7rptf/sin 7rf tf may be regarded as the conversion factor which
converts square -wave analysis into Fourier analysis when latter is
based on the square pulse.
The magnitude of the conversion factor is shown in Figure 23. It
may be concluded that for most applications of square -wave analysis
in television, the conversion factor may be taken equal to 1. The delay
characteristics as determined by square -wave analysis and Fourier
analysis are identical.
www.americanradiohistory.com
TRANSMISSION OF TELEVISION SOUND ON THE
PICTURE CARRIER`t
BY
GORDON L. FREDENDALL, KURT SCHLESINGER$ AND A. C. SCHROEDER
Research Department, RCA Laboratories Division.
Princeton. N. J.
Summary- Several pulse methods for the transmission of television
hound on the picture carrier during the line- blanking intervals are analyzed
from the points of view of signal -to -noise ratio, audio fidelity, and transmitter and receiver design.
The advantages of duplex transmission are: (1) elimination of a
separate sound transmitter; (2) elimination of the ambiguity and difficulty
which may occur when a standard frequency- modulated sound signal is
tuned in; (3) freedom of the audio output from the type of distortion which
occurs in frequency -modulated receivers as a consequence of excessive drift
of the frequency of the local oscillator; and (4) improvement of the phase
characteristic of the picture intermediate -frequency amplifier resulting from
elimination of trap circuits.
The highest audio- modulation frequency in duplex systems must not
exceed one half of the line-scanning frequency. This is a disadvantage
under the present television standards which specify a line frequency of
15,750 cycles per second.
With the exception of pulsed frequency modulation, the signal -to -noise
ratios of sound in duplex systems are not so great as the ratio offered by
the transmission of a standard frequency-modulated carrier. The comparison
is subject to the condition that the amplitude of the frequency-modulated
carrier is 0.7 of the peak amplitude of the duplex carrier. The signal-tonoise ratio of a pulsed frequency- modulated signal may equal the ratio of a
standard frequency- modulated signal up to a critical distance from the
transmitter, but is less at greater distance.
INTRODUCTION
FROM time to time, proposals have been made that the sound
accompaniment for television may be transmitted by a modulation of the picture carrier during the line- blanking intervals when
no picture detail is transmitted. Improved reception of picture and
sound, decreased investment in receivers and transmitters, and greater
channel width for the picture signal are mentioned as possibilities of a
"duplex" transmission of picture and sound. The purpose of this paper
is to assist engineers in their evaluation of duplex transmission as a
*
Decimal classification: R583.
t Reprinted
from Proc. I.R.E., February, 1946.
Columbia Broadcasting System, New York, N. Y. Work covered in
this paper was done while the author was a member of RCA Laboratories,
Purdue University, Lafayette, Ind.
$
147
www.americanradiohistory.com
TELEVISION, Volume IV
148
practicable system by offering an analysis of several methods of
duplexing.
A review of the method of sound transmission and reception in use
is helpful as a setting for the discussion. In the present arrangement,
sound is transmitted by a frequency- modulated sound transmitter operating on a carrier frequency which is 4.5 megacycles above the picture
carrier. At the position of the sound carrier, the recommended standards' state that the field strength of the picture sidebands shall not
exceed 0.0005 of the picture carrier. There is essentially nothing at
the transmitting point to distinguish a television sound transmitter
1
=ALONG
Mt SYIC
PULS(
INTERVAL
SASE
RSA
PRIMAL
INTEW/L
ECWIRPO
CAPA1ER
RACK LEVEL
WHITE LEVEL
CUPID
SYNC
PO3(3
-
WI IDO%
why 123)%
x-'F-W 3 M-1
W 3
1
IERO
IOITAL
03
MATACE
°/
444302M:ó
`..
-0
Dz
AVM
VERT !CAL
KA SLAPPING
Ru0060 0
Cf
PET13E
OS/ 1C.2/
E0E NOTE
3
NIDS)
-
BOTTOM
10E
TEA OF PCTIME
2
WMZORN 0E01306 NM 10 SCALE
3
RA01
Mpl
LEVEL
ST NC
6100
1001
0203
MlxY
3-3042
i
--I/1001
1OIC1
MATE
IERO
LEVEL- -'
5001
In0020 MM
'
1/13 Of YAY
5
ux
~
M)0004N
MOTE'
N100040
1000040
PAX.
-MTAt 11181
START OF VOL
2
-VTHE 1(011
START GE
3
-1(4004
1
SANS0
MAi
I/4 01
PULP
/
II)0040
SYNE
IT1050---
5EE NOTEO
WYOO
D1007M 10014
TO
STMT
OE OCST LAIC
START Of
0[01
00111
0(030060
430 TRALING SLOPES O NOR120NTA. BLAMING
W5T H STEEP 04401100 TO P ESERr0 MN 440 NAS
VALUES 01n``I ANOoIDNDER ALL CONDITdS 01
C TORE CANTENl
WHOA
MK PULSE
LIE
ME FELD 30
TRAILING EDGES 01 0[81401
COMPLETE 6 LESS 1,041040
AMO
H
4 -LFA304G
MAX SYNC
- EQUALIZING
((1NL HIWEEN
440140 MM
5
DAMNING
0004H
RACK LEAL
p4q0lIIL
LEVEL
M
j
SPEC
RACK
1,
J
CARRER/
4-41N
1/10 O1 MAX
5-561
RAN00NG
SEE NOTE
SYNC.
MAX
A
((UL Hi6E01
(OONO'1N1L
REAR SUSS
ART
woW4M
NEA
SEE NOTE
((TAIL (EINSEN
4
64
5- 0114030615 watt) MT0 AN ASTEMS0 IDEATE THAT
TRERUOE511114 ARE PERMITTED HEY FOR LONG TINE
VARMT0S. AND lOT FOR SUCCC55AE 011CLE5
AREA SAVA H B OWEEN 045 AND
0- 0010110104 ORS(
05 DE THE AEA
i
-FEHR
TO TEAT
Cf A 001I2ONTAL
FOR
1UPTNER
SYNC
ORSE
EVLAIA1g.S
AND
10ENANCES
Fig.
1-Television
synchronizing wave form.
from a conventional frequency- modulation transmitter designed for
operation in the frequency -modulation band. The sound receiver is
likewise conventional and may share only the same heterodyne oscillator with the picture receiver.2
The picture transmitter is amplitude -modulated and radiates a
wave form illustrated by Figure 1. Picture content is transmitted
Final report of the Radio Technical Planning Board.
In some designs, the sound and picture signals are amplified at intermediate frequency in one or more common stages before branching off into
oeparate intermediate -frequency amplifiers occurs.
2
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
149
during about 85 per cent of the total "time on the air." No picture
detail is transmitted during the blanking intervals of the scanning
tubes in the transmitter and receiver. Such "idle" intervals amount to
about 15 per cent of the total time.
Proposals3.4 for duplexing have been directed, therefore, toward
the utilization of some part of the blanking interval for sound transmission. Thus the television transmitter would be converted into a
picture -sound duplex transmitter which radiates picture intelligence
during 85 per cent of the time and sound intelligence during some part
of the remaining 15 per cent, using only one antenna and one radio frequency power amplifier. There would be a synchronized electronic
switch in the receiver for the opening of the sound channel to the
video signal sometime during the blanking interval.
Factors which are involved in a comparison of duplex methods and
the present method of continuous transmission of sound are the following: (1) audio fidelity; (2) signal -to -noise ratio; (3) amount of interaction between video and audio signals; (4) permanency of receiver
alignment; (5) picture quality; (6) ease of receiver tuning; (7) cost
of receiver; (8) cost of transmitter.
Certain advantages and disadvantages of a duplex system may be
predicted in advance of a theoretical and experimental analysis. First,
there is no separate sound transmitter and antenna. This economic
advantage can not be accorded much weight unless there is a resultant
economy in the television receiver because the ratio of receivers to
transmitters is so great that the economics of the receiver is the controlling factor.
A significant advantage of a duplex receiver is the freedom of the
audio output from the type of distortion which occurs in frequency modulation receivers as a consequence of excessive drift of the frequency of the local oscillator. Audio modulation is conveyed in a duplex
system by the envelope of a radio -frequency carrier and is relatively
unaffected by instability of the local oscillator.
Sound -rejector circuits in the picture intermediate -frequency amplifier can be removed with a resulting reduction in phase distortion and
some improvement in picture quality.
The ambiguity involved in tuning -in a conventional frequency modulation signal is removed. In the present system a choice must be
3 H. E. Kallman, "Audio
and video on a single carrier ", Electronics, vol.
14, pp. 39 -42; May, 1941.
4 Numerous patents including:
U. S. patents, No. 1,655,543, R. A.
Heising; No. 1,887,237, J. L. Finch; No. 2,061,734, R. D. Kell; No. 2,083,245,
H. Shore and J. N. Whittaker; No. 2,086,918, D. G. C. Luck; No. 2,089,639,
A. V. Bedford; No. 2,227,108, H. A. Rosenstein; No. 2,257,562, H. Branson;
No. 20,153, E. F. W. Alexanderson (reissue).
www.americanradiohistory.com
TELEVISION, Volume IV
150
made of tuning for minimum interference in the picture or minimum
noise in the sound in receivers which are somewhat misaligned.
There is also available a small extension of the video band into the
space now assigned as a guard band for the sound; this amounts to
about one quarter of a megacycle.
It may be anticipated that the signal -to-noise ratio with duplex
sound would be unfavorable as a consequence of the reduced time for
transmission of sound. This is a serious obstacle in some duplex systems. A further limitation is the imposition of a maximum audio frequency that may be transmitted without the introduction of spurious
frequencies into the audio spectrum. This restriction has been noted
before in pulse transmission of sound. Finally, there is the complication of synchronizing the electronic sound selector in the receiver with
the line -scanning frequency.
(A)
[I
fl
H
M(k)
f1
(B)
r
(A)
(B)
(C)
r,
2-
Fig.
Amplitude- modulated duplex wave forms.
pulse carrier
audio wave form M(t) and amplitude -modulated pulse carrier
amplitude -modulated pulses combined with television wave form
DUPLEX METHODS
Amplitude-Modulated Pulses
One of the most obvious duplex systems is the modulation of the
amplitude of a rectangular pulse wave in accordance with the audio
signal and the insertion of the modulated pulses in the line -blanking
interval of the video signal. Figure 2 illustrates the successive steps
at the transmitter. The pulse wave form shown in (A) of Figure 2
is amplitude modulated by the audio signal M(t), as illustrated in (B).
In (C) the modulated pulses have been inserted in the part of the
blanking interval following the synchronizing pulse. Such a composite
wave would be applied as modulation of the picture carrier.
If it is granted that the synchronized electronic switch in the
receiver is able to select the pulses from blanking so that the pulse
wave in Figure 2 (B) is recovered, the audio fidelity of the duplex
1.
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
151
system can be found from the solution for the frequency components
of (B). An analysis shows that the spectrum consists of the applied
audio -modulation frequency fo and a large number of sidebands
(nf0 ± fo) in which fc is the funda(fa fo), (2f, -!- fo) (3f, ± fo)
mental frequency (line frequency) of the pulse wave.' The amplitude
of each group obeys the damped sine -wave law sin n7rr /n where n is the
order of the sideband and r is the ratio of the width of the pulse to
the fundamental period. In a television application, r could not exceed
about 0.06 and hence the amplitude factor, sin nor/n changes slowly.
When fo exceeds 1 /2fc, there is overlapping of the first -order lower
sideband and the audio -frequency component, as well as general overlapping of adjacent sidebands of higher order. We do not have knowledge of any detector whereby an undistorted audio signal can be recovered from this multiplicity of overlapping sidebands. However, if the
frequency of the audio modulation is restricted by a low -pass filter at
the transmitter to less than one half the fundamental pulse frequency,
this confusion is avoided. A similar filter must be installed in the
receiver for the rejection of frequencies above f0/2. Such a low-pass
filter in the receiver functions as a distortionless detector of the audio
modulation. Therefore, the theoretical upper limit of the audio bandwidth is equal to one -half of line frequency, or 7875 cycles with the
present standards.
Vertically scanned pictures would allow a greater maximum audio
frequency of 4/3 X 7875, or 10,500 cycles.' However, it has been
observed in laboratory tests that moving subjects scanned vertically
with an interlaced pattern do not, in general, reproduce with as much
detail as horizontally scanned subjects. This is due, probably, to the
predominance of horizontal motion in average subject matter. Hence,
it appears that vertical scanning must be rejected as a means of increasing the maximum audio frequency.
The most promising way of increasing the upper audio limit in a
monochrome system is an increase in the video bandwidth. The two
quantities are related by the formula?
fa = Kl V fv
(1)
in which
fa = maximum audio frequency
Appendix I.
The line frequency in a vertical scanning system is 4/3 the line
frequency of the standard horizontal scanning system with an aspect ratio
of 4 to 3.
7 Appendix II.
5
www.americanradiohistory.com
TELEVISION, Volume IV
152
f = video bandwidth
K1= a constant.
Figure 3 shows the correlation between sound band, video band,
and number of lines for monochrome and color transmissions. The
latter is assumed to be a sequential tricolor system with an interlace
ratio of 2:1 and a color -field frequency of 120 cycles.' For a given
video bandwidth, the quality of duplex sound in terms of audio bandwidth may be made 1.4 times better for color than for monochrome
television. Thus, high -fidelity sound (11,000 cycles) may be duplexed
along with a color picture of about 360 lines over a video channel of
4 megacycles, while a monochrome picture of 525 lines, which occupies
the same video band, accommodates only about 7800 cycles. A maxi-
lip!
1200
600
*ON E
W
N
ONO
1000 Z
3
400 aoo
If
0
II
S
10
c
w
600 m
Z
20 400
Z
100 200
IS
o
20
V1D20 IANDWIDTN - MC.
Fig.
audio frequency versus
3- Maximum and
number of lines.
video bandwidth
mum audio frequency of 11,000 cycles would require over 700 lines in
a monochrome system.
The corresponding radio-frequency channel in all cases is approximately 30 per cent greater than the video bandwidth as a consequence
of the additional space required by the vestigial sideband.
A proposal for sound transmission has been disclosed which removes
the limitation on the maximum audio frequency of one -half line frequency.' In effect, the system provides for the modulation of a rectangular pulse carrier of two times line frequency and the subsequent
delay of alternate pulses to a time position which permits transmission
of pairs of pulses during horizontal blanking time. At the receiving
' P. C. Goldmark, J. N. Dyer, E. R. Piore, and J. M. Hollywood, "Color
television ", Proc. I.R.E., vol. 30, pp. 162 -182; April, 1942.
9 A. V. Bedford, U. S. Patent No. 2,089,639.
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
153
point, the previously undelayed pulses are delayed before detection, thus
restoring the modulated pulse signal to its original form as a wave of
double line frequency. The maximum audio frequency has thereby
been increased to line frequency. This system, however, does not appear
to be economically feasible from the point of view of receiver design
at the present time.
TABLE I
(root- mean -square)
SIGNAL -TO -NOISE RATIOS
(1)
Method of
Transmission
Standard
Amplitude
Modulation
-
AmplitudeModulated
Pulses
fae12
4
N
3w
P
2l,
Dissymetrical
Width-Modulated Pulses
2
Frequency
Modulation
fa
J!
3w
_
4w
P
4 ,/-2.w
t, '1f,, N
_
P
Sfa,12
N
3J6Jrfd
P
8f."2
N
_ 3w
J 21, Jfo
is
fd
J6Jr
Jf
fa
2/23J6Jrfd
8ía3,2
-
217
d2
42 .Jr K
24
None
K
0.417
d2
3w
t,
K
fd =150
fd =150
kilocycles kilocycles
K
4/ Jfa
-
N
3J6Jrfd
3
None
d2
-- -
1,/b 1\
N
f.
,1
Symmetrical
Width -Modulated Pulses
Pulses of
-
0.023
d2
J3fd
Jett
:(Id)312
)312
J2 Jr P
3
fa
NI
J3Id P
Modulation
K
2
None
Jfa N
Frequency
--
(Critical)
2P
Standard
Signal -to- Signal-toSignal -toNoise
Noise
Noise
P/N =K /d2 P /N =K /d2 (Critical)
Signal -toNoise
Signal -toNoise
(5)
(4)
(3)
(2)
0.003
K
None
d2
d2
K
K
0.012-
d2
62
d2
- 0.016K
K
d'°
d2
87
id =150 fd =150
kilocycles kilocycles
K
dl
K
0.052-
107
d2
fd =1200
kilocycles kilocycles
K
0.417
2420
fd =1200
Pulses of
Frequency
Modulation
116
Jr
/fd yi23-46Jrfd
K
/)
1\
fa
8!312
d2
d2
The signal -to-noise formulas in columns (1), (3), and (4) of the table for standard
frequency modulation, width -modulated pulses, and pulses of frequency modulation
during postblanking, are valid only for ratios higher than the critical ratio since the
formulas are derived with the assumption that noise limiting is effective. Thus it
may appear, with only a casual reading of the table, that the ratio is alwa s the
same for standard frequency modulation and pulses of frequency modulation
fd =1200 kilocycles. The fact is that the two types of transmission yield equal signal to -noise ratios only when the critical ratio for pulses of frequency modulation is
exceeded. At greater distances from the transmitter standard frequency modulation
is superior.
Values of constants:
cycles per second.
f = 7500 cycles per second; r =0.06;
w /t,
=15.4;
f.= 4 X 106
Success or failure of a method of transmission often rests on the
degree of immunity to noise. The signal -to -noise ratios appearing in
Table I provide a direct comparison of amplitude -modulation pulse
transmission and other systems. Comments on the significance of the
ratios, and the bearing on modulated pulses as an audio service for
www.americanradiohistory.com
TELEVISION. Volume IV
154
television are made later. It is clear that the amplitude -modulation
pulses suffer a disadvantage in that superimposed noise cannot be
reduced by amplitude limiting to the extent possible in certain other
systems.
2. Width- Modulated
Rectangular Pulses
A more promising form of pulse modulation, from the standpoint
of signal-to -noise ratio, is a constant -amplitude pulse system wherein
the width of a pulse is proportional to the amplitude of the audio
signal. Two examples of width -modulated signals are illustrated in
Figure 4(B) and (C). In type (1), a dissymmetrical modulation, the
leading edges of the pulses occur periodically, but the widths are pro(t)
(B)
I
-TYPE (11
TYPE (2)
(c)
(o)
(A)
(B)
(C)
(D)
n;
n_
n;p
ni_
4-
Fig.
Width -modulated duplex wave forms.
audio modulation M(t)
pulse carrier dissymmetrically width -modulated by M(t)type (1)
pulse carrier symmetrically width -modulated by M(t) -type (2)
combination of type (2) and television wave form
portional to the instantaneous amplitude of the audio signal M(t) at
the instant of the leading edge: that is, only the trailing edge is "modulated." Type (2) is a symmetrical modulation, the width of a pulse
being proportional to the instantaneous amplitude at the instant corresponding to the center line of the unmodulated pulse. The center lines
are periodically spaced; thus, both leading and trailing edges of type
(2) are modulated. Such pulse waves may be inverted in polarity and
combined with the standard television wave form as shown in Figure
4 (D) for type (2) modulation. Any amplitude variation of the pulse
due to noise may be removed by limiting in the receiver following
separation of the pulse from the video signal if the peak noise does not
exceed one half of the pulse amplitude.
Equation (28) in Appendix III is the expression for a pulse wave
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
155
width -modulated in the symmetrical manner (type 2) by a sine wave.
The similarity to standard frequency modulation is noticeable in the
sequence of sidebands which are generated. Thus when a pulse carrier
of fundamental frequency fD is width -modulated at a rate of fo cycles
per second, the resultant wave contains component frequencies fe,
(f0 + fo), (f0
fo), (fc + 2fo), (f0-2f0), etc., as well as corresponding sidebands for each harmonic of the fundamental f0; namely 2fD,
(2f, + fe), (2fr fo), (2f0 + 2f 0), (2f0 2f0), etc. In addition, the
frequency terms containing only the modulating frequency fo and its
harmonics 2f0, 3f0, etc. appear. The general term is (Mf0 - Nfo)
-
-
-
TYPES (I) AND(2)
!
(f,-f,)
_
I.
-TYPEN
TYPE(1) 2E0
(fc-2fe).
01
E
E
C
........TYPE(
iJ
.........- TYPE
-...
E
>
00
(2)
2
(fc-2fe)
/
.....
f
TYPE (2 3fe
/
.
/
[
....
0001
e-
s.
TYPE
(2)(fc-3fe)
TYPE (2) 4fo
.
0.2
0.4
06
fc /fe
Fig.
5- Frequency components resulting from
f,=
width modulation of a
pulse carrier wave by a sine wave.
Type (1) = dissymmetrical modulation
Type (2) = symmetrical modulation
pulse frequency fo = modulating frequency
where M and N are positive integers or zero.
Since overlapping of fo and the sideband (f0 fo) must be prevented, the modulating frequency should not exceed f0/2. At the receiving end, the modulated pulse -signal may be applied to a low -pass
filter which rejects all sidebands and harmonics exceeding one half the
frequency of the fundamental f6. However, all distortion terms are not
thereby excluded; harmonics of fo and the sidebands of higher order
(fa 2f0), (f0 3f0), may fall within the pass band. The magnitudes
of the most important distortion terms in (28) have been plotted in
Figure 5. The modulation constant « was taken equal to 1, the value
-
-
-
www.americanradiohistory.com
TELEVISION, Volume IV
156
corresponding to maximum modulation. A value of 3 per cent was
assigned to w, the unmodulated pulse width. Therefore, the widths of
the pulses in the modulated wave vary from 0 to 6 per cent of the
period of the carrier. This is substantially the maximum variation
under the specifications given for the television synchronizing wave
form (Figure 1). The broken-line portion of a curve indicates the
range of the component which is suppressed by the low-pass filter in
the receiver. Thus the component (f0 -2fo) is suppressed for the
range fo < f0/4 but is transmitted when fo > fc/4. In a reverse manner the component 2f0 is transmitted when fo < f,. /4 and suppressed
when fo > fD/4. The maximum value attained by either component is
approximately 0.05 per cent of the amplitude of the audio component f 0.
316NA`
TI'VIDEO
N SYNC< 1
IANNINO
LLL---MIYfq
(A)
2
TO ACTUAL
TqANSMITT ER
TRANSMITTER
I
OUTPUT
11
1
II
1
ti
LIMITERI
OUTPUT
LIMITER PL
1
~L
p
'I
I
MII%ER
n
1
01
1'
1
1
SPECIAL
VERTICAL
I
SYNC.
SLOTTIN
(B) GENERATION OF WIDTH -MODULATED PULSE
Fig.
6-Width- modulated
duplex -system transmitter.
An analysis of a dissymmetrical width -modulated pulse wave (Appendix III, equation (31)) displays the same general characteristics
as the symmetrical modulation. Harmonics of the audio frequency, as
well as numerous sidebands, are present, but the magnitudes shown in
Figure 5 are greater than in type (2) modulation. The largest contribution of any distortion term is 2.5 per cent.
Signal -to -noise ratios calculated according to the derivations in
Appendix IV appear in Table I and Figure 15.
Figure 6(A) illustrates one method for the production of dissymmetrical width -modulated pulses. The starting point is a wave of
narrow triangular pulses (Figure 6 (B) ) which is derived from driving
pulses at line frequency normally generated by the synchronizing generator. To the triangular pulses is added the audio signal from which
the frequency components higher than one -half the line frequency have
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
157
been removed by a low-pass filter. Limiter No. 1 removes the audio
wave below the base line as shown in Figure 6(B). The residue is
greatly amplified and then acted upon by limiter No. 2, with the result
that width -modulated pulses of substantially rectangular shape are
produced. These are inserted in the line -blanking interval following
the synchronizing pulse (postblanking) The standard field synchronizing pulse must be slotted down to black level during the line pulses,
as shown in Figure 6 (B), in order that the width -modulated pulses
when applied may extend to white level. The combined video signal is
applied to the picture transmitter in the customary manner.
Figure 7(A) illustrates the functional arrangement of the receiver.
The selection of the width -modulated pulses from the video signal and
rejection of picture components is performed by an electronic switch,
.
ELECTRONIC
SWITCH
TIMING XRO.A
LINE SCANNING
GENERATOR
MULTIVIBRATOR
DUPLEX
VIDEO SIGNAL
(A)
,GRID
GRID
DUPLEX RECEIVER
3
I
OUTPUT
(B)
Fig.
IOPERATION
7- Width-modulated
OF ELECTRONIC SWITCH
duplex system. Type (1) receiver.
usually a vacuum tube having two control grids. A keying pulse
originating in a multivibrator which is synchronized by the line scanning circuit is impressed on grid 1 of the switch. The width and
timing of the pulse is critical for the most favorable signal -to-noise
ratio. The duplex video applied to grid 3 causes plate current to flow
only when the tube is keyed on. In this way the width-modulated pulses
are isolated.
Amplitude noise is removed by limiting if the peak noise does not
exceed one-half of the pulse amplitude but the variation of the pulse
width due tonoise is not removable. Such variation constitutes a width
modulation and is reproduced as audible noise. When the synchronization of the receiver is impaired by noise, the timing of the switch is
likewise affected, and parts of blanking and picture may be admitted
into the audio amplifier and appear as noise in the loud speaker. There
www.americanradiohistory.com
TELEVISION, Volume IV
158
is a marked increase in the immunity of the system to noise if the
receiver is synchronized by automatic frequency control."
Audio components in excess of one-half line frequency are removed
by a low-pass filter.
Additional kinescope blanking must be provided in the receiver
since the duplex signal extends to white level during the sound pulse.
In Figure 7(A) blanking is derived from the multivibrator simultane-
ously with the keying pulses.
Means for excluding signal from the audio circuits when the receiver is not in synchronism is very desirable. Without such a device,
video components are admitted to the audio system with an annoying
audible result. A circuit may be devised which is sensitive to the
changed character of the signal passed by the electronic switch during
intervals of missynchronization and applies a bias beyond cutoff to
the audio amplifier.
In September, 1943, television signals containing width-modulated
pulses of the dissymmetrical type were transmitted by television station
WNBT, and successfully received in Princeton, New Jersey, using the
system outlined above.
Pulse Time Modulation
Another form of modulation known as "pulse time modulation" is
related to width modulation." In pulse time modulation the pulse
amplitude and width remain constant, but the time interval between
successive pulses is varied in accordance with the instantaneous amplitude of the audio signal and the rate of this variation corresponds to
the instantaneous frequency of the signal. Such a pulse wave may be
regarded as the sum of two width -modulated pulse waves of the dissymmetrical type of opposite polarities as illustrated in Figure 17.
The frequency components of the pulse time wave are therefore solvable from (31).
3.
Pulsed Frequency Modulation
In contrast with the foregoing duplex methods involving rectangular pulses for the transmission of sound during the line -blanking
interval, there is a method which may be called "pulsed frequency
modulation," that employs wave bursts of a frequ.ency- modulated sub carrier for the same purpose. The bursts are generated at the trans4.
10 K. R. Wendt and G. L. Fredendall, "Automatic frequency and phase
control of synchronization in television receivers ", Proc. I.R.E., vol. 31, pp.
7 -15;
January, 1943.
E. M. Deloraine and Emil Labin, "Pulse time modulation ", Elec.
Commun., vol. 22, pp. 91-98; 1944.
11
www.americanradiohistory.com
0
TELEVISION SOUND TRANSMISSION
159
mitter by a sine -wave oscillator which is operative only during line
blanking and is frequency modulated by the audio signal during this
interval. The center frequency and deviation are chosen so that the
essential sidebands lie within the video band. These subcarrier bursts
are combined with the video wave form as modulation of either the
line synchronizing pulses or the post -blanking (Figure 8). From the
point of view of signal -to-noise ratio, Figure 8 (B) is preferable. In
either case, the composite signal is applied as amplitude modulation
of the radio-frequency carrier.
In the receiver, the bursts are first isolated at the video level from
the picture part of the composite wave, then amplitude limited for
removal of noise, and finally applied to a conventional balanced frequency-modulation discriminator centered at the frequency of the sub carrier. The output of the discriminator is an amplitude-modulated
pulse wave. The audio signal is derived from the pulse output of the
discriminator by removing all components in excess of one -half line
1
IL
.111_
iNÌ
,.
f-Y.,,, .1.,ß
8-Pulsed
Fig.
(A)
9L
frequency -modulation duplex wave forms.
(B) in post blanking
in synchronizing pulse
frequency with a low-pass filter.
As the result of experimental and theoretical work with pulsed
frequency modulation, certain features of the technique were discovered which could escape a casual study. Such matters are treated in
the following discussion.
A. Transient response of pulsed frequency -modulated circuits: If
a pulsed frequency- modulated system is to function properly, the peak
value of the detected pulses should depend solely on the instantaneous
frequency of the subcarrier. This means that the various tuned circuits
involved should complete their transients in a time which is short
compared with the total duration of the wave burst. Figure 9 shows
the response of a simple tuned circuit to a wave burst of constant
amplitude that starts and stops with zero phase and lasts TP seconds.
The circuit has a build -up time r, which may be adjusted by the damping resistor R and is correlated with the bandwidth b in the form
www.americanradiohistory.com
TELEVISION, Volume IV
160
r
= 2RC =1/7rb.
(2)
The minimum bandwidth is determined by the time allowed for the
transients. If these are to be complete within p percent of the pulse
time,
b?
100
(3)
apTP
The total number n of cycles per pulse, as well as the number
occurring before the steady state is attained, are
n=
A97
(4)
TvfB
An =718
(5)
WAVE BURST
Fig.
9- Response of a tuned circuit to a pulse of frequency modulation.
Equation (5) holds regardless of the subcarrier frequency fa. The
following set of constants is representative of a typical circuit designed
for pulsed frequency -modulation operation:
Pulse time T
Time of build -up r
Circuit capacitance C
Circuit resistance R
Bandwidth b
Subcarrier frequency
Q
factor
5
microseconds
0.5 microsecond
25 micromicrofarads
10,000 ohms
400 kilocycles
f8
4
megacycles
10
Cycles per pulse n
20
An
2 cycles
B. Generation of phasing of the subcarrier at the transmitter:
According to the calculation above, a'total variation of no more than
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
161
subcarrier cycles is sufficient to produce peak modulation in the
receiver. Hence it follows that the instantaneous wave forms of the
subcarrier bursts must be closely similar at the beginning. Unless the
initial phase of each burst is repeated with extreme accuracy, the
otherwise random initial phases may introduce audible beat notes and
noise in the detected signal.
_In this connection, the keying of the subcarrier for part -time modulation presents a major problem. If an independent subcarrier oscillator
supplying a continuous frequency- modulated wave is used, an electronic
on -off switch controlled by the main synchronizing generator must be
provided. This switch cuts into the subcarrier and admits sections of
its wave train for modulation of the synchronizing pulses (or blanking) . It is obvious that the timing of this switch would have to be
accurate within small fractions of one subcarrier cycle, or about 0.1
microsecond, in order that the initial phases of all pulses be substantially equal. The noise susceptibility of this method is high, because
the subcarrier modulation is keyed on and off at full amplitude.
The problem of precise keying is further aggravated by the fact
that the repetition frequency of television pulses is not constant. In
all practical television synchronizing generators, the line frequency is
subjected to continuous frequency control so that it constitutes, at any
instant, a definite multiple of the field frequency. The field frequency
is synchronized with a 60 -cycle power supply which is inherently
variable around a well- defined average. As a result, beat notes of
variable pitch are bound to occur if a subcarrier source with continuous
frequency modulation and constant center frequency is subjected to
keying from a synchronizing generator unless special precautions are
taken.
In the system described below, such spurious signals have been
effectively eliminated. A continuous subcarrier generator is not used
instead, the bursts are supplied from a start -stop oscillator which is
switched on and off by the line blanking pulses. The start -stop sub carrier oscillator shown at (4) in Figure 10 is active only when plate
voltage is applied in the form of a pulse from the control tube (3).
Pulses of appropriate wave shape at line frequency are derived directly
from the synchronizing generator and impressed on the grid of the
control tube. Hence throughout the line- scanning interval the sub carrier oscillator is inoperative, but at the end of each line it receives
a plate -voltage pulse. As a result, subcarrier oscillations are built up
with exactly the same initial phase conditions each time. Since the
plate-power pulse is derived from the line-blanking pulse, it participates automatically in any variations of the line frequency. The power
2
;
www.americanradiohistory.com
TELEVISION, Volume IV
162
START- STOP
OSCILLATOR
0
E
REACTANCE
TUBE
ice_\
¡¡
Li
SYNC-
-. TRIPLER._
GENERATOR
.r+
MICRO-
PHONE
OUTPUT
AMPL
.-
©
AMPLIFIER
FILTER
h
10
CAMERA
H
Fig.
VIDEO
AMPLIFIER
10- Transmitter
MIXER
PFM DUPLEX
SIGNAL TO
TRANSMITTER
for pulsed frequency modulation.
pulses may also be preshaped in such a manner that the plate supply
ceases in time to allow the subcarrier oscillations to decay within the
allotted duration of sound transmission.
Figure 11 illustrates the subcarrier burst without and with frequency modulation.
In Figure 11.(B), which shows a large number of frequency -modulated pulses in superposition, the first half of the wave burst is sharp
while the wave trace appears increasingly blurred toward the end. This
verifies the fact that the initial phase is substantially identical for all
bursts regardless of the frequency modulation: that is, the pulse
fronts are "coherent."
C. Pulsed frequency-modulation transmitter: Figure 10 shows a
possible arrangement of components in a pulsed frequency -modulation
(A)
subcarrier burst
(B)
subcarrier burst. frequency-modulated
Fig. 11
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
163
transmitter. The start -stop oscillator is coupled to a reactance tube (5)
which is controlled continuously by the audio signal. A low -pass filter
(10) with a cutoff at one -half of the line frequency prevents the generation of overlapping sidebands of the subcarrier that would interfere
with audio fidelity. In an experimental transmitter, the master oscillator generated about 10 cycles at a frequency of 2 megacycles during
each burst with a deviation of ± 100 kilocycles. At the output of the
doubler stage (6) the center frequency becomes 4 megacycles and the
deviation ± 200 kilocycles. The subcarrier burst is amplified and combined with the video signal at (13).
From the point of view of pulsed frequency modulation the field
synchronizing pulse and the equalizing pulses interfere with the regular sequence of horizontal synchronizing pulses. Some modification of
the standard television wave form (Figure 1) is necessary to allow the
transmission of wave bursts of constant duration. Interruptions in the
sequence result in the generation of a narrow 60 -cycle pulse that contains harmonics of 60 cycles extending throughout the audible
PULSED FREQUENCY
MODULATION
Sz
Si
!'S2
$
..-.
FIELD
1
FIELD
41-
Fig.
12- Modification
of television wave form for pulsed frequency
modulation on line-synchronizing pulses.
spectrum.
If the bursts occur during postblanking (Figure 8(B)) the field synchronizing pulse may be slotted as in Figure 6(B), but if the line synchronizing pulses are modulated by the bursts, the modification
shown in Figure 12 is desirable. Here the slots S1 isolate the sub carrier bursts from the field signal so that separation of the sound may
take place in the receiver. The slots S, act as equalizers for maintenance of interlacing. Figure 13 shows the modified television wave
form carrying pulsed frequency-modulation duplex on the line synchronizing.
D. Pulsed frequency- modulation receiver: A complete pulsed frequency- modulation receiver is shown in Figure 14. Isolation of the
frequency-modulation bursts (whether in line- synchronizing pulses as
in Figure 8(A) or in post -blanking, as in Figure 8(B)) is performed
by a selector such as a tube with two control grids. The selector is
biased off by a suitable pulse signal generated by a multivibrator which
is synchronized from the line- deflection generator or the line -synchron-
www.americanradiohistory.com
TELEVISION, Volume IV
164
Fig.
13- Combined
video signal and pulsed frequency modulation
of line- synchronizing pulses.
izing circuits of the video receiver. A limiter removes the amplitude
noise to substantially the same extent as in conventional frequency modulation systems. Demodulation of the bursts is accomplished in a
conventional discriminator circuit centered at the subcarrier frequency. All audio components above a frequency of one -half line frequency are removed by a low -pass filter as in the other duplex systems
mentioned above. A locally generated blanking signal is required for
biasing off the kinescope when the wave form of Figure 8 (B) is used.
SIGNAL -TO -NOISE RATIOS OF DUPLEX AND STANDARD SYSTEMS
Formulas for the signal -to -noise ratios of duplex and standard systems are derived in Appendix IV. A comparison of the various ratios
requires the assumption of a numerical relationship between the amplitudes of the respective carriers.
The usual practice in television installations is to establish the
amplitude S of the standard frequency-modulated sound carrier at
about 0.7 of the peak amplitude P of the picture carrier. For convenience the ratio S/P will be taken as 1/V 22. When amplitude modu-
TECTOR
ANT.
PULSE
SELECTOR
LIMITER
MULTI VIBRATOR
Fig.
14- Pulsed
DI SCRIM
INATOR
LOW PASS
FILTER
LOUD
SPEAKER
SLANR IOU
frequency -modulation receiver.
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
165
lation was standard for sound transmission prior to the adoption of
frequency modulation, the same ratio 1/V 2 was customary.
The amplitudes in duplex transmission are fixed by the amplitude
of the picture carrier.
The unmodulated amplitude h of the amplitude -modulated pulse
carrier is one half the amplitude of blanking. In the standard wave
form (Figure 1) blanking is three fourths of the peak amplitude of
the picture carrier. Hence h may be taken as 3P/8. The amplitude of
the pulsed frequency -modulation signal during post-blanking is also
3P/8. The amplitude H of width -modulated pulses is equal to the full
amplitude of blanking or 3P/4.
Column (1) of Table I lists the audio signal -to -noise ratios for the
various methods of transmission of sound in terms of P/N and other
dimensions which are associated with a particular method. P is the
amplitude of the picture carrier and N is the noise factor.
Column (2) lists the critical signal -to -noise ratios below which the
formulas are no longer valid. This limit exists in the case of width modulated pulses when the peak noise is higher than one half the pulse
amplitude. There is no limit in standard amplitude modulation and
amplitude -modulated pulses since limiting is not applied. The limit
occurs in standard frequency modulation and pulsed frequency modulation when the peak amplitudes of noise and signal are equal.
If the noise is assumed to remain constant, the signal, and therefore the signal-to -noise ratio, varies with distance from the transmitter
according to the law for the propagation of television signals. Hence
P/N may be replaced by K /d2 as shown in column (3) where d is the
distance from the transmitter and K is a proportionality constant. If
the distance d exceeds the line -of-sight distance, a somewhat higher
power of d would be appropriate.
Columns (4) and (5) show the forms taken by (2) and (3) when
values are substituted.
Figure 15 illustrates the variation of the signal -to -noise ratios with
distance from the transmitter. Comparisons made of the various
methods of sound transmission from Figure 15 are necessarily on a
relative basis since the unit of distance is d/ fK.
Standard frequency modulation with a deviation of 150 kilocycles
yields the most favorable signal -to-noise ratio within 0.044 units of
distance. An equal ratio may be obtained over a more limited distance
of 0.013 units with pulsed frequency modulation during postblanking
if the maximum deviation is of the order of 1.2 megacycles. A greater
deviation is required in pulsed frequency modulation for equality,
because the audio signal which may be recovered from a pulsed-
www.americanradiohistory.com
TELEVISION, Volume IV
166
frequency -modulation wave is proportional to the pulse width, whereas
the audio noise is proportional to the square root of the width.12 The
maximum distance from the transmitter at which limiting of a pulsed
frequency-modulation signal is effective in 'removing noise (that is, the
critical distance) is necessarily less because the noise voltage admitted
to the receiver is greater as a consequence of the greater deviation.
Pulsed frequency modulation during postblanking with the customary deviation of 150 kilocycles is intermediate between standard
1111111
Ilk
.0001
2
c
a
Il
001
Q
01
STANCE FROM TRANSMITTER
OF DISTANCE
a/`rKÌ
(UNIT
Fig. 15- Signal-to -noise ratios for sound transmission.
= amplitude -modulated pulses
= width -modulated pulses (symmetrical)
C= width -modulated pulses (dissymmetrical)
D = standard amplitude modulation
E = pulsed frequency modulation (fa = 150 kilocycles)
F = pulsed frequency modulation (fa = 1200 kilocycles)
G = standard frequency modulation (fa = 150 kilocycles)
H = standard frequency modulation (fa= 50 kilocycles;
A
B
bandwidth = 150 kilocycles).
amplitude modulation and standard frequency' modulation.
Width -modulated pulses are intermediate between amplitude -modulation pulses and standard amplitude modulation.
Amplitude -modulated pulses rank lowest, largely as a consequence
of not being susceptible to limiting.
OTHER RECEIVER CONSIDERATIONS
A duplex receiver is "no
12
better" than its sound pulse selector.
Appendix IV, equation (43).
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
167
Audible noise can be introduced into the audio system of a duplex
receiver when portions of the video signal, representing picture, are
selected along with the desired sound signal. This occurs when the
accuracy of synchronization of the selector is reduced sufficiently by
noise and interference. In this respect, the automatic frequency control of synchronization was found to be definitely superior to conventional triggered synchronization.10 The flywheel effect of the automatic frequency- control circuit tends to minimize the disturbing effect of
noise on synchronization.
It appears that with automatic -frequency-control synchronization
the major part of the total audible noise in a well- designed duplex system may be attributed to the inherent noise characteristics discussed in
Appendix IV rather than to inaccurate selection of the sound signal.
The stability of duplex circuits was not studied, but it is clear that
drifts in the values of circuit elements that affect the accuracy of sound
selection would be detrimental.
An exhaustive study of the relative costs of a television receiver
designed for duplex sound on a conventional receiver intended for
reception of standard frequency modulation was not included in the
scope of this project. However, an analysis of two experimental receivers constructed according to the arrangements in Figures 7 (A)
and 14 indicates that the cost of a commercial. duplex receiver is not
likely to exceed that of a standard receiver.
APPENDIX I
rectangular -pulse wave of unit amplitude may be expressed as
cosine series
A
e
(t) = r +
2
sin
7rr
cos wit
+
sin 27rr
cos
2Û,ct }
2
7r
sin n7r
cos tk,ct
n
0)c
a
+
}.
.
=27rfc
r = pulse width per pulse period.
Modulation of e(t) by an audio signal M(t) has the result
f(t)
-----
[1
+M(t)]e(t)
www.americanradiohistory.com
(6)
TELEVISION, Volume IV
168
/
original \
unmodulated /I
pulse wave
1
= e (t)
/audio signal
I\
+
of
and its harmonics
diminished
by r
1
1
co
rM(t)
+ 1
n -1
APPENDIX
2
-M(t)
sin nlrr
a
(7)
cosn,.,ct.
II
A well -known formula12 expressing the video bandwidth required
for equal horizontal and vertical resolution is
f=2KL2Na
(8)
in which
f = video bandwidth
L=
N
=
number of scanning lines
frame repetition rate
= aspect ratio
K = experimental factor often taken equal to 0.6.
a
Since the maximum audio frequency fa which may be transmitted by
a pulse carrier is LN /2, the combination of this formula and (8) gives
fa=
-.K, v
fvN
2Ka
(9)
Equation (9) should be regarded chiefly as an expression of proportionality between the quantities because the value of K depends upon
the criterion for equal resolutions, which is not a precise concept.
APPENDIX
III
FREQUENCY COMPONENTS RESULTING FROM SYMMETRICAL WIDTH
MODULATION OF A RECTANGULAR -PULSE CARRIER BY A SINE WAVE
The problem is the calculation of the amplitude and frequency of
each component in a rectangular -pulse carrier which is width -modulated in a symmetrical manner by a sine wave. The width of a pulse
is proportional to the amplitude of the modulating wave at the instant
corresponding to the center line of the pulse. Hence, the width ap of
13
R. D. Kell, A. V. Bedford, and M. A. Trainer, "An experimental
television system, Part II. The transmitter ", Proc. I.R.E.,
vol. 22, pp.
1246 -1266; November, 1934.
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
169
the pth pulse is
at,
=w
/
T,
cos 2p --«
\
I
(10)
1
To
in which
= width of unmodulated pulse
T, = 1 /f, = period of pulse wave
To = 1 /f0 = period of modulating
« = modulation factor.
w
wave
In the general case, the modulated carrier wave will not repeat
precisely at the end of each audio cycle, but after some time greater
than the period To there will be repetition. Let this time be called T
and the corresponding frequency, f. The equation of the modulated
pulse wave may be deduced by regarding the wave as the summation
of a large number of pulse waves of equal period T. Each component
wave will be characterized by a certain pulse width which is constant
for the particular component. Thus there is a wave starting at the
origin and characterized by a pulse width ao, a wave of width al and
phase To, a wave of width a2 and phase 2T,, etc.
The pth wave has a width ap and phase pT,. There are (f,T -1)
waves to sum. The equation of the pth wave is
Et,
A
T
sin nirat,f
2
at,
=
+a
n =1
cos 2irnf (t
n
- pT,)
.
(11)
summation over p yields the equation of the modulated pulse wave.
e'=
fcT-1
at,
p=0
T
= ea.c.
+- fcT-i
p=0
2
7r
co
sin naat,f
cos 27rnf (t
-
pTo)
n=1
(12)
e.
Since the direct -current component is not of interest, it need not be
considered further. If a certain frequency component of e is sought,
the contributions of each of the p waves must be summed in the form
www.americanradiohistory.com
TELEVISION, Volume IV
170
- feT-1
2
a
sin nrravf
cos
p=o
2irfn(t -pTc).
The amplitude of the component (fn) is V Al2
7r
B1
=
-
The remainder of the
and B1. It is expected
zero values for Al and
inserted in Al and B1.
B12
in which
cos 2irfnpT,.
(14)
sin 2irfnpT,..
(15)
p =0
2 fer
a
-I-
sin (niravf)
2 feT -1
-
=
Al
(13)
n
-i sin
(niravf)
p =0
derivation is devoted to an examination of Al
that only certain values of n will lead to nonB1. Before summation, the expression for a9 is
There results
=-2 fa-1 -n sin [2nirfe (1- « cos 2lrTcfop)
1
Al
p =0
7r
B1
2 fcT -11
=T.
p =0
-n sin
[2n7rfE
] cos
2irfnpT, (16)
(1- « cos 27rTjop) ] sin 2irfnpTc.
(17)
Expansion of A1, in terms of Bessel functions, yields
Al
=
- fa-1 -n r
2
7r
1
p =0
+ cos Rp cos
cos Rp sin
A
2 ( -1)8/218(B) cos sC9
+ Jo (B)
2(- 1) (v +l) /2Jv(B) cos vCfl l
(18)
s(even)=2
v(uud)=1
and
B1
=
- fcT-1 r sin Rp sin
2
7r
1
p=0
n,
+ sin Rp cos A
A
v(oth) =1
s(even)=2
2(
2
(- 1) 8/2J8 (B)
-1) (v +1) /24 cos vC
in which
www.americanradiohistory.com
cos sCa
(19)
TELEVISION SOUND TRANSMISSION
,rnfw =A
n7rw
A typical
cc
27rTcfc
f =B
171
=C
27rfnTc
(20)
=R.
term in Al involving the summation over s is
4( -1)8/2
fcT -1
J. sin A
cos Rp cos sCp.
(21)
p=0
7rn
The expression
fcT -1
r =o
(22)
cos Rp cos sCp
is a finite trigonometric sum which is known to have the value
1
cos [(fcT
-1) (R - sC) /2] sin
sin [ (R
2
1
+
cos
-
[fUT(R- sC) /2]
sC) /2]
[(f.T -1) (R +sC) /2] sin [fUT(R +sC) /2]
sin
2
[
(R + sC) /2]
(23)
The above sum may be abbreviated
S=
S1
+
2
If the expressions for R and
52
(24)
2
C in (20)
are introduced,
S1 in (24)
becomes
cos 7r
S1=
If
[sTjo-nT.f +n-sTfo]
n sin
7r
(nTcf
sin7r
- sTcf0)
[n- sTfo]
(25)
have a nonzero solution, the denominator must be zero at
least for some values of n. This follows from the observation that
sin A (n sTfo) is always zero. From inspection, it is seen that the
denominator of (25) is zero when
S1 is to
-
n= i- MTfe + sTfo
(26)
in which M is the positive integer. When (26) is inserted in (25), the
indeterminancy may be reduced to
www.americanradiohistory.com
TELEVISION, Volume IV
172
S1
=
1
(27)
±M
sT0fo
Similar reasoning leads to explicit forms for Al and B1 in (14) and
(15). Finally, (12), for the modulated wave, may be written in the
form
2
[
11=0
sin
{J
-
17rw
[7rw
(Mfr + vf0)
M
'
101
+
« (Mfe + vfo) ] }
-M r/2)]
cos
27r
vfo /f0
(Mfr + vfo)
(1
(28)
FREQUENCY COMPONENTS RESULTING FROM DISSYMMETRICAL WIDTH
MODULATION OF A RECTANGULAR -PULSE CARRIER BY A SINE WAVE
When each pulse of a symmetrically-modulated
translated to the right (or left) on the time axis by
to one half the width of the modulated pulse, the
unsymmetrically modulated. Therefore (12) may be
e' =
frT-]
aP
2 frT-1
ao
sin
f
n7r[G
p=0
p=0
in which
a0
n=]
a
cos 27rnf
P
I
n
=w(1 -«
A mathematical process similar to
pulse carrier is
an amount equal
carrier becomes
modified to read
t- pT,--2 J
cos 27rpT0fo)
(29)
(30)
that outlined in Appendix II yields
the result
e
=
- (-1)
1
v(odd) =1
+
J v [27rw a
1
1) (v+1)/2
v(odd)=1
2
7r
+vfo)]
(Mf0
M + vfo/fr
cos 27r [(Mfr+vfo) (t
..
co
+
(v +1)/2
Jv[27rw
M-vfo/fo
7r
(Mfe-vfo)(t-w)]
cos27r [
(_1)(p+13-2)/2
Jn[w a
E
(Mfc +
.
f0) ]Js[27re «
M
sin 27r [ (Mfr + Uo) (t
2
+
--sin
-w)]
a (Mfr-vfo)]
(7rMf0w) cos 27rf0M(t
(M
+ Ïs) ]
fo/fe
- w)
-w /2)
M7r
www.americanradiohistory.com
(31)
TELEVISION SOUND TRANSMISSION
= (p +ß),
p
(p
173
-ß), -p +ß), ( -p -ß)
(
and ß are odd positive integers.
Other symbols have the same significance previously assigned.
APPENDIX IV
SIGNAL -TO -NOISE RATIOS
1.
Standard Amplitude Modulation and Standard Frequency Modula-
tion
The root- mean -square signal -to-noise ratio for 100 per cent amplitude modulation is
S
NV
f =2v2
S
(32)
N
4
fa
in which
S = unmodulated amplitude of carrier
N
\/
fa
= highest audio frequency
fa
=
(=
peak amplitude of noise
4 X
root -mean -square noise) 14
Crosby's and others have shown that
signal
\
V
noise /I FM
6f,ß
/signal
2fa I\ noise
AM
(33)
is
of
noise
The
amplitude
in which fd is 2 times frequency deviation.
assumed to be below the threshold value.
From (32) and (33),
V
signal
11
noise /
FM
6
S
,
f
vfd 1/77
Nfa
d
(34)
fa
2.. Amplitude-Modulated Pulses
In a 100 per cent amplitude -modulated pulse system the root -meansquare value of the audio signal detected by means of a low-pass filter
14 Vernon
D. Landon, "The distribution of amplitude with time in
fluctuation noise ", Proc. I.R.E., vol. 29, pp. 50 -55; February, 1941.
15 Murray G.
Crosby, "Frequency- modulation noise characteristics ",
Proc. I.R.E., vol. 25, pp. 472 -517; April, 1937.
www.americanradiohistory.com
TELEVISION, Volume IV
174
in the receiver (see Appendix I) is
rh
(35)
in which
h
= unmodulated height of the pulse
r = ratio of pulse width to period of pulse carrier.
If the assumption is made that the pulse wave is applied to the detector
(low -pass filter) only during the time of the pulses, the root-meansquare noise is
(36)
4
Fig.
16-Width -modulated pulses.
Hence
(::l)
=2/2
z1 NI
pulses
N V fd
Width -Modulated Pulses
Before detection, the noisy signal is clipped, or limited, top and
bottom so that only a comparatively narrow section near the center of
each pulse is selected. Hence, it is assumed that noise effects are
introduced into width-modulated pulses chiefly by the random displacement of the sides of the pulses. The following derivation applies when
the peak noise is less than one half of the amplitude of the pulses. In
Figure 16, let
3.
w
= unmodulated width
of pulse
www.americanradiohistory.com
TELEVISION SOUND TRANSMISSION
H
= height
t8
= time of rise of pulse side
b
= displacement
175
of pulse
of side in seconds due to a root- mean -square noise
voltage
f = video -frequency bandwidth.
Then the slope of the side of a pulse is
(37)
b
4
t8
from which
b=
fo t8
N
4
W
(I)
-O.
i
(38)
H
W
UNMODULATED WIDTH W,
MODULATION FACTORO
T
MODULATED WIDTH WL+
MODULATION FACTOR .E 1.1
2)
Fig. 17- Decomposition of a pulse time wave into width -modulated pulses
(1) and (2) of the dissymmetrical type. Shaded lines intended to simulate
an oscillogram.
The audio signal recovered from the pulse wave, whether symmetrically
or dissymmetrically modulated, is
CwH'
(39)
.\./2
In this expression, C is a proportional constant. H' is the new height
of the pulse after the center section of the pulse has been selected out
and the remainder rejected (as shown in Figure 16). Amplitude due
to the random displacement of one side of a pulse in dissymmetrical
modulation is
CbH'.
www.americanradiohistory.com
(40)
TELEVISION, Volume IV
176
Hence the signal -to -noise ratio is
CwH'
Cb
H'=
2
22
t8N V
2
\
wH
(41)
f
In symmetrical modulation, both sides of a pulse are subject to random
displacement due to noise. The noise voltage given in (40) must therefore be multiplied by
The signal -to -noise ratio for symmetrical
modulation is therefore
2wH
.
(42)
Pulses of Frequency Modulation
The audio signal which may be recovered from a keyed frequency modulated wave by means of a discriminator followed by a low -pass
filter (Figure 14) is proportional to the pulse width. However, the
noise appearing in the audio output is proportional to the square root
of the width. Hence
4.
signal
noise
/I
P
-F-M =/r
(signal
noise
standard
www.americanradiohistory.com
F -M
(43)
METHOD OF MEASURING THE DEGREE
OF MODULATION OF A TELEVISION SIGNAL "t
A
BY
T. J. BUZALSKI
Engineering Department, National Broadcasting Company, Inc.
New York, N. Y.
Summary-A method of measuring the degree of modulation on a standard television signal is described. The double sideband output of the trans
mitter energizes a linear diode monitor, the output of which contains a
direct current component in addition to the visual signal. Means are pro
vided to interrupt this composite signal periodically by short -circuiting the
diode output load impedance for a brief interval, thus establishing a reference zero signal. The resultant modified signal, including the zero reference
level, may be observed by means of a cathode ray oscilloscope capable of
handling only alternating current signals. The trace on the face of the
oscilloscope will contain all of the information required to measure the
degree of modulation attained.
INTRODUCTION
HE need for determining the degree of modulation which was
attained on the signal radiated by a television transmitter was
apparent very soon after experiments were begun with television transmission. Most of the modulation monitoring methods which
were developed for sound broadcasting were not applicable to television broadcasting. The method of measuring the degree of modulation by observing the carrier frequency envelope on a cathode -ray
oscilloscope was applicable to television provided that the information
given by the trace was properly interpreted. The current television
standards require that the carrier envelope achieve maximum amplitude at the peak of the synchronizing signal and that this maximum
amplitude shall be independent of light and shade in the picture signal.
As a consequence of this method of operation, the peak carrier envelope
amplitude becomes a constant, whereas the average carrier envelope
amplitude becomes a variable dependent upon the content of the picture
signal. Therefore, modulation measurements under existing standards
for television transmission must be made in terms of the peak carrier
envelope amplitude, in contrast to sound broadcasting practice, wherein
such measurements would be referred to the constant which in that case
would be average carrier envelope amplitude.
When the radio frequency envelope of the visual transmitter was
*
Decimal Classification: R254.1 X R583.
Reprinted from RCA REVIEW, June, 1946.
177
www.americanradiohistory.com
178
TELEVISION, Volume IV
monitored on a cathode -ray oscilloscope, the operators were in a position
to assert with confidence that the signals being radiated were in accordance with the current standards. This method was reasonably satisfactory, but the location of the cathode -ray oscilloscope was determined by
the probable accuracy of results rather than by operating convenience.
The cathode -ray oscilloscope, a relatively expensive piece of equipment,
was made unavailable for other purposes when frequent monitoring of
the radio frequency envelope was considered necessary. A more expedient method of obtaining the information offered by the cathode -ray
oscilloscope envelope monitoring method had been sought for some time.
An article by A. W. Russell' suggested the use of a vibrating switch
to "preserve the direct current level in oscillograph amplifiers." While
the usefulness of this method in studying the operating characteristics of many vacuum tube circuits was immediately evident, its application to the measurement of modulation was not conceived until several
months had elapsed. During the course of the experimenting which followed, the switching mechanism which was used became identified as
the "Vibroswitch."
A diode rectifier, which derived its signal from the coaxial radio
frequency transmission line between the transmitter and the vestigial
sideband filter, has been used for many years as a radio monitor. The
quality of the picture was observed on a kinescope while the wave form
and amplitude of the composite signal were observed on a cathode ray
oscilloscope as a regular operating procedure. The "Vibroswitch" was
applied to the diode monitoring system.
THEORY
The circuit diagram of the diode rectifier, "Vibroswitch," and
cathode ray oscilloscope arrangement is shown in Figure 1. When the
circuit constants have been properly chosen, the instantaneous potential
difference developed across the diode load impedance Z, is substantially
proportional to the instantaneous carrier envelope amplitude. In a constant peak carrier amplitude system of modulation (direct current
transmission), which is currently standard for television, the peak
carrier amplitude is attained during the synchronizing pulse interval.
The minimum carrier amplitude occurs when a maximum white signal
is present. If the modulation were complete during a given maximum
white interval, the concurrent instantaneous carrier envelope amplitude
would be zero, and as a result the concurrent instantaneous potential
Level in Oscillograph Ampli1 A. W. Russell, "Preserving the D. C.
fiers", Electronic Eng., Vol. XV, No. 175, page 173, Sept., 1942.
www.americanradiohistory.com
MEASUREMENT OF MODULATION
179
difference across Z0 would also be zero. It, therefore, appears that if
we periodically short -circuit Ze, we will artificially create the conditions
which would obtain during complete modulation. If the rate at which
the short -circuiting occurs is sufficiently rapid, the resultant revised
signal will be passed by the cathode -ray oscilloscope amplifiers, and the
amplitude of the resultant trace should be proportional to the instantaneous potential drop across Z, and, therefore, within certain limitations, proportional to the instantaneous carrier envelope amplitude.
One limitation is imposed by the degree of linearity possible between
the voltage applied to the diode circuit and the resultant current.
Another limitation is imposed by the effective diode circuit time constant. These circuits must be so designed as to permit the rate of
CATHODE RAY
0501LO5COPE
o
-a....
//owl
¿awl
Fig. 1 -Diode rectifier, "Vibroswitch," and cathode ray oscilloscope circuit
arrangement.
change of potential difference across Z, to follow the rate of change of
carrier envelope amplitude required to transmit the desired intelligence. Further, the information being transmitted during the short circuiting interval cannot be recorded by the cathode -ray oscilloscope.
The interpretation of the results must be made in the light of these
limitations.
THE "VIBROSWITCH"
The original "Vibroswitch" was a standard vibrator such as is used
in automobile receiver power supply units, but revised for 60 cycle
alternating current operation. However, the contact spring tension
varied with use to a degree that rendered this instrument too unreliable for regular use under operating conditions. Experimentation
then proceeded through the use of a motor driven segmented disc, a
motor driven cam, a loudspeaker element equipped with contacts and,
more recently, a specially constructed switch using the coil and magnet
www.americanradiohistory.com
TELEVISION, Volume IV
180
from a Baldwin headset. The mechanical schematic diagram of this
unit is shown in Figure 2. The physical appearance is evident in
CONTACT
CTL. SCREW
DAMPING
CTL. SCREW
RUBBER
Fig.
2-Mechanical
schematic diagram of the "Vibroswitch."
Figure 3. The fundamental problem insofar as the "Vibroswitch" is
concerned is to obtain a short closed contact period with clean make
Fig.
3
-A recent
physical form of the "Vibroswitch."
and break. Most of the earlier models suffered from mechanical oscillation of the swinger, causing variation in contact resistance at the
www.americanradiohistory.com
MEASUREMENT OF MODULATION
181
instant that the contact was closed. This led to a confused trace on the
oscilloscope.
INTERPRETATION OF THE OSCILLOGRAMS
Figure 4 gives the expected oscilloscope traces. The actual appearance of the trace on an oscilloscope is shown in the photographs
/00
SYNC. PEAK
7S
DLACX
1
/6
P/N/TE
O
ZERO
(a) Horizontal deflection rate approximately one half the field repetition
rate.
/m
5YNC. PEAK
71
BLACK
/5.
WHITE
O.
(b)
-
ZERO
Horizontal deflection rate approximately one half the line repetition
rate.
Fig.
4- Representation
of the expected Oscilloscope Trace:
included in Figure 5. If the vertical deflection circuit of the monitoring oscilloscope operates with the direct current component of the
signal re-inserted, it is possible to set up a scale reading 0 to 100
on the face of the oscilloscope and using the zero carrier level indication provided by the short -circuiting interval of the "Vibroswitch"
cycle, set the gain of the oscilloscope amplifier so that the peak of sync
falls at 100 and the zero carrier dot or line falls at zero- The amplitude
www.americanradiohistory.com
TELEVISION, Volume IV
182
of the white signal and black level can then be read directly in per cent
of peak carrier envelope amplitude. Similarly, variation of black level
or peak carrier as a function of average brightness can be observed and
read in per cent of peak cal ries envelope amplitude.
GENERAL COMMENTS
The optimum repetition rate of switching would probably vary with
-
(a) Horizontal deflection rate one
half the field repetition rate
"Vibroswitch" not operating.
(b) Same horizontal deflection rate
-"Vibroswitch" operating.
-
(c) Horizontal deflection rate one
half the line repetition rate
"Vibroswitch" operating.
Fig. 5- Photographs of oscilloscope traces:
each application.
Experience with monitoring standard television
transmissions indicates that a repetition rate in the order of 800 to
1000 cycles per second is acceptable. The switching rate should be
nearly, but not exactly, in synchronism with the signal being observed.
www.americanradiohistory.com
MEASUREMENT OF MODULATION
183
The mark or short -circuiting interval should be short, perhaps on the
order of 10 per cent, but long enough so that there can be no doubt that
the circuit has been fully *discharged and that a positive mark is evident at the zero carrier level. The cathode ray oscilloscope amplifiers
must be linear over a sufficient swing to pass the composite signal
without compression.
Measurements of black level in per cent of peak carrier envelope
amplitude, white signal in per cent of peak carrier envelope amplitude,
and variation of black level as a function of average brightness using
the "Vibroswitch" technique have been checked against the envelope
cathode ray oscilloscope method. The results of the two methods were
found to be in substantial agreement.
This device permits measurements on low power equipment which
would not provide sufficient voltage to deflect the plates of an envelope
cathode ray oscilloscope directly.
ACKNOWLEDGMENT
The actual device described herein is the result of the work of many
engineers who have been associated with the author. Their contributions have been directly responsible for the processing of an "idea"
into a practical and useful tool.
www.americanradiohistory.com
FACTORS GOVERNING PERFORMANCE OF
ELECTRON GUNS IN TELEVISION
CATHODE -RAY TUBES*t
BY
R. R. LAW/
Research Laboratories, RCA Manufacturing Co., Inc.,
Harrison, N. J.
Summary-On the basis of Langmuir's1 limiting- current -density relationship, it is shown that the useful beam current in a convential television
cathode-ray tube has an upper limit defined by
Is
=
1.13
e
A2
kT
N2
i0E,-
tan24,
where
= the beam current;
= the cathode -current density;
E2.= the second -anode voltage relative to the cathode;
e = the electron charge;
k = Boltzmann's constant;
T = the absolute temperature of the cathode;
A = the aperture of the final focusing system;
N = the number of scanning lines; and,
4, = the equivalent deflection angle.
I,
io
This result is derived for the case of an ideal electron gun with no defining
apertures. In practice this upper limit of beam current is not attained because of aberrations and space -charge mutual repulsion effects.
INTRODUCTION
iN
DISCUSSIONS of the performance of electron guns in television
cathode -ray tubes, questions frequently arise as to what will be the
effect of changing this or that parameter. For example, such
questions are asked as : how does the brightness of the picture depend
upon the resolution ?; does wide -angle deflection offer other advantages
than reduction in tube length ?; or, what will be the effect of increasing
the operating voltage? Although the answers to these and many other
questions may be derived from the fundamental principles of electron
Decimal classification: R583.6 X R138.3.
from Proc. I.R.E., February,
t$ Reprinted
Now with the Research Department,
1942.
RCA Laboratories Division,
Princeton, N. J.
1 D.
B. Langmuir, "Theoretical limitations of cathode -ray tubes,"
Proc. I.R.E., vol. 25, pp. 977-991; August, 1937.
184
www.americanradiohistory.com
ELECTRON GUNS
185
optics,1 -4 there is need for a simple, easily interpreted relationship
correlating the various factors governing electron -gun performance.
It is the purpose of this paper to present such a relationship.
THEORETICAL ANALYSIS
To formulate the problem, consider a conventional' electron gun of
the form illustrated schematically in Figure 1. This device operates
in the following manner. First, the cathode -region or first- crossoverforming lens L1 concentrates the electron beam into a small diameter
at a first crossover. Second, the electrons emerging from this first
crossover are refocused to a small spot on the fluorescent screen by
the final focusing lens L2. The electron beam so formed is deflected by
electrostatic or electromagnetic means to trace out the picture raster.
What factors determine the performance of this device? The
7,
_¡.:
ß1i11::
FIRST
ATHODE
ANODE
O
ECLECTRODE
,ru0D ANODL
Fig.
ACCELERATING
ELECTRODE
1- Schematic representation of a conventional electron gun.
analysis will be facilitated by the use of suitable nomenclature and
symbols. Let
H = picture height
s=
N
scanning -spot diameter
number of scanning lines to be resolved
A
equivalent deflection angle
aperture of final focusing lens
=
¢=
O
=
= half angle of beam spread
focused electron
2 L. Jacob, "Electron distribution in electron -opticallybeams," Phil. Mag., vol. 28, pp. 81 -98; July, 1939.
Phys.,
3 E. G. Ramberg and G. A. Morton, "Electron optics," Jour. Appl.
vol. 10, pp. 465-478; July, 1939.
densities in electron beams," Jour.
4 J. R. Pierce, "Limiting current
Appl. Phys., vol. 10, pp. 715 -724; October, 1939.
6 V. K. Zworykin, "Description of an experimental television system
and kinescope," Proc. I.R.E., vol. 21, pp. 1655 -1673; December, 1933.
www.americanradiohistory.com
TELEVISION, Volume IV
186
a= first -crossover to final -focusing -lens distance
b= final- focusing -lens to screen distance
d = first -crossover diameter
D
= cathode diameter
= cathode- current density
it = first -crossover current density
E1= first -anode potential
E2 = second -anode potential
io
e= electronic
k
Fig.
,
charge
= Boltzmann's constant
2- Schematic
representation of geometric factors determining performance of the electron gun in television cathode -ray tubes.
T = absolute temperature of the cathode
The definitions of symbols having to do with the geometric configuration of the structure are further clarified in the schematic drawing of
Figure 2.
Consider the performance of this device. By definition, if the picture height is H and the number of scanning lines to be resolved is N
(the meaning of resolution will be amplified later on when the question
of light distribution across a beam trace is examined), the effective
diameter of the scanning spot may be stated as
s=H/N
(1)
This scanning spot will be an image of the first crossover. If space
charge is neglected, and the familiars electron-optical magnification
www.americanradiohistory.com
.
ELECTRON GUNS
187
formula is applied, an ideal electron gun would produce such a scanning spot from a first crossover which had an effective diameter of
H
E2
a
N
E1
b
d=-
(2)
The distances between gun and screen and between first crossover
and final focusing lens may be expressed in terms of the equivalent
deflection angle, the aperture of the final focusing lens, and the spread
of the beam as it enters the final focusing lens. If is the equivalent
deflection angle,
/
b=
H
(3)
2tan
If 9 is the half angle of beam spread and
A is
the aperture of the final
focusing lens,
a
A
=
2
(4)
tan
O
Equation (2) may then be written
d
E2 tan
=-NA
E1
95
(5)
tan 9
How much beam current may be concentrated into a crossover of
given size? What will be the light distribution across a beam trace
when the scanning spot is an image of this first crossover? In terms
of the present nomenclature, Langmuir' has shown that the current
density in a crossover in an ideal electron optical system is
11
d2.
(
sin=
i,:
= io s
i n'-
9! 14- E
1i
9
1-
; E
d2
sin2
D2
e
i
1-F-
- sin°
d"
n-
bT
D2
e
kT
-
--
d^
IF
9
siu -o
( 6)
9
J J
«
In general 1 << E (e, kT), and (d2,1D 2) sing O
1. If the value of d
given by (5) is substituted in this expression and it is remembered
www.americanradiohistory.com
TELEVISION, Volume IV
188
that sin O
tan
9
between the limits
I
=
for the angles commonly encountered, integration
0
and d/2 gives
-rrD2
e
1--
io
E kT
e
Az
N2D2
tane 0
)
(7)
4
where
7TD2i0/4
I is the current within a spot of effective size H /N. But
= I8, where I8 is the total beam current if the system contains
no limiting
apertures. Equation (7) may, therefore, be written
e
A2
kT
N^D2
tan'
ç5
(8)
This result warrants further examination. I/I8 is the ratio of the
beam current within a particular zone to the total beam current. This
zone is to be of such width as to give the resolution N. But how shall
resolution be defined? In the absence of defining apertures, the spot
has no definite boundary, and irrespective of the spacing, the scanning
lines must overlap to a certain extent. For a given resolution, how
much may they overlap? To answer these questions, it is necessary to
know the brightness distribution across a beam trace.
The author' has shown that (6) may be expressed in the form
gr
=
CIE
s
-4
d2
(9)
To determine the brightness distribution across an individual beam
trace, let x be the co- ordinate expressing distance from the center of
the spot perpendicular to the direction of scanning, and let y be the
co- ordinate
expressing distance from the center of the spot in the
direction of scanning. The excitation occasioned by a single beam
trace will be
excitation
= C2E-The
+
E-HY2
dy
= C;;
-co
E'
2
.
(10)
If the light output of the phosphor is directly proportional to the excitation, this equation represents the brightness distribution also. In
terms of a given resolution, how much may these traces overlap? Can
the degree of overlap be expressed in terms of the ratio I/I8?
6
R. R. Law, "High current electron gun for projection kinescopes,"
Proc. I.R.E., vol. 25, pp. 954-976; August, 1937.
www.americanradiohistory.com
ELECTRON GUNS
,v.
189
Figure 3 shows the resultant brightness distribution in a reproduction of a portion of a scene containing relatively large adjacent white
and dark areas. The brightness distribution is shown for three values
of the ratio I/1,. When I/I, = 0.5, the reproduction is substantially
equivalent to that which would be obtained from a cosine -squared distribution having the same maximum value and a total width equal to
twice the spacing between adjacent lines. Under these conditions the
cosine-squared distribution would give a flat field.7.8 Although the
present exponential function does not give a flat field, the practical
limiting resolution may be said to occur when the spot size is such
that the brightness of the beam trace drops to one half its maximum
value in one half the distance between the centers of adjacent scanning
lines. As may be seen from Figure 3, this condition is satisfied when
REPRODUCTION
12
Y
BRIOHTNE95 PATTERN
¢
TO
B
I
Q
BE
/
.
/,
/
2
DISTANCE (LINE -WIDTH
Fig.
3-Fidelity of reproduction as a
//
L\
.I't,'.
''
3
UN TO
function of degree of overlap.
I/I, = 0.5. But if the degree of overlap at which N lines may just be
resolved is defined by the condition I/1, = 0.5, (8) gives
1,
= 1.13i0E2
e
A2
tan'
4).
kT N2
I
therefore, represents an upper limit to the useful beam current
that may be obtained with an ideal electron gun having no defining
apertures. With suitable defining apertures the useful beam current
7 P. Mertz and F. Gray, "A theory of scanning," Bell Sys. Tech. Jour.,
vol. 13, pp. 464-515; July, 1934.
8 H. A. Wheeler and A. V. Loughren, "The fine structure of television
images," Proc. I.R.E., vol. 26, pp. 540-575; May, 1938.
www.americanradiohistory.com
190
TELEVISION, Volume IV
may be increased. This comes about because the peak current density
may be maintained over the entire spot. For example, with a circular
spot of uniform intensity overlapping to such an extent as to give a
flat field as before, the beam current may be increased by the factor
7r/1.13. Inasmuch as aberrations and space- charge mutual repulsion
effects opérate to increase spot size, these values will not be realized
in practice. For any given structure, the ratio of the measured beam
current to the computed limiting value affords a figure of merit
describing the performance of the gun. This ratio is ordinarily about
one tenth,' but by minimizing the effects of space charge in the first crossover- forming region, Pierce" has obtained current densities of
over one half the limiting value.
DISCUSSION
By virtue of this simple relationship, describing the performance
of an ideal electron gun in a television cathode -ray tube, it is now
possible to answer the original questions as to what will be the effect
of changing this or that parameter. Thus, if a linear relationship
between picture brightness and the product of beam current is
assumed, the picture brightness will vary inversely as the square of
the number of lines, and directly with the square of the voltage. Wideangle deflection does offer other advantages than reduction in tube
length provided deflection introduces no defocusing; this analysis indicates that the picture brightness should vary directly with the square
of the tangent of the equivalent deflection angle. In addition to correlating the various factors governing electron-gun performance, this
relationship may prove useful in evaluating the performance of developmental models of electron guns by affording a direct means of computing the ideal performance of the particular structure.
9 J. R. Pierce, "Rectilinear electron flow in beams," Jour.
Appl. Phys.,
vol. 11, pp. 548 -554; August, 1940.
www.americanradiohistory.com
TELEVISION RECEPTION WITH BUILT -IN
ANTENNAS FOR HORIZONTALLY AND
VERTICALLY POLARIZED WAVES
BY W. L. CARLSON:I:
RCA Manufacturing Company, Inc., Camden, N. J.
Summary -Television antennas suitable for mounting within a console
receiving cabinet are described. A small loaded dipole was found to be more
sensitive than a loop of equal size.
Data are given for reception in buildings on receivers with built -in
antennas. Reflections caused standing waves, which affected reception of
both horizontally and vertically polarized waves. The presence of people
near the receiver had the most effect on the signal strength received when
vertically polarized waves were utilized. Good reception in steel-frame
buildings was limited to the side of the building having an unobstructed
path to the transmitter. Normal obstructions in the vicinity of the antenna,
such as might be encountered in residential locations, were found to attenuate vertically polarized waves more than horizontally polarized waves.
A field survey of wave propagation through normal city obstructions
is recorded. A close agreement with theoretical open -country propagation
characteristics was obtained.
HE loop antenna enjoyed a few years of popularity in the early
days of broadcasting, but was later discarded in favor of the
better performing outdoor antenna. Recently, changed listening
habits of the public, higher -power broadcast stations, technical improvements in receivers, and other factors have contributed to the
revival of the built -in loop antenna for standard -broadcast reception.
It is natural to ask whether the future trend of the television
receiving antenna will follow the history of the standard -broadcast
antenna. It seems likely that the popularity of the built -in antenna for
standard broadcast will stimulate a demand for a built -in antenna for
television. Factors related to this question such as the propagation of
ultra- high -frequency waves through buildings and their reception on
small antennas have been recently investigated. The results obtained
are reported in this paper.
Before taking up these results, it may be well to review the work
of others which seems most pertinent to the subject. It has been shown
by Trevor and Carter', Norton, and Brown" that for outdoor reception
free from obstructions and at ultra -high frequencies such as 50 megacycles, the field strength near the ground is substantially stronger for
Decimal Classification R583.7 X R326.6.
:
f Reprinted from RCA REVIEW, April, 1942.
Now with the Research Department, RCA Laboratories Division,
Princeton, N. J.
' Trevor and Carter, "Notes on Propagation of Waves Below Ten
Meters in Length," Proc. I.R.E., March, 1933.
2
K. A. Norton, "Statement on Ultra- High -Frequency Propagation,"
Television Hearing Before FCC, Jan. 15, 1940.
191
www.americanradiohistory.com
TELEVISION, Volume IV
192
vertically polarized waves than for horizontally polarized waves. Data
are presented in the present paper which show substantially the same
relative response, as found by the above mentioned investigators, for
waves received at an outdoor location free from nearby obstructions
after having been propagated through low buildings such as are found
in a city residential district. Brown' further shows that as the receiving antenna is raised approximately 30 feet above ground, the two
types of polarization yield practically identical field intensities, when
the transmitting antenna is at least one wavelength above ground. Also
the usual radio -noise fields in the ultra- high- frequency range are
stronger in the vertical than in the horizontal plane. Therefore, in
spite of the preponderance of vertically polarized field near the surface
of the earth, horizontally polarized waves yield a more favorable signal to- noise ratio for television and aural broadcast services (between 30
and 100 megacycles) where the transmitting antenna is at least a few
wavelengths above ground level.
Wickizer' found 4.3 db higher average field strength for horizontally than for vertically polarized waves during a survey along
highways with the receiving antenna 10 feet above ground. This can
be explained by assuming that the normal obstructions encountered
along the highway attenuated vertically polarized waves more than
horizontally polarized waves. Englund, Crawford, and Mumford5
showed that trees along the roadside absorbed and reflected vertically
polarized waves. Data are presented in this present paper which indicate that trees do not materially affect horizontally polarized waves at
69 megacycles.
Jones" showed field- strength contours within a dwelling with reception of vertically polarized waves. Data are presented in this present
paper which indicate that wood frame houses interfere more with
vertically polarized waves than with horizontally polarized waves.
ANTENNA DESIGNS
A television receiving antenna, confined within a console cabinet,
may be directional with means for orienting its reception characteristics or it may be nondirectional. A vertical loop may be employed
as a bi- directional antenna for reception of vertically polarized waves
or a horizontal dipole may be employed for bi- directional reception of
3 G. H. Brown, "Vertical versus Horizontal Polarization," Electronics,
Oct., 1940.
' G. S. Wickizer, "Mobile Field Strength Recordings of 49.5, 83.5, and
142 Mc from Empire State Bldg. Horizontal and Vertical Polarization,"
RCA REVIEW, April, 1940.
5 Englund, Crawford, and Mumford, "Some Results of a Study of
Ultra -Short Wave Transmission Phenomena," Proc. I.R.E., March, 1933.
L. F. Jones, "A Study of the Propagation of Wavelengths Between
Three and Eight Meters," Proc. I.R.E., March, 1933.
R
www.americanradiohistory.com
BUILT -IN ANTENNAS
193
horizontally polarized waves. For nondirectional reception a vertical
dipole or a capacitive element terminated through a coupling inductance to chassis ground may be employed for vertically polarized waves
or a horizontal loop or folded dipole may be employed for nondirectional reception of horizontally polarized waves.
A directional built -in antenna with means for rotating it can be
employed to discriminate against interference, including undesired
Fig.
1
reflections. The nondirectional type of built -in antenna is less expensive and usually will occupy less cabinet space.
Figures 1 and 2 are photographs of two experimental types of
built -in antennas which were adapted to the RCA TRK-120 television receiver chassis. Figure 1 shows a vertical loop -type antenna. The two
turns are in parallel and are connected to an inductor through a
wave -change switch. The antenna circuit functions as a full -wave
resonant circuit and is coupled to a conventional, resonant grid circuit.
Thé circuits are designed to give a band -pass characteristic of about
5 megacycles width. Figure 2 shows a horizontal dipole with end -capaci-
www.americanradiohistory.com
194
TELEVISION, Volume IV
tance load. It connects to an inductor which couples to a resonant grid
circuit as in the case of the loop design. These antennas both have
a figure -eight reception pattern in the horizontal plane. Both
are
rotatable about a vertical axis. The loop is 10 by 141/2 inches. The
dipole ends are 81/2 inches square and are separated by 12 inches. The
same cabinet space will accommodate either antenna. Provision is
also
made through a wave -change switch for operating the sets on
conventional antennas through a transmission line.
Fig.
2
The relative sensitivity of these antennas and of a half-wave dipole
connected to the receiver through a short transmission line of negligible
loss is given below. The measurements were made at 69 Mc in an
open field with horizontal -wave polarization. The loop was in a horizontal position for this test.
Type Antenna
Half -Wave Dipole
Loaded Dipole
Loop
www.americanradiohistory.com
Relative
Sensitivity
6
3
2
BUILT-IN ANTENNAS
195
The greater sensitivity of the loaded dipole as compared with that
of the loop works out as an advantage for directional reception of
horizontally polarized waves (dipole in horizontal position) and as an
advantage for nondirectional reception of vertically polarized waves
(dipole in vertical position) .
EFFECT OF WOOD -FRAME HOUSE ON RECEPTION
survey was made comparing the reception on the two receivers
1 and 2 in a typical wood -frame dwelling. A small portable
transmitter with loop antenna was set up in four different locations:
1T, 2T, 3T, and 4T adjacent to the dwelling. See Figure 3. The two
television chassis, with loop and loaded -dipole antennas, were tested in
three different locations within the dwelling on the first floor and in
one location in the field adjacent to the house. These locations are
shown as 1R, 2R, 3R, and 4R. Maximum and minimum antenna microvolts (obtained by rotating the antennas around a vertical axis) were
recorded for both antennas and with both polarizations. It should be
noted that in these tests the dipole was always in a horizontal position
and the loop, in a vertical position. The data obtained are as follows:
A
of Figures
FIELD TEST FROM PORTABLE TRANSMITTER ON 69 Mc
7'rans.
Position
IT
IT
IT
2T
2T
2T
3T
3T
3T
4T
1T
*
Vertical Polarization
Rec.
Position
1R
2R
3R
1R
2R
3R
1R
2R
3R
4R
4R
Horizontal Polarization
Dipole
Max.
Min.
Loop
Max.
Min.
Dipole
Max.
Min.
221
125
125
161
161
68
177
87
215
125
38
113
50
13
20
51
23
38
75
125
101
161
17
40
18
40
44
45
110
55
42
62
45
16
10
75
17
75
62
189
62
57
76
88
169
26
204
225
377
155
Loop
Max. Min.
5
63
17
23
60
247
195
210
17
45
10
45
38
44
130
32
63
125
33
88
161
163
29
33
95
93
51
6
13
26
13
41
5
13
32
13
33
38
28
Average for transmitter positions 1T, 2T, 3T and receiver positions
for 100 -foot separation.
1R, 2R, 3R, corrected
The figures in the table indicate microvolts output from the receiving antennas. The transmitter loop was 21/2 feet above ground. The
receiver loop and loaded dipole were 61/4 feet above ground for all
tests. The field strength of the vertically polarized wave at receiver
position 4R from transmitter position 1T was 3.5 times the field
strength of the horizontally polarized wave. This field- strength ratio
in favor of vertical polarization is abnormally high. The ratio from a
normal distant transmitter would be approximately. as indicated in
www.americanradiohistory.com
TELEVISION, Volume IV
196
Figure 6. The loaded dipole is
at a given field strength.
1.5
times more sensitive than the loop
The average effect of the house on reception is obtained by a comparison of the readings obtained outdoors (with the transmitter in
positions 1T and 4T and the receivers in position 4R) with the readings obtained with the receivers indoors (with the transmitter in
positions 1T, 2T, and 3T and the receivers in positions 1R, 2R, and
3R). The last line of the chart contains the average for all the indoor
readings, with corrections for the difference in transmission distances
compared to the outdoor readings for positions 1T and 4R.
4R.
DWE LING
2R
3R
o0
Os_
3T
0
o
N
17
l
2T
Fig.
47
3
The individual readings for the different transmitter and receiver
positions varied widely, indicating the presence of standing waves
within the house for both wave polarizations.
For horizontally polarized waves the average indoor readings were
substantially the same as the outdoor readings.
For vertically polarized waves the loop maximum signal dropped
from 377 microvolts outdoors to an average of 155 microvolts (40 per
cent) indoors. This indicates that the polarization plane of the waves
was distorted or that the waves were attenuated. The dipole maximum
signal increased from 62 microvolts outdoors to an average of 189
microvolts indoors. This change indicates that the polarization plane
of the waves was distorted so as to have a substantial component in
the horizontal polarization plane.
The maximum voltage recorded outdoors for vertically polarized
waves on the look was 2.3 times the maximum voltage recorded for
www.americanradiohistory.com
BUILT -IN ANTENNAS
197
horizontally polarized waves on the loaded dipole. Indoors, the respective average maximum voltages were substantially equal. This suggests the possibility that rain pipes, electric wiring, and plumbing as
they are situated in wooden frame houses may adversely affect vertically-polarized waves more than horizontally-polarized waves.
During the tests it was noted that the maximum signal on the
vertical loop, for reception of horizontal waves, occurred when the Ioop
was turned broadside to the arriving wave. Mr. A. H. Turner offered
the theory that this response was due to the differences in field strength
at the top and bottom of the loop, i.e., due to the vertical voltage
gradient of the horizontally polarized wave. If correct, this theory
would'require that the response remain constant with height of loop
above ground so long as the rate of change of field strength with
height remains constant. This conclusion was verified by experiments
which appear to confirm the voltage-gradient theory for vertical loop
reception of horizontally polarized waves.
BODY EFFECT ON RECEPTION
It was observed that persons moving about in the vicinity of the
receiving antenna affected the reception, the greatest effect on the
received signal strength occurring when a vertical dipole or vertical
loop was being used. This result is to be expected since the body acts
as a vertical dipole. The body effect was further investigated as follows:
The first tests were made in the open field with the portable transmitter located at point 4T and a half -wave receiving dipole located
at point 4R of Figure 3. The receiving dipole was 6 feet, 3 inches high
at its center. A vertical dipole was used for vertically polarized wave
reception and a horizontal dipole rotated for normal maximum reception was used for horizontally polarized wave reception. A man 6 feet
tall stood on a wooden support 22 inches above ground in positions
at 10 -inch intervals in front and in back of the receiving antenna.
The recorded data are shown in Figure 4 for 69 megacycles and 45
megacycles. When the man's arms were raised parallel to his shoulders
and parallel to the dipole, the effect on horizontally polarized wave
reception was increased.
For each test the receiver gain was first adjusted to give the same
arbitrarily chosen output of 100 microamperes without the presence
of the man in the vicinity of the antenna. The new meter reading
caused by the presence of the man was then recorded. The curves are,
therefore, only an indication of the relative change in output due to
the presence of the man.
The tests were also made with the man in positions along a line at
www.americanradiohistory.com
TELEVISION, Volume IV
198
right angles to the direction of wave propagation. The variations in
signal recorded under this condition were never greater than those
indicated for positions in line with the direction of wave propagation.
A second set of tests at 69 Mc were conducted with the receiving
antenna located near the middle of the living room of the dwelling
as illustrated in Figure 3. In these tests the man stood on the floor.
The same horizontal dipole was used for horizontally polarized wave
reception. Two vertical dipoles, spaced 40 inches apart and cross connected, were used for vertically polarized-wave reception. This type of
L
EFFECT OF
13
12
II
J
10
2
MAN
111113..C.41,
NEAR ANTENNA
PClA4i2ATiCMI_
PHI NM
Ill-ralle
/
EMI
gli.1111.1.111.111
1014
9
/1
7
WA
T
90
7
G
5
DISTANCE
D
i
T
ANSMI'fT4.
TION
--
4
32
OF
MAN
H
i41 Z 7NTA l
O
FROM
II/
2
RECEIVING
3
4
ANTENNA
5
G
IN
7
FEET.
Fig. 4
antenna gives the same bidirectional reception for vertically polarized
waves as a vertical loop.
When the antennas were oriented for maximum signal strength,
the results were substantially the same indoors as outdoors. In one
test, with the antennas rotated 45 degrees from the maximum-gain
position, the response varied 2 -to-1 for vertical polarization as the
man walked across the room. For horizontal polarization the gain
varied only 10 per cent. As the antennas were oriented towards the
minimum- reception position the effect of the body became more pronounced for both polarizations.
These tests confirm the opinion that the movements of people in
the vicinity of receivers operating on frequencies of the order of 70
megacycles with a built-in antenna are more likely to interfere with
the reception from vertically polarized waves than with that from
horizontally polarized waves. The greatest effect will be observed on
the minimum response from bidirectional antennas oriented to reduce
multiple images and other interferences.
www.americanradiohistory.com
13UILT -IN ANTENNAS
199
RECEPTION IN STEEL -FRAME BUILDINGS
A number of field tests were conducted in New York City on television reception from Station W2XBS, Empire State Building, operating on the former No. 1 channel (44 to 50 megacycles). The results
obtained at three locations on the receiver with the loaded -dipole
antenna were as follows:
At 26 East Ninety -Third Street in a tenth -floor apartment, an input
of 95 microvolts was obtained in a room location which gave poor
results. At another location removed 15 feet, in the same room, an
input of 560 microvolts gave a fair -quality picture when the antenna
was oriented to reduce multiple -image responses. This location was on
the side of the apartment away from the transmitter. The distance was
3 miles from the transmitter. A better picture was obtained on an
outdoor antenna located on the roof.
At 75 Varick Street on the sixteenth floor facing the transmitter,
an input of 1550 microvolts was recorded. This signal gave an excellent
picture. Moving the receiver back towards the middle of the building
gave poor results. The distance was 2 miles from the transmitter.
At the RCA Building on the fifty-third floor facing the transmitter,
an input of 3000 microvolts was recorded. This gave a good picture.
Another location on the opposite side of the building gave an input of
150 microvolts and a very poor picture due to multiple images. The
distance to transmitter was 0.7 mile.
This survey indicates that in office buildings and apartment houses
of steel construction, dependable service using built -in antennas will
probably be found in locations facing the transmitter and preferably
within line of sight. A bidirectional antenna is desirable to reduce
multiple images.
EFFECT OF CITY OBSTRUCTIONS
The relative field strength of vertically and horizontally polarized
waves passing mainly through residential areas was also investigated.
For these tests a half-wave dipole receiving antenna was located
remote from the receiver and buildings so as to minmize the effect of
nearby obstructions.
The small test transmitter previously referred to was placed 5 feet
above ground in a residential location at Haddonfield, New Jersey.
The receiving dipole antenna was placed in three different locations in
a field, at heights ranging from 4 to 121/2 feet. At the transmitter
site the ground was about 40 feet higher in elevation than the receiving locations, most of the ground rise occurring near the transmitter.
The transmission distance was 0.6 mile. The receiving locations were
about 150 feet from each other and about the same distance from the
nearest trees and metal fences. There were eight rows of detached
www.americanradiohistory.com
TELEVISION, Volume IV
200
dwellings between transmitter and receiver. The nearest houses in line
with the propagation path were 300 feet from the receiving locations.
The data obtained are recorded in Figure 5. The dots are for
vertically polarized waves and the circles are for horizontally polarized'
waves. The solid -line curves A and B were plotted from the results
200
Vf0. ICAL
liflhliP!
140
140
Z
I
V0 A0.1 AT10
iar
180
11111111
0
1114!!!!
i
III
Amin
80
LL
60
40
20
''
0
i
,
/
-
FIELD 5UR.VEY
1
2
4
R= CEIVING
8
ANTENNA
Fig.
i
1
AT
69
MC
1
K
o
HEIGHT
IN
14
FEET
5
of theoretical calculations3, in which a ground dielectric constant of 15
and a transmitter height of 43 feet were assumed. With a transmitter
antenna height of 5 feet, the response to horizontally polarized waves
relative to vertically polarized waves would be approximately in the
ratio of Curve B' to Curve A.
Further tests were conducted with the portable transmitter located
10 feet above the roof of Building No. 5, RCA Manufacturing Company, Camden, New Jersey. The loop antenna was about 110 feet above
www.americanradiohistory.com
BUILT -IN ANTENNAS
201
ground. Figure 6 gives the field strengths recorded at the Camden
Airport, a distance of 2.5 miles. As in Figure 5, the solid curves represent the theoretical calculations. Most of the intervening buildings
along the transmission path were of brick and metal -frame construction. The terr.ain was practically level throughout the transmission
path.
200
lao
180
340
Ñ 120
z
rWz
10
c
W
LL
80
o
FIELD
20
0
2
4
RECEIVER
SUR.VEY
HEIGHT
Fig.
69
12
IO
$
G
ANTENNA
AT
IN
MC
14
FEET
6
A test run in back of the airport gave the same field strength for
horizontally polarized waves as recorded in Figure 6. The field strength
for vertically polarized waves was about the same as for horizontally
polarized waves. Around the receiving location the only obstruction
which might have caused this drop in vertically polarized signal was
a long 6 -foot high metal fence 1000 feet away from the receiver in
the direction of the transmitter. A reflected wave from some distant
object may have caused this result.
With the transmitter in the same location, another group of obser-
www.americanradiohistory.com
202
TELEVISION, Volilme IV
vations were made with the receiver located in Knight Park, Collingswood, New Jersey. On vertically polarized reception the field strength
was normal in one location which was 200 feet remote from any
obstacle, see Curve A in Figure 7. The field strength (Curve A') was
considerably reduced for the second location surrounded by trees.
These observations were made in November. There was close agreement in the recorded data for horizontally polarized waves at the two
locations. See recorded data B and B' in Figure 7. The terrain was
.
FIELD SURVEY
140
'
120
AT
69
MC
I
--
I
-
I
/Q
ICA Li
iiiiiii
1
BI
tB
AITI
..,,a'
.
'
4
2
i
I
HORIZONTAL
.
POLARIZATION
A.Ei
m
0
4
2
FZECÉIVINO
0
8
ANTENNA
Fig.
Q
E0.TICAL
POLAiL1ZAT10
10
HEIGHT
2
IN
4
FEET
7
fairly regular over the
31/2 -mile transmission path. There were numerous dwellings and miscellaneous buildings between transmitting and
receiving locations.
The close agreement between experimental data and theoretical
calculations as recorded in Figures 5, 6, and Curves B and B' in Figure 7 indicate that low buildings and other city obstructions in the
transmission path do not 'materially affect the relative field strengths
of horizontally and vertically polarized waves. The observations which
did not agree with the theoretical calculations can usually be accounted
for by objects in the vicinity of the receiving location which absorbed
and reflected vertically polarized waves more than they did horizontally
polarized waves.
The author expresses appreciation for the assistance of Dr. G. H.
Brown and Messrs. E. O. Johnson, V. D. Landon, and A. H. Turner in
connection with this investigation.
www.americanradiohistory.com
AUTOMATIC FREQUENCY AND PHASE CONTROL
OF SYNCHRONIZATION IN TELEVISION
RECEIVERS','t
BY
K. R. WENDT AND G. L. FREDENDALL
Research Department, RCA Laboratories Division,
Princeton, N. J.
Summary-One of the problems in the reception of television images is
satisfactory synchronization in the presence of noise. During the
past several years considerable experience has been gained with respect to
this problem under various receiving conditions. The system of synchronization which has given satisfactory results up to the present time has depended for its operation on the reception and separation of individual pulses.
In general, it can be said that with this system satisfactory synchronization
can be obtained from those signals which will in all other respects provide
an entirely acceptable picture. However, for limiting conditions of service,
particularly during early operation where field strength may be low, an
improvement in synchronization will be effective and desirable provided
that it does not involve other complications or disadvantages.
This paper describes a synchronizing means at the receiver that employs a new principle in the field of synchronization. The principle is automatic frequency and phase control of the saw -tooth scanning voltages. In
such a system, synchrnoziation depends on the average of many regularly
recurring synchronizing pulses. Noise has insufficient energy at the scanning
frequencies to effect control through the direct- current link from which all
but relatively long -time variations are filtered out.
Experimental receivers, in which automatic phase and frequency control
of the scanning oscillators has been incorporated, have operated with high
immunity to noise. The degree of immunity is of a different order of
magnitude from that found in conventional synchronizing systems.
Noise cannot affect horizontal resolution or interlacing. An intrinsic
property of the new system, is perfect interlacing. The return line in an
automatic -frequency -controlled system may start before synchronization.
Consideration of this new development indicates that its use would
result in several improvements in television services: (1) when severe noise
conditions occur, an improved picture is obtainable at points within the
present service area; (2) under such noise conditions the useful service area
is extended; (3) the maximum resolution permitted by a television channel
is realizable at locations having low field strengths. It is expected that these
improved results will be attained without increase in the cost of the television receiver.
to provide
*
Decimal classification R583.5.
f Presented at the Summer I.R.E. Convention,
July 1, 1942. Reprinted from Proc. I.R.E., January,
in Cleveland, Ohio, on
1943.
203
www.americanradiohistory.com
TELEVISION, Volume IV
204
CONVENTIONAL SYNCHRONIZING SYSTEMS
N THE operation of present commercial television receivers, the
natural frequencies of the horizontal and vertical scanning oscillators, in the absence of a synchronizing signal, are lower than the
line or field frequencies, respectively, at the transmitter. The application of a transmitted pulse initiates or "triggers" a new cycle of the
oscillator before one would otherwise occur. The period of the horizontal- scanning oscillator is shortened to conform to line frequency
and the period of the vertical oscillator to field frequency. Thus, triggering is required for each successive horizontal and vertical scan.
This is the basic principle of operation of conventional synchronizing
systems.
Figure 1 shows a scanning oscillator of a typical commercial television receiver. A cycle of operation in the absence of a synchronizing
signal is shown in Figure 2. As the grid potential eD of the tube T1
Fig.
1-Conventional
triggered scanning oscillator.
reaches the cutoff point as a consequence of leakage of charge through
resistance R from the previously charged capacitance C, plate current
iv begins to flow. A short time later, the induced voltage et causes the
capacitance C to take a large charge which, in turn, lowers eg to .a
high negative value. Plate current does not flow again until sufficient
current has leaked through R. The excursions of eg above the cutoff
point of the tube T2 are responsible for the generation of a saw -tooth
wave eg in the plate circuit of that tube.
Assume now that a synchronizing pulse eo is applied between point
A and ground in Figure 1. The effect is a premature rise of the potential ea to the cutoff voltage of Tl as shown in full lines in Figure 3a.
A pulse of plate current i9 occurs earlier than in the absence of a synchronizing signal. The dotted wave in Figure 3a shows the variations
of currents and voltages in the absence of a pulse as in Figure 2.
Wave e, may represent current variations in the coils of an electro-
www.americanradiohistory.com
SYNCHRONIZATION
205
magnetically deflected tube or voltage variations across the plates of an
electrostatically deflected tube.
The behavior of the triggered oscillator when noise is present in the
signal is the primary interest here. Hence, we shall wish to determine
how the frequency or phase of the scanning voltages are affected when
the picture signal is accompanied by noise peaks which exceed black
level and therefore appear in the synchronizing signal as shown in
Figure 3b. Noise peak "a" superimposed on the normal grid potential
e9 curve is insufficient to raise the potential above the cutoff of tube T1;
hence, the peak is ignored by the oscillator. Noise peak "b ", how-
Fig.
2- Operation
of the scanning oscillator.
ever, does have sufficient amplitude to cause eg to rise above the cutoff
potential, and therefore initiates a new cycle of oscillation prematurely. The legitimate synchronizing pulse at "c" would have caused
the normal cycle shown in dashed lines. The deflection signal es shown
in full lines corresponds to the premature synchronization. The dashed
lines represent the desired signal.
If es represents the horizontal- deflection signal, the observer interprets the misplacement of e8 as a line out of the normal position on the
viewing screen. If e8 represents the vertical deflection signal, he
observes a vertical movement of the picture.
It will be realized that the immunity of the system to noise is least
when eg is near cutoff because lower noise peaks are sufficient to initi-
www.americanradiohistory.com
TELF,VIII)N, VnLme IV
(a)
-_-
TELEVISION
SIGNAL
_grING LEVEL
SEPARATED
ée
SYNC. SIGNAL
,
TIME--
`L,.
/
SCANNING
_
I
GG
OSCILLATOR
GUT_OFF__
/
//
1
1
II
\""-Ii
SAWTOOTH
SIGNAL
n
c
- -- --
SIGNAL
PLUS NOISE
SEPARATED
SYNC. SIGNAL
b
c
An
..__PREMnTURC TRIPPING
ro,,
/A
DUE TO NOISE
/A
PEAK b
CUTOFF
SAWTOOTH
OSCILLATO
I
53
//
1
I
0TH
SIGNAL
/ //
I
I I
II
e,
--
II
,SCANNING
,....\- DISPLACED
LINE
.....-.-
V
(h)
Fig. 3
scanning oscillator.
(b)- Operation of scanning oscillator when noise is present.
(a)- Operation of
www.americanradiohistory.com
SYNCHRONIZATION
207
ate a new cycle of the oscillator. In the event that a synchronizing
pulse is obliterated by noise, the oscillator may remain inactive until
e9 reaches the cutoff potential of the tube and thus initiates a new
cycle which is late relative to the normal position. It is clear,, therefore,
that triggered synchronizing as described above is subject to noise
limitations that are inherent in the principle of operation.
AUTOMATIC FREQUENCY- AND PHASE -CONTROLLED SYNCHRONIZING
SYSTEMS
Figure 4 is a block diagram of the essential components of an automatic frequency- and phase -controlled synchronizing system. Since the
same principle is involved in the operation of the horizontal and verAFPC CIRCUIT
A
PHASE
D
B DETECTOR
FILTER
T-
SCANNING
VER. DEFL.
OSCILLATOR
CIRCUITS
SCANNING
I-IOR. DEFL
OSCILLATOR
CIRCUITS
r
11
-)- TELEVISION
ER.SSNC
PULSES
RECEIVER
1IR.SYNC
MO
III
PuL5E5
PHASE
B
D
DETECTOR
FILTER
F
I
AFPC CIRCUIT
Fig.
4 -Block
diagram of an automatic frequency- and
phase -controlled system.
tical circuits, it is unnecessary to specify a particular circuit. The
phase detector receives the synchronizing signal at A and a saw -tooth
wave at B taken from the output of the scanning oscillator. A control
voltage produced at D by the phase detector contains information regarding the phase of the saw -tooth wave relative to the synchronizing
pulses. The phase detectors described below respond to changes in
relative phase that may exist at the time of arrival of each pulse.
However, only the slowly varying components of the control voltage
are passed by the filter following the phase detector. Rapid variations
corresponding to rapid or erratic changes in relative phase are eliminated. Thus, the control voltage at F may be regarded as a direct
voltage which is applied to the scanning oscillator in order to restore
the phase of the oscillator relative to the synchronizing pulses when
www.americanradiohistory.com
TELEVISION, Volume IV
208
there is a long -time trend in phase away from the equilibrium state
established by the speed control of the scanning oscillator. Such
changes in phase and frequency of the saw -tooth as occur as a result
of the action of the control voltage are conducted back to the phase
detector through the feedback path in order to provide further correction.
In the presence of noise of sufficient magnitude, the phase detector
may register the relative phase of a noise peak and the corresponding
saw -tooth cycle. Such spurious components in the control voltages at
D usually lie in the range of frequencies beyond cutoff of the filter
and are therefore effectively removed from the voltage at F. The
noise immunity of automatic frequency- and phase -controlled circuits
is a consequence largely of the action of the filter. Further insight into
the theory of automatic frequency- and phase -controlled synchronization may be obtained from a more detailed account of the operation
of specific circuits.
----- r
PHASE DETECTOR
RECEIVER
FILTER
SCANNING OSCILLATOR
r---i
---1
1
I
T
I
SYNC.
SIGNAL
I
Ta
I
D
R
IC,
RrI
I
I
I
C,
Rs
I
e+11
Fig.
C
ETR:
II
_J L_ __J L__=
I
G
F
II
-I-
Twi
A=
TS
I
I
I
ye+
5-Automatic frequency- and phase -controlled
IJ
circuit.
Figure 5 shows a circuit which may be used for automatic frequency and phase control of a horizontal or a vertical oscillator. Here,
synchronizing pulses are supplied to the terminals Al -A9 of the phase
detector by means of a balanced circuit. A fraction of the output of
the scanning oscillator is introduced at point B of the phase detector
in order to form the composite signals shown in Figure 6 (a) and (b)
In practice, when the automatic frequency- and phase -controlled system
is in equilibrium, the synchronizing pulse must occur sometime during
the return line, that is, during the steep portion of the saw -tooth wave.
This restriction is necessary for the viewing of a television picture in
the correctly framed position on the screen of the cathode -ray tube.
If it is assumed that a state of equilibrium is attained, the condition is maintained in thé following manner. Tubes T1 and T2 are diode
.
www.americanradiohistory.com
SYNCHRONIZATION
209
rectifiers which may be idealized for simplicity of explanation. We
shall assume that the circuit composed of the resistance R2 and the
capacitance C2, associated with the diode T1, maintains a potential
variation (Figure 6 (c) ) at the cathode of the diode that resembles the
wave in Figure 6b in every respect except that the peak amplitudes
of the pulses are definitely located at
E0 volts. In popular terms,
-
EL-
IF SYNC
I-
POTENTIAL
OF A¿
TIME
POTENTIAL
OF AI
-E2
O
POTENTIAL =
CATHODE T
IV
'V
IV
I
I
I
-Ee
I
1
I
I
I
1
I
I
I
I
I
I
1
POTENTIAL
PLATE
TL
-T
POTENTIAL
OF
e
Fig.
6- Composite
I-1
signals for possible equilibrium position.
the diode is said to "set direct current." The values of R0 and C0 must
be chosen with the view of causing Ti to act as a direct -current setter
or peak rectifier.
In a similar manner, the diode To in combination with its associated
elements Ro and C2 maintains the potential variation at the plate of
T2 as shown in Figure 6c in which the peak amplitude of the synchron-
www.americanradiohistory.com
210
TELEVISION, Volume IV
-E
izing pulse is maintained at a potential of
volts with respect to
ground. The potential with respect to ground of the mid -point of the
resistance R3 shown in Figure 6(d) is the average of the potentials at
the end point of R3. An important observation to be made in Figure
6 (d) is that the synchronizing signal is balanced out. This leads to
the conclusion that the waveform and the direct -current component of
the signal at point D are independent of the amplitude assumed for
the synchronizing signal, but that the direct -current component is
dependent upon the phase relation of the synchronizing signal and
saw -tooth wave.
The low -pass filter in the plate circuit of the amplifier tube T3
transmits the direct -current component of the signal at point D and
greatly attenuates the alternating- current components. The amplified
direct -current component or control signal at point F is applied as a
positive bias to the grid of the scanning oscillator tube T4. The frequency of the oscillator is a function of the grid bias. Consequently,
the saw-tooth wave generated by tube T, has a frequency controlled
by the resistance R6 and the control voltage at point F. This signal is
applied to the phase detector at point B by way of the feedback path.
The capability of the circuit for controlling the frequency and phase
of the saw-tooth wave may be understood with the aid of Figure 7.
Assume, as was done above, that Figure 6 expresses a state of equilibrium in the circuit and that the frequency and phase of the saw tooth will remain indefinitely as shown if the circuit is not disturbed.
Let Figure 7(a) represent a departure from the equilibrium condition
as a result of some disturbance such as drifting in the values of circuit constants or voltages. The relative phase of the synchronizing
signal and the saw -tooth wave differs from the equilibrium phase relation (Figure 6 (a) ) by an amount AT. As before, the peak rectifiers
hold the peaks of the synchronizing signal at a potential
E0. Hence,
the alternating- current axis of the control signal at point H (Figure
7(d)) is lowered by an amount AE. The direct -current component at
point H therefore amounts to
(E0 + AE) volts or an increment of
AE volts over the equilibrium value in Figure 6(d). This increment
tends to increase the frequency of the oscillator and thus to shift the
saw -tooth wave toward the position of equilibrium. Similarly, a departure from equilibrium shown in Figure 8 (a) and (b) gives rise to an
increment of + AE volts in the control voltage which acts to decrease
the frequency of the oscillator and thus again restore the equilibrium
of the circuit.
The amount of control or hold -in power available for overcoming
-
-
-
www.americanradiohistory.com
SYNCHRONIZATION
211
phase and frequency deviations of the saw -tooth wave is proportional
to the gain of the direct -current amplifier and to the difference between
the direct -current components of the control signals at H corresponding to the two extreme phase conditions. Extreme conditions occur
when a synchronizing pulse occurs either at the maximum or minimum
points of the saw -tooth wave. This difference is equal to the amplitude
I
SYIa1C
I
I
-I
I
i
-I
E
VP'
Ez
POTENTIAL
I
I
OF Ax
TIME
Cit
O
POTE NTIAL e"
CATHODE
T1
-E0
POTE NTIAL cr
PLATE Tz
-EoCr
-Ee
Fig.
7
-Phase
of saw -tooth delayed from equilibrium position of Fig. 6.
of the saw -tooth signal introduced at point B except for negligible
voltage drops in the phase- detecting circuit.
When noise is present in the synchronizing signal, the voltage across
the terminals Ark, contains noise pulses which have been passed by
the synchronizing separators as sketched in Figure 9(a). The peak
amplitude of some noise pulses exceed the bias voltage of the diodes
and cause diode current to flow. Hence, the signal at point H resem-
www.americanradiohistory.com
TELEVISION, Volume IV
212
bles the erratic saw -tooth wave shown in Figure 9 (b) in which the
noise pulses themselves are balanced out. The filter in the plate circuit of T3 transmits only the direct -current component and the slowly
varying components of the grid voltage. All components above a few
cycles per second in frequency are effectively suppressed. The slowly
varying components represent a persistent trend in the alternating-
POTENTIAL
OF A,
POTENTIAL e
CATHODE T,
-Eo
- E.
ro-GE
Fig.
8
-Phase
--------
POTENTIAL.
OF
H
of saw -tooth advanced from equilibrium position of Fig. 6.
current axis away from the equilibrium position and give rise to a
change in the control voltage applied to the oscillator. The resulting
deviation in the phase and frequency of the oscillator is automatically
minimized by the restoring effect of the automatic frequency- and
phase- controlled circuit.
www.americanradiohistory.com
SYNCHRONIZATION
213
Almost identical constants have been used in the filters for the
horizontal and vertical circuits. The response of the filter is reduced
to about one third for a sine wave of 1 cycle per second and the
response to 60 cycles is practically zero. Therefore, individual scanning lines cannot be perceptibly displaced with respect to neighboring
lines. That is, the horizontal oscillator cannot respond to noise fast
enough to impair horizontal resolution. In general, a triggered oscillator is sensitive to noise to an extent that horizontal resolution is
decreased. Likewise the response to 30 cycles is so low that interlace
is essentially perfect, even in the presence of severe noise or an imperfect vertical -synchronizing signal.
Another form of phase detector is shown in Figure 10. This circuit uses four diodes and is inherently balanced, except for second order effects. The circuit is more complicated than the two -diode
circuit of Figure 5 but is somewhat easier to set up and adjust. The
I
Fig.
CONTROL SIGNAL
r
9- Synchronizing signal
T
M
and control signal when noise is present.
principle of operation is the same as the two -diode circuit, although the
details are different, and a description can be given more easily in
another manner. The four diodes may be considered as a single -pole
single -throw switch which connects the output capacitance C3 to the
input circuit resistance R2 during the synchronizing pulse interval.
This is accomplished as follows: the four diodes may be considered as
a bridge in which synchronizing pulses are applied in push -pull across
the diagonal AB with polarities such as to cause current conduction in
each diode. A direct -current biasing voltage, that maintains all diodes
in the nonconducting stage during the intervals between pulses, is built
up across the combination R1C1. This circuit is complete in itself ; no
battery is needed. The two corners C, D of the bridge are connected
www.americanradiohistory.com
TELEVISION, Volume IV
214
to the input and output circuit, respectively. It will be noted that each
of these corners connect within the bridge to both a cathode and an
anode. Thus, when the diodes are in a conducting state, current may
flow in either direction between the input and output circuits. Capacitance C3 in the output circuit receives a charge which brings the potential at point C nearly to the value existing at the input point D during
the synchronizing pulse interval. This voltage will, of course, depend
upon the phase of the synchronizing pulse and the saw -tooth signals.
In the circuit of Figure 10, the signal at C contains neither of
the input signals but only a direct -current component which is corrected once during each pulse in accordance with the phase relation
between the input signals. This voltage controls the output of the
SCANNING
PHASE_ DETECTOR
JI6,__
OSCILLATOR
A
Br 4-.
R,
DEFLECTION
SYNC.
SIGNAL
SAWTOOTH
C
FILTER
DE FLECTION
OUTPUT TUBE
PLATE
2
I
Fig.
10- Four -diode
ICS
ICA
5+
1
automatic frequency- and phase -controlled circuit.
direct -current amplifier which acts through the feed -back loop and
causes the phase of the saw -tooth wave to vary until an equilibrium
is reached. This equilibrium occurs on the positive slope of the saw tooth. In order that the phase relation for a properly framed picture
shall exist, the saw -tooth applied at point D must have the polarity
for which the return-line portion has a positive slope. The reason for
this requirement may be seen by tracing the operation of the circuit.
Assume that the local oscillator is out of equilibrium in such a direction that its frequency is low. The return line (positive slope) of the
saw -tooth will occur late, as shown dotted in Figure 11 and the synchronizing pulse will occur at a more negative point as shown at B,
www.americanradiohistory.com
SYNCHRONIZATION
215
rather than at A. The potential of capacitance
C3 then becomes more
negative and the plate of the amplifier tube becomes more positive
with the result that the frequency is increased and the equilibrium
restored.
The circuits shown in Figures 5 and 10 require the application of
saw -tooth signals of opposite polarity to the phase detector. Furthermore, the signal applied to the direct -current amplifier in Figure 5
contains a saw -tooth component whereas the circuit of Figure 10 does
not. Either circuit could be changed to operate the same as the other
in these two respects by reversing the input and output connections
of the phase detector. There are many possible variations in phase
detectors. However, the two described have been used satisfactorily
and serve to illustrate 'the important characteristics of this portion of
an automatic frequency- and phase -controlled system.
OUTPUT
LATE
SAwTOOTH
C.
NORMAL
A
NEGATIVE
4
SYNC
Fig.
11- Equilibrium
and off -equilibrium conditions for four -diode circuit.
In an automatic frequency- and phase -controlled system there is an
approximate equivalent to the degree of lock -in of a tripped oscillator.
The tripped oscillator is locked in tighter when the amplitude of the
synchronizing signal is increased. Greater susceptibility to noise is
noticed in the tightly locked -in oscillator. In the automatic frequency and phase -controlled system the amplitude of the synchronizing signal
is relatively unimportant but the alternating- current gain from the
phase detector to the oscillator influences the speed with which the
oscillator may be changed in frequency, that is, the tightness of lock-in.
If the automatic -frequency and phase -controlled oscillator can be
shifted rapidly, relatively few pulses are required to obtain control and
noise is averaged out over a short period. Conversely, if the gain is
low, many consecutive pulses are necessary to obtain control and
noise is averaged out over a long period, a condition which is obviously
desirable. A low alternating- current gain, unfortunately, has another
effect; i.e., when the receiver has just been turned on, or for some other
www.americanradiohistory.com
216
TELEVISION, Volume IV
reason is completely out of synchronism, the time for pulling into synchronism may be excessively long.
The mechanism of pull -in in an automatic frequency- and phase controlled system may be outlined as follows : Assume that when no signal is received, the speed -control setting is such that the frequency of
the oscillator" deviates from synchronous frequency by a small amount.
When a synchronizing signal is received, the phase detector generates
an alternating- current wave, the frequency of which is the difference
frequency of the synchronizing signal and the locally generated saw tooth wave. If the difference frequency is attenuated strongly by the
filter, there is little or no tendency for the oscillator to pull in since
no direct -current control signal is generated. If the filter is such that
some components of control voltage are passed without excessive phase
shift and attenuation, the frequency of the oscillator tends to follow
the instantaneous value of the control signal.
When the oscillator is pulled toward synchronism, the momentary
difference frequency is decreased and the oscillator tends to remain
longer in this phase than in the opposite phase during which the difference frequency is increased. This amounts to a distortion of the control signal which produces a new axis of the control voltage in the
direction to pull the oscillator toward synchronism. The oscillator may
be said to take three steps toward synchronism and two away, the
sequence continuing until the deviations become sufficiently small and
the oscillator falls into synchronism. Low phase shift through the filter and appreciable alternating-current gain are favorable for rapid
pull -in. However, since the system employs a feedback loop, care must
be taken to avoid self-oscillation. These requirements have led to the
unusual filter, shown in Figures 5 and 10.
An automatic frequency- and phase- controlled system, unlike a
triggered system, is not limited to a single phase relationship between
synchronizing and deflection. That is, the blanking bar may be caused
to occur, tightly locked in, anywhere on the screen. This means that
the deflection -return line may start at the beginning of the "front
porch, "1 ahead of the synchronizing pulse, thus greatly easing the
return -line time requirements in the deflection system. Shifting of the
blanking bar is accomplished as follows : Equilibrium, as previously
stated, occurs with the synchronizing pulse located on the return line
of the saw -tooth signal applied to the phase detector. This saw-tooth,
1
The term "front porch" refers to the part of the synchronizing signal
between the beginning of the blanking signal and the front edge of
the
synchronizing pulse.
www.americanradiohistory.com
SYNCHRONIZATION
217
however, need not be taken directly from the oscillator. The steep edge
of the original saw -tooth wave may easily be delayed and, since the
synchronizing pulse occurs in equilibrium on this delayed edge, the
blanking bar must necessarily occur after the oscillator has tripped.
An appropriate delay for horizontal deflection occurs automatically if
the saw -tooth is made by integrating the pulse on the plate of the
deflection tube, the delay being caused by the deflecting coil itself.
EXPERIMENTAL RECEIVERS
Several experimental receivers incorporating automatic frequencyand phase- controlled circuits have been constructed. These have been
tested both in the laboratory and the field and have given remarkably
Fig.
12- Automatic
-
frequency- and phase -controlled receiver
interference from high-frequency buzzer.
superior performance over conventional receivers for limiting conditions of service. Some pictures were taken in an attempt to show this
difference but obviously it is impossible to convey accurately information regarding accuracy of synchronization by means of a still photograph. Figure 12 shows a test pattern received by an automatic frequency and phase -controlled receiver when operating with a signal
above the hiss level but with interference from a high- frequency buzzer.
The interference can be seen only as short black lines, and the figure
serves mainly to show the resolution and interlace obtained essentially
in the absence of noise. Exposure for this and the other test -pattern
www.americanradiohistory.com
TELEVISION, Volume IV
218
Fig.
13- Conventional
receiver -hiss noise.
pictures was approximately 1/15 second, or 2 frames.
Figures 13 and 14 are test patterns showing the relative synchronizing capabilities of a conventional receiver and an automatic frequency- and phase -controlled receiver in the presence of about equal
amounts of hiss noise. The loss of horizontal resolution and lack of
interlace are plainly evident in Figure 13. Resolution in Figure 14 is
limited only by the modulation of the kinescope by noise. This obser-
Fig.
14- Automatic
frequency- and phase -controlled receiver -hiss noise.
www.americanradiohistory.com
SYNCHRONIZATION
15- Conventional
I
receiver -interference from electric razor.
vation was made when a noise-free driven synchronizing signal was
substituted but the picture signal left unchanged. Interlacing in Figure 14 is not perceptibly affected by noise.
The interference in Figures 15 and 16 was caused by an electric
razor. Synchronization in Figure 15 (conventional receiver) is lost
entirely during severe noise peaks. Horizontal resolution is also seriously affected. Figure 16, received on an automatic frequency- and
Fig.
16- Automatic
frequency- and phase- controlled receiver
interference from electric razor.
www.americanradiohistory.com
-
220
TELEVISION, Volume IV
phase -controlled receiver, does not exhibit the loss of resolution seen
in Figure 15.
The operating characteristics of an automatic frequency- and phase controlled receiver and of a conventional receiver are quite different.
During severe noise conditions a picture synchronized by automatic
frequency and phase control remains together as a whole but may
appear to move slightly about the equilibrium position in a random
manner. Single lines or groups of lines cannot tear out horizontally
because the filter in the horizontal circuit does not pass components of
the control signal which would cause abrupt changes in oscillator speed.
When synchronizing signals are obliterated for an appreciable length
of time, the vertical and horizontal oscillators run at the free- running
speed until synchronizing is re- established. When the receiver is properly adjusted, the free- running speeds are equal to or very close to the
synchronous speeds, a condition which favors pull -in.
Automatic frequency- and phase -controlled synchronization is not
entirely without disadvantages. For instance, reasonably good stability
must exist in the synchronizing generator at the transmitter. However, the tentative standard recommended by the Federal Communications Commission is deemed entirely adequate and present synchronizing generators are well within the recommended standard. Also, since
the system is slow to fall out of synchronism, it is likewise slow to
pull into synchronism. For instance, during a local thunderstorm, when
a multiple lightning stroke obliterates the signal for a considerable
portion of a second, the oscillators may fall out of synchronism and
require as much as a second to resynchronize.
The system is sensitive to line -voltage variations unless glow -tube
regulation is used. The regulated power required is small, however,
and the chief objection is that occasioned by the extra tubes and
sockets.
CONCLUSIONS
Superior reception resulting from the use of automatic frequencyand phase -controlled synchronizing has been experienced in field tests
under conditions of severe noise such as may exist occasionally even
within the normally useful service area of a television station. Horizontal resolution is found to be limited only by modulation of the
kinescope by noise. Noise does not destroy interlacing of scanning
lines. Tearing of the picture in horizontal strips and rapid vertical
movement which may occur in a conventional receiver during severe
noise bursts are essentially eliminated by the long time constants of an
automatic frequency- and phase-controlled system.
www.americanradiohistory.com
RADIO -FREQUENCY- OPERATED HIGH -VOLTAGE
SUPPLIES FOR CATHODE -RAY TUBES *t
BY
O. H. SCHADE$
Research and Engineering Department, RCA Manufacturing Company, Inc.,
Harrison, N. J.
Summary-The operation of tuned step -up transformers in self- excited
oscillator circuits as high-voltage sources for kinescopes is analyzed. General information and data are given for optimum radio- frequency -transformer design and operating conditions with specified rectifier loads.
Practical high-voltage supplies are illustrated ranging from 1 to 50 kilovolts
with power- output values of one-quarter watt to 50 watts, respectively.
The performance of these supplies in television equipment is discussed.
INTRODUCTION
HE operation of cathode -ray tubes for television requires highpotential direct -current sources, ranging in voltage from less
than 1 kilovolt for iconoscopes to 30 kilovolts and higher for
projection kinescopes.
The conventional high -voltage supply consists of an iron -core step up transformer energized from the 60-cycle power line, and a rectifier
circuit with smoothing filter. Mechanical and insulation problems make
it difficult to construct small 60 -cycle transformers with tightly packed
windings for voltages exceeding approximately 5 kilovolts. Practical
transformers, therefore, are relatively large and heavy and can furnish
currents considerably in excess of the usual requirements.
The use of high- frequency -power sources permits a substantial
reduction in transformer inductance and results in a relatively simple
transformer construction. The input power is generated by vacuum tube oscillators, which automatically limit the possible power output.
This characteristic and the low- energy storage in the small smoothing
reactances permit the construction of safe supplies provided the current requirements are not too high.
The theory of tuned step-up transformers points out the necessity
of constructing unusually high- impedance secondary circuits to obtain
Decimal Classification: R583 X R356.1.
the Sixteenth Annual I.R.E. Convention in New York,
N. Y., on January 10, 1941 and at the Winter Conference, New York, N. Y.,
January 28, 1943. Reprinted from Proc. I.R.E., April 1943.
$ Now with the Tube Department, RCA Victor Division, Harrison, N. J.
*
t Presented at
221
www.americanradiohistory.com
TELEVISION, Volume IV
222
efficient operation. The design of optimum high-voltage coils is, therefore, of prime importance in the construction of practical radio -fre-
quency- operated supplies.
A brief analysis of tuned step -up transformers in self-excited
oscillator circuits with rectifier loads will furnish design data for the
various circuit components and show their influence on the performance of the high -voltage supply.
THE TUNED STEP -UP TRANSFORMER
The exciting current of a transformer is determined by the reactance of the primary winding and its power factor. The power factor
is expressed at radio frequencies by its reciprocal value, the Q value
of the reactance. The loss component may be represented as a series
resistance r or a shunt resistance R (Figure 1). For Q values greater
than
5,
_
=
an+
ó
d .nanan.: Z.. QwL =R
Fig.
1
-Power factor and impedance
r=-X
Q
R = QX
Q
>
5
of tuned circuits.
and
X=o,LorX=1/o,C.
(1)
The magnetizing current of the transformer is canceled with
respect to the power source by the operation of tuning the transformer
primary. The resonant impedance, hence, of a tuned circuit is
Zo
= R.
(2)
The secondary of the transformer is tuned by the natural circuit
capacitances consisting of distributed coil capacitance, diode capacitance, and stray capacitance. The secondary circuit has, therefore, a
natural frequency 000 which determines the operating frequency of the
transformer.
A high -voltage radio -frequency transformer is a special case of
www.americanradiohistory.com
HIGH -VOLTAGE SUPPLIES
223
two coupled tuned circuits. The method of coupling is in general
immaterial; the circuit however, must be suitable for stable self excited oscillations, maintain a substantially constant secondary voltage under considerable external load variations, and load the oscillator
efficiently.
The use of critical coupling
K,=1
(3)
Q/Q//
furnishes a maximum voltage step -up for the no -load condition
E2/E1
= v Z/Z,
(4)
I.0
K= It,
0.7
=0.67.
.= _ 0o)
(m6 _
u= _.1
-
0.5
IIÀ
0.2
OR mo,x 6,0,
K»Fi=zs9-
A
K=2576
FOR 6I,>mo
t 6,02 =wo
0.I,
0.06
I.0
OS
Fig.
2- Frequency
15
characteristics of coupled circuits.
but it is not suitable for variable loads, because of its dependance on
the Q value and impedance of the secondary circuit (equations (3)
and (4) ). The maximum energy transfer into the secondary circuit is
limited to 50 per cent of the power input to the primary circuit. The
stability of the secondary voltage can be greatly improved by increasing the coupling to a value
K0 for a fully loaded circuit. The
theoretical efficiency limit then can increase to 100 per cent.
The overcoupled circuit has two coupling frequencies wi and c,.,
which cause a double -hump resonance curve as shown in Figure 2.
The spread of the peaks depends on the coupling,
K»
K=
1
-
1
+ (0)1/02)2
(01/0)2)2
www.americanradiohistory.com
(5)
TELEVISION, Volume IV
224
Ez
RL
d
TUNING CHARACTERISTIC
FOR
RS
Fig.
3- Oscillator
K»Kc
circuit for K < Ko, and unsuitable tuning
characteristic when
K,.
K»
The relative amplitude of the peaks depends on the relative frequencies
cool and wog to which the circuits are originally tuned before coupling.
It is, hence, possible to control the secondary voltage E2 by changing
the primary tuning without change of coupling or of secondary tuning.
The best voltage stability for variable loads is obtained by operation
at the lower coupling frequency, and maximum energy is obtained for
a tuning adjustment cool _. w02. The latter adjustment, however, is not
critical. It is, therefore, the desirable operating condition of the circuit. The voltage step -up is reduced to approximately one half of the
maximum obtainable in order to provide high efficiency and good
voltage regulation. The latter is in the order of 7 to 15 per cent from
no load to full load when the output is measured at the direct-current
terminals of practical kinescope supplies and includes oscillator performance. A coupling of K ? 20 K, is required dt full load.
REQUIREMENTS FOR SELF -EXCITATION, INDUCTANCE, AND Q VALUES
Self- excitation with feedback from the primary winding causes an
unstable tuning characteristics as indicated in Figure 3. A stable oscillation characteristic requires coupling of the grid- circuit inductance L2
to the secondary circuit L2 as shown in Figure 4. The circuit oscillates
at the lower frequency peak wl when the winding directions between
L1 and L3 are as in normal oscillator circuits. Reversal of L1 or L3
causes stable oscillation at w2.
-r
'
STABLE OSCILLATIONS AT
w, OR wa DEPENDING ON
PHASE OF FEEDBACK
Fig.
4- Oscillator
circuit for all values of K, and stable
Ko.
characteristic when
K»
www.americanradiohistory.com
HIGH -VOLTAGE SUPPLIES
225
The full -load Q values of primary and secondary circuits should be
high to obtain a large degree of over -coupling (K 20 K0) with moderate values of K which cannot be made very large because of insulation
requirements (K -- 25 per cent).
Desirable values are
Q1
?
10 when
transformer is shunted by the
(6)
reflected plate load Rp
QUI
?
transformer is shunted by the
equivalent rectifier load RL.
20 when
Corresponding inductance values are
(a)
(b)
+e
(c)
Fig.
5- Rectifier
circuits for kinescope supply voltages
t,Lj
0.1R
6)L2
0.05RL.
E
and
2.
(7)
The no -load Q values should of course be considerably higher than
the full -load values. A loss of 10 per cent per circuit requires ten times
the Q value given in (6) ; i.e.,
QLl
=100;
QL2
= 200.
(8)
THE EQUIVALENT RECTIFIER LOAD
The equivalent rectifier load RL depends on the rectifier circuit,
three types of which are shown in Figure 5. The alternating -current
www.americanradiohistory.com
,
22G
7'FLEVL4/ON, Voluine IV
load R,, is determined by the direct -current load resistance R, the direct current output voltage É, and the alternating peak voltage R., applied
to the rectifier tubes.
RL
=
(k2)2R
(9)
2É2
R,,
= 1 '2R
R,,
=1/8R for voltage -doubling circuits.
for half -wave rectifiers
(9a)
The direct -current load resistance R of a supply furnishing 1 milliampere at 10 kilovolts is R= 10 megohms. The secondary circuit
feeding a half-wave rectifier must, therefore, have an impedance Z0 =
R = 1ORL, i.e., 50 megohms for a secondary loss of 10 per cent. Secondary circuits of such high impedance are too expensive and large
for practical use and efficiency is, therefore, sacrificed in favor of size
and cost as shown later. Equation (9a) points out the advantage of a
doubling circuit, from an efficiency standpoint, because it requires only
one fourth the circuit impedance. The circuit of Figure 5(c) is similar
to a doubling circuit, except that Dl rectifies only part of the
coil
voltage. The voltage Ê tap is made slightly lower than the desired
focusing potential R, for electrostatic types of kinescopes. El
is
adjustable by means of the poteniometer P, which allows the addition
of B- supply voltage to the radio- frequency voltage. This circuit
has
a high efficiency, because it does not dissipate power in
a bleeder
resistance. It maintains also a very stable voltage at increased first
anode current.
THE REFLECTED PLATE LOAD
The primary-circuit constants are determined after the secondary
coil has been designed from the operating frequency, the total
power
output P.O. to be supplied by the oscillator, and the oscillator peak
voltage swing
4.
From these, the reflected load,
(Rd
Rp
2
=
2P.O.
(10)
The primary reactance wL1 is then obtained from (7). The problem
of
designing an optimum high -voltage coil and determining its operating
frequency may be approached in the following manner.
www.americanradiohistory.com
IIIGH-VOLTAGE SUPPLIES
227
HIGH -VOLTAGE COIL DESIGN
Physical Dimensions
The physical size of the coil depends on the required minimum
sparking distances and the power which must be dissipated. The latter
is at first unknown. A coil of desirable size for the particular purpose
is hence chosen and given a copper cross section consistent with voltage
requirements and high -Q values. The winding is subdivided into pies
Figure 6) with approximately 5 turns per layer, the pie spacing being
somewhat less than the pie height in order to maintain the same
potential gradient between coils as inside the winding. The coils
should be supported by strips of insulating material or by an inpregnated paper tubing, which is perforated to permit free circulation of
(
..
b=3/16
1122
MEN
113831
18221
®
®
a-
9/16'
CUM
Fig.
tom
ICC21
Ci751
6:831)
6-Dimensions of high -voltage
coil for 10 to 15 kilovolts.
air and to reduce dielectric losses. The coil size indicated in Figure
will dissipate approximately 6.5 watts in a horizontal position.
6
DESIGN FOR OPTIMUM ELECTRICAL CHARACTERISTICS
The power loss in the coil is given by
P=(E2)2/R.
The equivalent shunt -loss resistance R at resonance (see Figure
(1) ) can be written R = L /rC. Thus,
P=
(E2) 2
rC
-I.
www.americanradiohistory.com
(11)
1
and
(11a)
TELEVISION, Volume IV
228
For given values of secondary voltage E2 and tuning capacitance C,
which should be as small as possible, a minimum for the power loss
requires a high L/r ratio. At low frequencies, this ratio has a constant
value, depending only on the total copper cross section of the coil and
its shape. At higher frequencies, the coil resistance r increases because
of eddy currents, as follows:
r = ro(1 + k2)
(12)
where k is the eddy- current factor expressed as
k=
and ro
0.04N'd3
= direct -current resistance,
ohms
= constant for particular coil shape
N' = total number of insulated wire strands
d = strand diameter, inches
0.04
= effective length of coil (a + b
f = frequency, cycles per second.
l
in cross section of coil
in Figure 6), inches
It is apparent from (12) that operation at high frequencies requires
a small wire or strand diameter d. If it is desired to use Litz wire, we
may select No. 41 enamel wire as the smallest desirable wire for
strands, but are at liberty to use a single wire or parallel wires (Litz)
per turn, thus being able to vary L and f without affecting the copper
cross section or any of the remaining factors which determine k in
(12). The coil in Figure 6 contains 4200 strands of No. 41 wire in its
cross section; i.e., it may be given as 4200 turns of single No. 41 wire
or 2100 turns with 2 parallel strands of No. 41 wire, etc. The tuning
capacitance C is estimated to be C = 7 micromicrofarads (coil capacitance = 3 micromicrofarads).
The lowest operating frequency of the circuit with N = 4200 turns
(L = 387 millihenries) is 96 kilocycles, at which the eddy- current
factor k2 has still a negligible value (k2 = 0.037) . The equivalent shunt
resistance R has, therefore, the optimum value obtainable with this
coil size: R=22.5 megohms but the value of Q is only 97. Figure 7
shows the results of paralleling strands to vary L and f as explained
above. The shunt resistance R decreases to 50 per cent of its optimum
value at the frequency where Q goes through a maximum.
www.americanradiohistory.com
HIGH -VOLTAGE SUPPLIES
229
A good compromise between efficiency and voltage regulatiop, which
depends on coupling and Q values as explained, indicates 1400 turns of
3- strand Litz wire with L=43 millihenries, a resistance R =17.5
288 kilocycles.
megohms, Q = 227, and an operating frequency
time,
cost
of
wire,
etc.,
may influence
Other factors, such as winding
this choice. The maximum peak voltage for P= 6.5 watts is, hence,
15 kilovolts for this coil.
f=
E.=
TUBES AND CIRCUIT ASSEMBLY
Efficient operation of the oscillator tubes requires class C excitation
and low plate -voltage loss. Beam power tubes such as the 6L6 and
6Y6G are, therefore, especially suitable for use at low supply voltages.
DIMENSIONS, SEE FIG.6
TUNING CAPACITANCE C =7p}If
FOR COIL
TOTAL STRANDS
12
_TOTAL TURNS
N
N' =4200 Nó41 ETLWIRE
=N?OF PARALLLEL STRANDS
SPECIFIC VALUES FOR COIL
OF FIG.6
WIRE = 3/41'S LITZ
N = 1400
L=43
i
OW =265(f =500 KC)
R.0= 22.5 MEGONMS(f<100 KC)
W1
Q-
Z0-
D
J
=
28
R
=17.5
Q
=227
KC
MEG.
Qoa.
75
2
IN illinikh
a
50
25
AM
A
NUMBER OF PARALLEL STRANDS
Fig.
4
2
100
200
300
FREQUENCY
400
(f )- KILOCYCLES
500
600
7-Efficiency and Q values of coil (Fig. 6) versus number of
strands per turn for a fixed product (strands X turns).
The 6Y6G can furnish 15 watts power with 75 to 85 per cent efficiency
at voltages between 300 and 375 volts. The grid -leak bias should be
Eel = 2Ecc0. The screen -grid voltage is made self-regulating by a series
resistance R8 (Figure 8). It varies from approximately 65 volts at no
load to 120 volts at full load and, thus, aids the voltage regulation of
the supply. Larger output powers require parallel operation of tubes.
HIGH -VOLTAGE RECTIFIER TUBES
Standard high -voltage rectifiers such as the 2X2 or 2V3 -G require
considerable heater power and are not designed for high- frequency
operation. The development of special diodes for rectification of high
www.americanradiohistory.com
TELEVISION, Volume IV
230
radio -frequency voltages was therefore indicated. The RCA -8016 requires a cathode power of only one-quarter watt and thus permits
economical radio -frequency heating from the oscillator source.
SMOOTHING FILTER REQUIREMENTS
The filter capacitances have small values because of the high operating frequency (300 kilocycles for a 10- kilovolt supply). In contrast
to conditions with 60 -cycle operation, the ripple voltage is determined
substantially by the ratio of the sum of diode and stray capacitances
TICKLER
.004}if
HIGH VOLTAGE COIL
FRACTIONAL TURN
ECI
50000
.01
jir
TYPE 6Y6 -G
02 DEVELOPMENTAL
TYPE 8016
OHMS
DIODE
12
T-I
2
100
MH.
uf
0
SHIELD
1/
E2 (RIPPLE=
4.2 V. PEAK)
.0015
Rs
.00028
..UP
400
OHMS
..Ur
000
OHMS
2
.01
¡ CABLE
6001600
uuf
05
uF
E, (RIPPLE _
0.4V. PEAK)
100000 OHMS
250000
OHMS
l_ ó pf
B
B
FOCUSING CONTROL
OPERATING
Fig.
EB
IB
VOLTS
MA.
+325
+325
8- Circuit
CONDITIONS (APPROX.
E01
VOLTS
E1
E2
I2
VOLTS
KV.
KV.
MA.
31
+68
62
-36
2
10.6
0
+115
9.4
0.95
E02
and operating conditions of the 10-kilovolt
supply for kinescopes.
to the filter -condenser capacitance. The actual ripple percentages must
be of considerably lower value than in 60 -cycle filters to avoid capacitive coupling and interference with receiver operation. Typical values
are given in Figure 8 for a 10-kilovolt kinescope supply.
CIRCUIT ASSEMBLY
The particular form of the transformer assembly depends on the
type of circuit and the required sparking distances. Typical assemblies
are shown in the sketches of Figure 9. The operation at high radio frequency voltages emphasizes corona effects because of increased
www.americanradiohistory.com
HIGH-VOLTAGE SUPPLIES
DIODE
231
L,
PLATE
DIODE HEATER NUI
AND PLATE N_2
TO
CONDENSER
TAP
Li
DIODE
HEATER
N2
DIODE
HEATER
COIL ASSEMBLY FOR
NALF -WAVE DIODE CIRCUIT
Fig.
9- Typical
COIL ASSEMBLY FOR
VOLTAGE- DOUBLING CIRCUIT
high- frequency
transformer assemblies.
dielectric losses in ionized air. The fine -wire, high -potential ends of
transformer windings, must, therefore, be protected against power
loss and destructive effects due to corona by guard rings or conductors
of sufficient radius of curvature as illustrated in Figures 9 and 10. This
requirement also includes diode terminals and filter circuit.
DEVELOPMENTAL VOLTAGE SUPPLIES
A very small voltage supply for iconoscopes is shown in Figure 11.
It was built several years ago for battery operation and is housed in a
coil shield 21/2 inches in diameter. The 955 oscillator tube takes 8
milliamperes at 180 volts to supply 1 kilovolt to a bleeder circuit and
iconoscope. The operating- frequency is 1.2 megacycles. The small
diode is an experimental tube. The larger kinescope supply shown in
Figure 12 operates between 7 kilovolts and 12 kilovolts and measures
7% X 41/4 X 9 inches. The supply includes the oscillator tube, which
is separated by a heat shield from' the transformer assembly. The
housing is ventilated at the oscillator but otherwise closed, to prevent
HIGH VOLTAGE SOCKET
RESISTOR
CORONA SHIELD
POINT OF CORONA
/
1
RUBBER
INSULATED
CABLE
SHIELD
GOOD CABLE CONNECTION
Fig.
10- Corona
POOR CABLE CONNECTION
shielding of cable connections,
www.americanradiohistory.com
TELEVISION, Volume IV
232
Fig.
11
-A
1- kilovolt high -voltage
supply for iconoscopes.
dust precipitation on the high- voltage conductors. Operating data are
given on the circuit diagram in Figure 8.
A 30- kilovolt projection -tube supply with separate oscillator for
the focusing voltage is shown in Figure 13. Transformers and rectifier
assembly are housed in dust -tight shields. The outside dimensions of
the second anode supply are 11 X 11 X 12 inches high. The focusing
voltage can be varied from 4 to 7 kilovolts by tuning the primary of
its oscillator circuit. The main second -anode supply employs a voltage doubling circuit energized by three parallel 6Y6G oscillator tubes. Both
supplies are operated in series to maintain a desired voltage ratio
under varying load conditions. Circuit and performance are shown in
Figures 14 and 15.
Fig. 12-A 10-kilovolt high -voltage supply for kinescopes.
www.americanradiohistory.com
HIGH -VOLTAGE SUPPLIES
13-Arrangement
Fig.
233
of a 30- kilovolt voltage -doubling circuit.
A number of radio -frequency- operated supplies for various voltages
have given trouble -free service in the laboratory and in television
equipment. Voltage stability and focus regulation under actual operating conditions are quite satisfactory. Little difficulty was experienced
SECOND -ANODE SUPPLY
.-
yr.
FOCUSING VOLTAGE SUPPLY
50
OHMS
TYPE
50000
6Y6-G
OHMS
50 000
400
OHMS
II
50
OHMS
TYPE 8016
.01
6Y6-O
400
OHMS
yr
of
50
-i
50000
TYPE
6Y6 -G
400
OHMS
00028
yr.
OHMS
T
2 MH
Mr
2
--
yr
C
280
250
OHMS
OHMS
100000150000
RS
0
TYPE
OHMS
or
-JO,'ô 6a61,4o
pr
50
OHMS
400
OHMS
.01
IT]1
.004yC
TYPE 6Y6 -G
.05yF
IF-
C1,t600 -1600 yyr
.25
.01
yr
OHOMOSO
ÉZ
fur
25-30KV.
1--
\
SHIELDED LEAD
600 -1600
yyr
430000 OHMS
-+
02 LD3= DEVELOPMENTAL
DIODES
500000
OHMS
Ia
FOCUSING CONTROL
Es
Fig.
14- Circuit
ypfitir
0004
*
500000
OHMS
00028 pi"
E1
4 -7 KV
of the 30- kilovolt supply for projection kinescopes.
www.americanradiohistory.com
TELEVISION, Volume IV
234
preventing oscillator interference with television equipment but
isolating resistors or chokes may be required when a single source cf
filament or B supply is used.
in
CONCLUSIONS
Actual performance has proved that high -voltage supplies energized
by vacuum -tube oscillators at high frequencies are practical as kinescope and iconoscope supplies. The obtainable power output is limited
by the oscillator power. This method permits the construction of safe
supplies, where the current requirements are not too high. This lowpower reserve also protects the kinescope and rectifier in case of spark over or accidental short circuits because of the small short -circuiting
current. It must be remembered, however, that currents of dangerous
magnitude are obtainable, depending on the voltage step -up and the
-3
- - -2
40
4-
PARALLEL OSCILLATOR TUBES IN
SUPPLY
..
..
,,
375v.
30
.
M.
ÉB
Il
IW
10
0
400
360v.
\
-.
I.
I
300
2001
E.
2
Ó
<
100
3
LOAD CURRENT 72 (MA.)
Fig.
15-Regulation
characteristics of the 30- kilovolt supply.
oscillator power. For such conditions, due precautions for safety must
be taken.
The cost of high -frequency -operated supplies compares favorably
with 60 -cycle supplies when the oscillator power is moderate and when,
consequently, small oscillator tubes can be used. Kinescope supplies
for voltages up to 30 kilovolts and approximately 50 watts output are
in this range.
www.americanradiohistory.com
A
TYPE OF LIGHT VALVE FOR TELEVISION
REPRODUCTION't
BY
J. S. DONAL, JR.#
AND D. B. LANGMUIR$
Summary -The desirability of a light valve for the reproduction of
television pictures is discussed, and the use of a suspension of opaque
platelike particles for this purpose is shown to offer the particular advantages that the electron beam would be only a control mechanism and the
picture brightness would be limited only by the light source and lens system.
The theory of operation of such a suspension is described and it is
demonstrated that inertial effects may be neglected and that the rate of
orientation of the particles is independent of particle size and is a function
of the viscosity and dielectric constant of the suspending medium and of the
square of the applied voltage. The contrast ratio obtained may be made
very high, although the optical efficiency will decline as the contrast ratio
rises.
It is found that suspension resistivity must be considered in practical
application of the light valve, for if the field is applied through an insulating
wall the valve will respond only to changes in potential of the outside of
this wall, since leakage will prevent a constant wall potential from maintaining a field across the suspension.
From the results of tests, the conclusions are drawn that the fundamental optical behavior of the suspensions considered is in accordance with
the predictions of a theory based on simple assumptions, and that the suspensions fulfill the basic requirements of a television light valve.
INTRODUCTION
THE development of television has seen progress in the reproduction of images, from the early scanning disks to the modern
kinescopes, or cathode -ray tubes. Because of the limitations of
kinescopes in the production of very large images, alternative reproduction devices have been studied.' This paper presents a description
Decimal classification : R583 X R388.
Presented at the Fifteenth Annual I.R.E. Convention, in Boston,
Massachusetts, on June 29, 1940. Reprinted from Proc. I.R.E., May, 1943.
This paper reports work carried on, prior to 1940, as part of the televisiondevelopment program of the RCA Manufacturing Company, Inc.
# Research Department RCA Laboratories Division, Princeton, N. J.
$ Formerly, Research Laboratories, RCA Manufacturing Company,
Inc., Harrison, New Jersey; now the Office of Scientific Research and
Development,. Washington, D. C.
' Dr. Rosenthal has described an alternative system, "the Skiatron ",
which depends upon the development of opaque areas in microcrystalline
layers of ionic crystals under the action of electron bombardment. See A. H.
Rosenthal, "A system of large- screen television reception based on certain
electron phenomena in crystals ", Proc. I.R.E., vol. 28, pp. 203 -213; May,
*
1940.
235
www.americanradiohistory.com
236
TELEVISION, Volume IV
of a type of cathode-ray -controlled light valve which appears to offer
promise in this field.
Since a new type of television reproduction device is described, it
may be helpful to compare its principles with those of other reproduction systems. Three examples will be considered : the combination of
a scanning disk with a single light source modulated by a video signal;
a nitrobenzene Kerr cell combined with a scanning disk ; and a kinescope.
The system using the scanning disk has two major disadvantages.
The light which produces the picture must be generated by power
modulated at video frequencies. Also, light can be delivered to only
one picture element at a time and the resulting picture must suffer in
respect to brightness by a factor of many thousands relative to the
limiting brightness which can be obtained with a lantern slide where,
for example, light from every element reaches the observer's eye continuously.
The conventional Kerr cell has the theoretical advantage that the
video signal is used as a control rather than as a generator of light.
However, it suffers from the same unfavorable efficiency factor as does
the scanning disk in that light reaches the observer's eye from only
one picture element at a time.
In the kinescope, the light must be generated by modulated power,
and it can be generated at only one picture element at a time, so that
both of the disadvantages of the scanning disk are present. The success
of the kinescope in television reproduction is due chiefly to certain
special features of the device. Cathode -ray beams make it possible to
concentrate power into small areas very effectively. The combination
of such large power densities with phosphors which can efficiently
transform the power into light produces a satisfactory result in spite
of the theoretical disadvantages.
It is worth while to consider systems of television reproduction
which are free from both of the stated handicaps. An ideal device,
which may be called a television light valve, can be conceived as a
lantern slide which at every point has an opacity that can be controlled
instantaneously. Such a slide might be scanned so that the transparency varies from point to point in accordance with the picture signal.
If the transparency remained constant at the value set by the signal
until the return of the beam during the next scan, light would reach
the observer's eye continuously from all of the desired elements of the
picture. The brightness of the picture would be limited only by the
light source, the lens system, and the maximum transparency of the
light valve; the electron beam would be only a control mechanism.
www.americanradiohistory.com
LIGHT VALVE
237
In exploring the groundwork for such a hypothetical system the
variety of possibilities is large and it is difficult to limit them in any
preconceived manner. However, confining the scanning systems to
electron beams, and the controlling effects to voltages developed by the
beams, certain almost inescapable features of a practical light valve
can be defined in advance. First, the layer of optically controllable
material must be thin, of the order of a few thousandths of the width
or length. Second, the light -control effect must result from an electric
field which is parallel to the direction of light transmission. While
neither of these requirements is absolute or final, the complexities
introduced by deviating from either of them militate against the practicability of a light valve which fails to meet them.
It is illustrative to consider the conventional type of nitrobenzene
Kerr cell in the light of these conditions. Since such a cell must be at
least several millimeters thick in the direction of light propagation in
order to obtain useful transmission with the maximum electric field
permitted by the dielectric strength of nitrobenzene, and since the lines
of force must run perpendicular to this direction, neither basic requirement is fulfilled. The classical Kerr effect seems unsuited to the television light -valve problem.
In respect to frequency response the requirements for the optical
medium in the present problem are much less stringent than those for
a light valve used in conjunction with a scanning disk. Since any given
picture element need respond only once during each frame period, the
time of respose of the light valve may be 1/100 second or even longer.
The system may be considered to consist of a large nñmber (one for
each picture element) of parallel communication channels of very narrow bandwidth. Taken all together these transmit the same amount
of intelligence as a single wide -band channel.
An optically sensitive medium which satisfies the basic requirements outlined and which has been found to possess many advantageous
qualities consists of a suspension of small, flat, opaque particles in an
insulating fluid. If the suspended platelike particles have a dielectric
constant greater than that of the liquid, an electric field will cause
them to align themselves parallel to the field. The shadow cast by the
particles and, hence, also the light transmission of the suspension will
therefore be subject to control by the field.
Theoretical and experimental investigations of such suspensions
are described in this paper. A separate paper' treats the methods by
which cathode-ray beams can be made to control a light valve and
2 J.
S. Donal, Jr., "Cathode -ray control of television light valves ",
Proc. I.R.E., this issue, pp. 195 -208.
www.americanradiohistory.com
238
TELEVISION, Volume IV
describes a cathode -ray -controlled suspension light valve with which a
high- definition television picture has been projected on a screen.
THEORY
Consider a suspension in which there are n absorbing particles per
cubic centimeter, each particle having a projected area a on a plane
perpendicular to the x direction. Light traveling in the x direction
will vary in intensity L according to the familiar exponential absorption law
-=dL
naL.
dx
(1)
Integrating, we have
L
-_ena.x
(2)
Lo
Here L is the light intensity after traversing the distance x in the
suspension, while Lo is the incident light intensity at x equals zero.
The quantity nax is equal to the sum of the projected areas of all
particles in a suspension thickness x. Letting this total area equal A,
we have
L
-=e
-A.
(3)
Lo
Now let Al be the total projected area of the particles in the
unoriented condition and let A2 be their projected area after having
been aligned by the field, with L1 and L2 representing the corresponding light intensities after transmission through a suspension thickness
x. Then, from (3),
L2
=
L,
The ratio L. /L1 is equal to the maximum contrast ratio obtainable
while L., /Lo, the ratio of emergent to incident light when all particles
are lined up, may be called high -light transmission or optical efficiency
of the suspension. Designating these by C and E, respectively, and
substituting, we obtain
C- e-A2(1--(Ai/:Is))
(4)
= j¡i (1-(Ai/A2)),
(5)
www.americanradiohistory.com
LIGHT VALVE
239
This relationship between contrast, optical efficiency (or high -light
transmission L2 /4), and the shape factor Al /A2 of the individual
particles is presented graphically in Figure 1. It is seen that the contrast ratio is by no means limited to a value equal to the shape factor,
and that a contrast ratio of any desired magnitude can be obtained
simply by increasing the total area of the suspended particles per unit
area of the valve. This can be done, for example, by increasing the
thickness of the fluid layer, or by increasing the concentration of the
particles. The optical efficiency will decline as the contrast ratio rises
loo
16,
A2 =
50
20
4
6
lemeram
v 30
°
k
2
NNE
114111
1111
3
2
A
A2
I
.01
.02 .03 .05
HIGHLIGHT
0.1
.2 .3
TRANSMISSION
.5
i0
(E)
1-
Curves showing the contrast ratio obtainable from a suspension
as a function of its light transmission in the completely oriented state for
various values of the shape factor of the particles. The lines are theoretical.
Experimental points for a suspension of graphite in castor oil are
also shown.
Fig.
at a rate depending upon the shape factor of the particles.
The rate at which the particles are oriented by an applied field will
depend upon their size, shape, and moment of inertia, and upon the
fluid viscosity, dielectric constant, and density. An exact theory would
be complicated. It can be shown, however, that inertial forces in the
motion are very small compared to viscous ones, and need be considered
only when the Reynolds number3 R has a value of the order of magnitude of unity. The Reynolds number appropriate to this problem is
3 The significance of the Reynolds number in problems of this nature
is stated at length by Sir Horace Lamb in his "Hydro- dynamics ", sixth
edition, Cambridge University Press, Cambridge, England, 1932.
www.americanradiohistory.com
TELEVISION, Volume /P
240
R=
paw'
lA
where (in the centimeter -gram-second system of units)
p= fluid density
r = equivalent particle radius
w
= angular velocity of particle
p.
= coefficient of viscosity
Fig. 2 -A platelike particle is viewed edge on in an electric field. The field
component parallel to the surface induces charges on the particle and the
field component perpendicular to the surface acts upon these induced
charges to produce a torque which tends to rotate the particle so as to place
its long axis parallel to the electric field.
For the range of values of the four variables which might be practicable in a television light valve R is much less than 1, so that viscous
effects predominate overwhelmingly.
Since inertial effects may be neglected, the rate of orientation of
the particles can be calculated quite simply if conducting plates, such as
those of graphite, are considered. Such a particle is shown in an electric field in Figure 2. The plane of the flat body is shown perpendicular
to the page and its periphery may have any shape whatever.
Consider the field E resolved into components E cos 4) along OX
www.americanradiohistory.com
LIGHT VALVE
241
and E sin ¢ along OY. The former may be considered to build up
polarization charges on the plate so as to give it a double moment of
strength M. The latter may be regarded as exerting a torque upon
this dipole without affecting the value of M. If the size of the plate
(proportional to r) is varied, keeping the particle shape factor and the
electric field E constant, the surface charge density at corresponding
points will remain constant. The total polarization charges will therefore be proportional to the area of the plate and, of course, to E, the
dielectric constant of the medium. Since the separation of positive
and negative charges will be proportional to r, it is clear that
M = K1Er3E cos ¢
where K1 is a constant depending upon the shape of the periphery of
the plate.
The torque exerted on the plate by the field will be
LE= ME sin
4)
= K1E2er3 sin 0 cos ¢
= K2E2Er3 sin 2¢.
This torque will be opposed by a torque due to viscous drag, for if
the same plate as is shown in Figure 2 is considered to be rotating in
a viscous fluid, a torque will be exerted on it by the drag of the fluid.
Since the inertial forces may again be neglected, it can be shown from
dimensional reasoning that
4
Lm=K3p.-7.3
dt
where µ is the coefficient of viscosity.
The reason for the third -power variation with r can be summarized
qualitatively as follows : As r increases with constant µ and d¢ /dt, the
area upon which shearing or pressure forces are exerted increases as
r2, while the lever arms by which the forces must be multiplied to give
the torque increase as r. With constant d4 /dt the rate of shear at
corresponding points in the fluid remains constant as r increases since
the increased velocity of parts of the system is counterbalanced by the
increased scale of size of the system. Therefore, the net result is a
torque increasing as r3.
Equating these torques and rearranging, we obtain
www.americanradiohistory.com
TELEVISION, Volume IV
242
d¢
dt
e
=K-E'' sin 20.
µ
This equation can be integrated, yielding
tan 0 = tan 4)0-20
(7)
where c = K (e /p.) E2, t is the time, and 00 is the value of 0 when t = o.
Since both the torques involved increase as the cube of the particle
size, the rate of orientation is independent of the particle size. The
3-
Fig.
Theoretical curves of light transmission as a function of time t for
a suspension of particles of infinite shape factor, under a constant orienting
field. The constant e = E2 (5 /µ). At et = 0 all particles lie at 45 degrees
to the field.
time response of the suspension light valve depends therefore upon the
square of the applied voltage and upon the viscosity and dielectric constant of the suspending medium.
The projected area of the suspended particles is equal to Ao sin 4),
where Ao is the total area presented by the deoriented particles in 1
square centimeter of valve area. It is assumed that this projected area
equals zero when alignment is complete. Therefore, the transmission
coefficient T of the suspension as a function of time will be given by
T
= e-A. sin ¢
www.americanradiohistory.com
(g)
LIGHT VALVE
243
where tan 4) = tan 4)o e -2ct from (7).
The curves of Figure 3 show the relations, derived from (8),
between the transmission coefficient and the time for various initial
values of light transmission in the deoriented condition.
For comparing theory with observed data, it is convenient to eliminate 4) from (7) and (8). When this is done the equations may be
put into the form
lnln
-Ti -
lnAo =1/2 ln (1
+ Oct) .
(9)
o
X
N)
o
4
G
-I
3
4
-5
-6
-4
-3
-2
O
2
2cf
3
4
S
+a
4-
Fig.
Comparison between test and theory of light-transmission -versustime curves. The two adjustable constants used in fitting the experimental
points to the curve are discussed in the text.
4)o is arbitrary and has been set equal to 7r/4 in this case.
This equation is shown plotted in Figure 4, together with observed
points from two suspensions. The constant c, which is indeterminate
because of its dependence upon the shape factor of the particles, has
been adjusted to give the best fit. In addition the points have been
shifted horizontally by means of the constant a so as to make the curves
coincide. The latter procedure is necessary since the time scale for the
theoretical curve is essentially a relative one. An absolute origin for
the time is indeterminate because we have no way of measuring the
value of 4) when the electric field is first applied. The theory is at best
The constant
www.americanradiohistory.com
TELEVISION, Volume IV
244
approximate because of the adjustable constants and because of the
assumption (not consistent with actual suspension characteristics)
that the transmission coefficient approaches unity with complete orientation. The general behavior of the suspensions is seen, however, to
be in conformity with that predicted by simple calculations.
TESTS
Observations of suspension characteristics were made using cells
with glass walls. The thickness of the fluid layer was varied from a
few millimeters down to about 0.1 millimeter. Potentials were applied
by means of semitransparent sputtered metal coatings in direct contact
with the suspension. Light intensities were measured with a photo-
,,
hill
..NM
moo
s LARGE
PART ICLES IN OIL
NElI
,.,.
0
..
°SMALL PARTICLES IN BE NZOL
['SMALL PARTICLES IN OIL
mg
.-....
TIME REQUIRED FOR
507
ORIENTATION
111140
- SECONDS
1.0
Fig. 5 -Time required for ligtht transmission to rise to half of its final value
as a function of the parameter EV e%µ. This curve checks the square -law
response to field strength, the effect of dielectric constant and viscosity, and
the independence of particle size upon rate of orientation.
tube. When transient effects were studied this tube was connected to
the input of a direct- current amplifier which controlled an oscilloscope
having a long-persistent screen on which single traces could be studied.
In one series of tests the light transmission in the deoriented and
completely oriented states was measured for several different values of
opacity controlled by the concentration of particles or the thickness of
the cell. The results for a suspension of graphite in castor oil are
shown as the crosses in Figure 1. The theoretical relationship between
contrast and optical efficiency is seen to be verified. The effective shape
factor of the particles is apparently between 3 and 4. A contrast ratio
of 25 with an efficiency of 0.2 is seen to be obtainable.
Figure 5 is a plot of the time required for the transmission of the
www.americanradiohistory.com
LIGHT VALVE
245
suspension to reach half of its final value against the parameter EVE /p..
The parameter was varied not only by applying different field strengths
to the suspension, but by using fluids of different viscosity and dielectrict constant. Suspended particles of differing sizes were also used.
The time of orientation is seen to be proportional to the inverse square
of ENTE771, and to be independent of the particle size, as predicted by
theory.
Figure 4 showed a check of the detailed form of the transmission versus -time curves. The theoretical curve as represented by (9) is
shown by the solid line. Points measured experimentally with two different suspensions are also shown, and the agreement with theory is
satisfactory.
POLARIZATION EFFECTS
In the experiments just described the potential difference was
applied to the suspension from electrodes in contact with the fluid. The
behavior observed was not a function of the conductivity of the suspension, the various values of fluid conductivity resulting merely in the
passage of a greater or smaller conduction current. However, in important practical cases, as already described,' the potential differences may
not be applied directly to the fluid. One wall of the light valve might
consist, for example, of mica or of thin glass which is charged on its
outer face by an electron beam. In that case the conductivity of the
suspension will have a marked effect upon the performance therefore,
it seems worth while to discuss the broader aspects of this phenomenon
at this point.
Consider the behavior of a cell of which one electrode is in direct
contact with the suspension and is grounded while the other electrode
is the charged outer surface of a perfectly insulating wall. Two cases
will be discussed. In the first, the potential difference between the two
electrodes is held constant for a time long compared to the relaxation
period of the suspension, while in the second case the charged electrode
is allowed to float after bringing its potential to a certain value. In
both cases the general behavior of the suspension will be the same;
that is, when the charge is first applied a field will be established across
the valve causing it to light up, but the potential drop across the leaky
fluid will decline with time. If the particles are now deoriented the
valve will remain dark until the charge is removed from the electrodes,
at which time another flash will occur due to the depolarization field
set up when the charge on the inner wall of the valve leaks off.
The differences in detail, depending upon whether the outside of
the insulating wall is held at a constant potential during the relaxation
;
www.americanradiohistory.com
TELEVISION, Volume IV
246
period, may be analyzed quantitatively in terms of the circuit of Figure 6. The condenser C1 corresponds to the insulating wall, while the
leaky condenser C.7 represents the valve fluid.
Suppose a voltage is applied suddenly to the circuit (switch in
position 1), held constant for a time long compared to the relaxation
period, which in this case is R (C1 + C2), and is then removed suddenly
by changing the switch to position 3. The resulting changes of the
2
13
4
2
TIME
l
-
1-Nl
12
1
12
1
K
1
3I:
II
1I
II
t
li
2
V2
3'12
II
TIME--B-
II
11
II
II
v2
TIME
-;
Fig. 6-Analogue to the behavior of a leaky suspension.
(a) and (b) show the potential changes (assumed to be with respect
to ground) when a constant potential difference is applied through an
insulating wall and then removed; (c) and (d) show the potential changes
when the suspension is actuated by charging the outside of an insulating
wall which is then allowed to float until the charge is removed. The numerals
on the abscissas indicate the positions of the switch in the diagram above.
potential of the point P, with respect to ground, are represented in
Figure 6 (a). The potential differences across CZ may be represented
by the variations of the potential of the point Q with respect to ground
shown in Figure 6(b).
After the initial potential difference across C2 has fallen to zero,
the full applied voltage will be supported by the insulating wall. At
the removal of the applied potential the voltage across C2 due to
www.americanradiohistory.com
LIGHT VALVE
247
depolarization will be equal and opposite to that resulting from the
initial voltage application. This reversed voltage falls to zero by leakage and the valve returns to its original potential distribution. The
single removal of the applied potential by grounding point P completely discharges the valve provided P is kept grounded until the
reversed potential across the suspension is reduced substantially to
zero by leakage. Regrounding of point P at a later time would have
no additional effect if there is no recharging in the meantime. Also,
a subsequent application of a potential difference across C1 and C2 of
a value equal to that applied initially (as shown in the second cycles
of Figure 6 (a) and Figure 6(b)), or of a different value, would produce proportionate potential differences across the valve suspension,
uninfluenced by the potential differences of any preceding cycle.
The second of the cases mentioned above will now be considered.
If instead of holding the voltage across the cell constant, we place the
switch in position 2 after bringing the voltage to a steady value, the
potential differences across the suspension will differ, as shown in
Figure 6 (c) and Figure 6 (d), from those considered above. This case
may be realized in practice if an area of the light valve is charged by
an electron beam which then goes on to scan other valve areas. If an
amount of charge sufficient to cause an initial potential difference Vo
between P and ground is put on condenser C1 at P, the initial potential
across C2 is of course VoC1/ (C1 + C2) as before. However, this will
decay exponentially at a rate more rapid (Figure 6(d)) than in the
case considered above since C1 and C2 are no longer effectively in
parallel between point Q and ground and the time constant will now
be reduced to RC2. Instead of the entire initial potential difference
being impressed across C1 as a result of leakage through thesuspension, the potential difference across C1 will remain unchanged by the
leakage through R, and, therefore, the potential of P above ground will
fall from Vo to the asymptotic value VoC2/ (C1 + C2) as indicated in
Figure 6 (c) . If now the valve is discharged by bringing the point P
to ground potential suddenly and then allowing it to float again, a
reversed potential difference will appear across C2, (Figure 6 (d) ), but
of a magnitude smaller than that which occurred in the first direction,
since the change in potential of P by grounding it is now only VoC2,
(C1 + C2) instead of Vo, and the fraction C1/ (C1 + C2) of this appears
across C2. This potential difference is then reduced to zero by suspension leakage.
It will be remembered that in the first case considered, the changes
in potential produced by subsequent charging or discharging of the
point P were independent of the first cycle just considered because the
www.americanradiohistory.com
248'
TELEVISION, Volume IV
point P was kept grounded until the reversed potential difference
across C. had disappeared. Now, however, the situation is different,
for the potential of P drifts away from ground (Figure 6 (c) ) as the
charge on CZ associated with the reversed potential difference leaks off.
Thus, if the point P is again carried to ground potential (Figure 6(c) )
without an intermediate recharging, a second although still smaller
reversed potential difference appears across C2 (Figure 6(d)). On
repeated discharge of P to ground, these reversed potential differences
across C2 would approach zero asymptotically. Furthermore, if before
this has occurred and before the potential of P is in equilibrium at
ground potential, a new potential difference V1 (not shown in the
figure) is applied between P and ground, the change in potential of P
will be less than V, and, hence, the new potential difference developed
across C2 will be less than V1C1/ (C1 + C2).
From the practical standpoint, this phenomenon results in a delay
in the response of the light valve coming into equilibrium with the
impulses applied. This subject is discussed at greater length elsewhere.'
DISCUSSION
From the foregoing it will be seen that the suspension of opaque
platelike particles fulfills to a considerable degree the requirements of
the ideal light valve in which the opacity of a thin sheet may be varied
from point to point to reproduce the lights and shades of a picture.
The change in opacity of the suspension under the action of an
applied potential difference, which limits the contrast obtainable, may
be made as high as desired by increasing the concentration of the suspended' particles. Although the optical efficiency declines with increasing contrast, the results show that with particles of easily obtainable
shape factor and with sheets of suspension thin enough to afford the
possibility of reasonable resolutions, high contrast can be obtained
with an optical efficiency which is not unreasonably low.
Since any given point of the layer of suspension must respond only
once in each frame time, the rate of response of the suspensions investigated appears to be fully adequate for television purposes. This is
particularly true since the sheet of suspension must be thin, of the
order of a line width in thickness, if adequate resolution is to be
obtained. This requirement means that high fields will result from
relatively moderate potential differences across the suspension layer.
Since the rate of orientation was found to be independent of the
particle size, the principle limitation on particle size is that the individual particles shall not be evident in the reproduced picture. In gen-
www.americanradiohistory.com
LIGHT VALVE
249
eral this condition will be satisfied if the particles are small compared
to the size of a picture element.
Although polarization effects have been shown to result in a delay
in the attaining of an equilibrium between the response of the valve
and the potential differences applied, the effects of these residual
potential differences may be expected to be reduced in importance by
the fact that the suspension responds to the square of the applied difference in potential.
www.americanradiohistory.com
REFLECTIVE OPTICS IN PROJECTION
TELEVISION *t
BY
I. G. MALOFF# AND D. W. EPSTEIN$
Summary-Development of a process for molding large aspherical
correcting lenses from clear plastics now makes projection television techniques economical and practical for home receivers as well as theater
systems. Optical principles, mechanical mounting problems, design of
correcting lenses, molding methods and a receiver console arrangement
are presented.
TT HAS been known for a long time that aspherical surfaces in combination with either spherical or aspherical mirrors may be
arranged into optical systems of high aperture and high definition.
Astronomers made use of this principle in an arrangement consisting
of a spherical mirror and an aspherical lens; however, high costs and
difficulties in constructing such systems prevented their general utili-
zation.
In searching for efficient optical systems for projecting television
images originating on screens of cathode -ray tubes, the principle of
reflective optical systems has been made a subject of concentrated
study and experimentation. This has resulted in the development of
a number of reflective optical systems suitable for projecting television images with diagonals ranging from 25 inches to 25 feet. RCA
systems consist of a spherical front surface mirror and an aspherical
lens, positive in the central portion and gradually changing into negative near its periphery. The gain in illumination on the viewing
screen with the new systems is about six or seven to one when compared with a conventional f/2 lens. The quality of the images obtained
is comparable with images produced by conventional projection lenses.
The main handicap of the new system, the high cost of the aspherical lens, has been overcome by the development of machines for making aspherical molds and by development of a process for molding
aspherical lenses from plastics. RCA reflective optical systems are
designed for a fixed image distance and require cathode -ray tubes havDecimal Classification : R583 X R138.3.
Presented at the National Electronics Conference, Chicago, Ill., 1944.
Reprinted from Electronics, December, 1944.
# Home Instrument Department, RCA Victor Division, Camden, N. J.
$ Research Department, RCA Laboratories Division, Princeton, N. J.
*
250
www.americanradiohistory.com
PROJECTION OPTICS
251
ing face -curvatures fixed in relation to the curvature of the mirrors
in the system. The last two factors, while limiting the versatility of a
given system, appear to be a small price to pay for the manifold gain
in light. The design, manufacturing, installation and servicing of the
RCA reflective optical systems have been improved and simplified to
such a point that these systems can be considered as proven tools in
television and oscillographic techniques. Reflective systems designed
for infinite throw have been already applied successfully to television
outdoor pickup cameras with the same manifold gain in light.
ANALYSIS OF THE PROBLEM
The problem of projecting images originating on the screens of
cathode -ray tubes has received a great deal of attention from investigators here and abroad over a period "of years. It has been shown'
that the space distribution of light emitted by the screen of a cathode ray tube follows very closely the cosine or Lambert law of perfectly
diffusing surfaces. When a lens such as the conventional motion picture projection lens is used to project a cathode -ray tube image
onto a viewing screen, the overall efficiency of such a system is extremely low.
In motion -picture projection most of the light striking the film
is delivered to the viewing screen, except of course for the light
absorbed by the darkened portions of the film, thus creating the picture
itself. However, when projecting light from a perfectly diffusing surface onto a viewing screen by means of the same lens, much of the
light is lost. For large magnifications the following relation holds:
(lumens on viewing screen)
(lumens on tube)
100 %
=K
1
100%
4f2
where K is the transmission coefficient of the lens and f is the f/ number of the lens.' Good, commercially available, treated projection
lenses having a relative aperture of f/2 and a transmission coefficient
of nearly 100 percent, collect from the tube and deliver at large magnification to the viewing screen only 6.25 percent of the light generated.3
The image on the face of the cathode -ray tube is obtained at a
(1) Orth, R. T., Richards, P. A., and Headrick, L. B., Development of
Cathode -Ray Tubes for Oscillographic Purposes. Proc. I.R.E., 23, p. 1316,
Nov., 1935.
(2) Maloff, I. G., and Epstein, D. W. ELECTRON OPTICS IN
TELEVISION, McGraw -Hill Book Co., New York, N. Y., 1938.
(3) Maloff, I. G., and Tolson, W. A., A Resume of the Technical
Aspects of RCA Theatre Television, RCA REVIEW, 6, p. 6, July, 1941.
www.americanradiohistory.com
252
TELEVISION, Volume IV
relatively high cost in equipment, effort and power. Any increase in
the brightness of this image may be obtained only at great cost from
the standpoint of design and operation. For this reason, the problem
of providing a more efficient optical projection system has received a
great deal of attention. Improvement of a few percent was of no interest. A manifold increase in the percentage of light delivered to the
screen was sought. The answer was finally found in modifying a principle known to astronomers and adapting it to the problem on hand.
For quite a long time,1 5,' astronomers and opticians have known
that optical systems combining spherical and aspherical mirrors and
surfaces are capable of working at very high relative apertures and
at the same time are remarkably free from optical defects. Schmidt'
applied this principle to astrophotography. The so- called Schmidt
camera is an optical system (Figure 1) comprising a spherical mirror
A and a weak aspherical lens B at the center of curvature of the
mirror. Images of distant objects are formed on an image plate C,
which in itself is part of a sphere of radius slightly larger than half
the radius of the mirror and located at the focal point of the system.'
SYSTEM USED IN ASTRONOMY
Of the outstanding defects of the images formed by optical systems
(spherical and chromatic aberation, coma, astigmatism, curvature of
the field and distortion), only spherical aberration is distributed uniformly over the whole image field; all other defects increase with the
distance from the axis. A spherical mirror has no axis and is, of
course, achromatic. If a small aperture is placed at the center of
curvature of a spherical mirror, then any narrow beam of parallel
light coming from any direction through this aperture onto the mirror
will focus at a point located on a sphere whose radius is equal to half
the radius of curvature of the mirror. If the aperture is increased,
spherical aberration becomes apparent and the quality of the image
deteriorates.
The correcting lens in the Schmidt arrangement (shown in Figure 1) introduces into the incident beam an amount of spherical aberration which is equal to that introduced by the mirror but is opposite
in sign. Thus, by placing a suitably shaped correcting lens at the center
(4) Schwarzschild, K., "Theorie der Spiegeltelescope," Gottingen
Abhandlungen, 2, 1905.
(5) Kellner, G. A. H., U. S. Patent 969,785, granted Sept. 1910.
(6) Schmidt, Bernard, Mitt, Hamburger Sternwarte in Bergedorf, 7,
No. 36, 1932.
(7) Hendrix, D. O., and Christie, W. H., Some Applications of Schmidt
Principle in Optical Design, Sei. Amer., 161, p. 118, Aug., 1939.
www.americanradiohistory.com
PROJECTION OPTICS
253
of curvature of the mirror the non -aberration condition for all rays
arriving at the mirror from distant objects may be retained. The system is then free from the spherical aberration, while coma, astigmatism and chromatic aberration introduced by the correcting lens are
minimized by proper shaping of this lens.
SYSTEMS FOR TELEVISION PROJECTION
When a reflective optical system is used for projecting images originating on luminescent screens of cathode-ray tubes, the requirements
which the optical system must fulfill are considerably different from
those of the Schmidt camera. The most important difference is that
the light from a point on the luminescent screen does not emerge
from the optical system as a bundle of parallel light. On the contrary, it emerges as a bundle converging to a point or focus at a
definite distance. This finite throw system is radically different from
Fig.
1- Optical system
of the so- called Schmidt astronomical camera,
adapted by RCA for use in projection television systems.
that of the infinite throw. The other difference is that the thickness of the glass face plate of the cathode -ray tube introduces a certain
amount of spherical aberration, which has to be taken into account
when balancing the spherical aberration of the correcting lens against
that of the mirror.
The outstanding advantage of an optical system such as that shown
in Figure 1 over a more conventional optical system is its ability to
focus a large field (large tube diameter) with a large relative aperture.
As was mentioned already, such a system possesses this property primarily because a spherical mirror with an aperture located at the
center of curvature of the mirror suffers from only two aberrations,
spherical aberration which is uniform all over the field, and curvature of the field. This may be seen from Figures 2 and 3, where C is
the center of curvature of the mirror and 01 and O, are object points
located on the axis and off the axis respectively.
www.americanradiohistory.com
TELEVISION, Volume IV
Fig.
2-Spherical
mirror with an aperture at its center of curvature.
Figure 2 shows the ray paths for these two object points with the
aperture located at the center of curvature. It is seen that the image
or rather the circle of least confusion, since spherical aberration is
present, is practically of the same size and symmetry for both object
points. The reason for this is that the principal ray, i.e., the ray passing through the object point and center of the aperture also passes
through the center of curvature of the mirror, and is therefore also an
axis of symmetry for the sphere. The only difference is that the circular aperture mounted perpendicular to the principal axis and therefore symmetrically located with respect to the principal axis is
non -symmetrically located with respect to the auxiliary axis. This
causes some non -symmetry in the light distribution of the circle of
least confusion but this non-symmetry becomes of importance only in
the case of very large fields (large objects).
4perfure- forming
Fig.
3- Spherical
mask
mirror with an aperture that is not at the center
of curvature.
www.americanradiohistory.com
PROJECTION OPTICS
255
Figure 3 shows the imaging properties of a mirror with the aperture located not at the center of curvature. It is seen that there is
barely any sign of image formation for the off -axis object point.
PURPOSE OF CORRECTING
LAS
The object of the correcting lens is to correct for the spherical
aberration of the mirror without introducing any serious aberrations
of- itself. This is accomplished by making the correcting lens as weak
as possible and locating it in the plane of the aperture at the center of
curvature. In this way, the symmetry property of the spherical mirror
is least disturbed. The curvature of the field is not corrected as it is
actually used to good advantage in cathode -ray tube projection.
The spherical aberration of the mirror may be interpreted as focusing by means of zones, each zone having a different focal length. The
correcting lens has to be such that each zone of the lens has a different
focal length, compensating for the various focal lengths of the mirror
and resulting in a focusing system with all zones of the same focal
length.
The shape of the correcting lens will thus depend upon the zonal
focal length of the mirror one chooses as the focal length of the
optical system (mirror plus correcting lens). Since theoretically there
are an infinite number of zones on the mirror, there are theoretically
an infinite number of correcting lens shapes that will produce a system in which all zones have the same focal length.
Since the mirror with an aperture at the center of curvature has
no extra -axial or chromatic aberrations, such aberrations are caused
by the correcting lens itself, i.e., by the power or slopes on the correcting lens. From the standpoint of these aberrations, therefore, that
shape should be chosen whose maximum slope is the least. Thus if the
paraxial (central) focal length of the mirror is chosen as that of the
system, then the central focal length of the correcting lens is infinite
and the shape of the curve is concave. Alternately, if a zonal focal
length of the mirror is chosen as that of the system there will be a
zonal focal length of the correcting lens which is infinite and the
shape of the curve is convex at the center and concave past this zone.
If a peripheral focal length is chosen, the required correcting lens is
convex. The maximum slope is least for a convex -flat -concave curve.
The shape and size of the correcting lens depend upon the throw
or magnification for which the system is to be used. For a given focal
length and relative aperture, the correcting lens aperture decreases
as the magnification decreases. That this must be so, may be sur-
www.americanradiohistory.com
TELEVISION, Volume IV
256
mized from the fact that for unity magnification the plate aperture is
zero, since object and image coincide at the center of curvature.
Figure 4 shows the variation of correcting lens aperture and mirror
aperture with magnification. Thus, a different correcting lens is required for each throw or magnification. If a high relative aperture
astronomical Schmidt camera is used for projection at a throw only
a few times the focal length, the resulting image is of poor quality.
The reason is that a high relative aperture optical system can be well corrected for only one position of object and image. The throw or
magnification tolerance for a given correcting lens decreases with
increased relative aperture for a given resolution.
To obtain a flat image field, i.e., focus on a flat viewing screen, it is
necessary that the object field or tube face be curved. Calculations
show that in general the shape of tube face depends on the throw
-a
0,7
0.6
Semiaper/ure of mirror
'for ,axid/ point
0.5
t 0.4
0.3
E
0.2
0.1
l
/
ÇSemiaper/ure
of
correcting /ens
0
10
100
500
Magnification
Fig.
4- Manner
in which the semi -apertures of the mirror and correcting
lens vary with magnification.
sphere for infinite throw and an ellipsoid for finite throw. The eccentricity of the ellipsoid is sufficiently small, however, so that even for
finite throw the tube face may be made spherical with a radius of
curvature equal to that of the focal length of the system.
DESIGN OF CORRECTING LENS
The shape of the correcting lens must be such that all rays emanating from an object point O, and reflected by the mirror, shall meet
at the image point I located at a distance S from the correcting lens.
Figure 5 shows three rays emanating from O and striking the mirror
at different apertures. Without the presence of the correcting lens,
rays 1, 2, 3 would intersect the axis at distances ql, q2, and q3 from the
center of curvature. The slopes on the correcting lens have to be such
(approximately as shown on Figure 5) that all three rays intersect at
www.americanradiohistory.com
PROJECTION OPTICS
257
I; hence, the correcting lens has
a flat zone at the point where ray 2
passes, negative slope where ray 1 passes and positive slope where ray
3 passes.
Considered from the point of view of spherical aberration, if the
zone where ray 2 strikes the mirror is taken as a reference, then the
mirror has negative spherical aberration for smaller apertures and
thus requires a positive lens for correction, and positive spherical
aberration for larger apertures and thus requires a negative lens.
The shape of the curve of the correcting lens for any throw may be
calculated to about the same accuracy as that obtained with the equation given by Hendrix and Christie for infinite throw, from the
formula
Fig.
5- Diagram
illustrating how a suitably designed correcting lens makes
the required corrections for spherical aberration.
d=
1 4(1p
N1
p
l2x'-2I 2- s -p/
x21
(1)
x2
(2)
or from its equivalent
d=
1
N-1
m,-S
1
C
4(
1
x4
S
/
2(
2S-m1
S
J
where d is the depth of the curve at the zone of radius x, p is the distance between object (tube face) and center of curvature of the mirror,
www.americanradiohistory.com
258
TELEVISION, Volume IV
S is the distance between image (viewing screen) and center of curvature of the mirror, and in is the magnification. The relation between
the quantities p, S, na and the focal length f of the system is given by
= mp
p=f (m-1)/m
(3)
S=f
(5)
S
(m--1)
(4)
All distances in the above equations are measured in terms of the
radius of curvature of the mirror, i.e., the radius of curvature is taken
as the unit of length.
In applications such as projection television, the light emitted by
the luminescent screen first passes through a thickness of glass constituting the tube face. Although the effect of the tube face is small in
cases of high f/number, it becomes quite appreciable for a low f /number system. The fact that the tube face is curved endows it with some
power and actually alters the magnification of the system slightly.
However, the largest effect of the tube face is caused by its spherical
aberration. The presence of the tube face necessitates a change in the
shape of the correcting lens. For a convex -concave correcting lens, the
spherical aberration of the tube face calls for greater correction from
the convex portion and smaller correction from the concave portion.
Equations (1) and (2) are not sufficiently accurate to determine
the shape of the correcting lens for systems with high relative apertures. It was found that the best method of determining accurately
the shape of the correcting lens is the old reliable and rather tedious,
but very accurate method of tracing rays through the system consisting of the tube face, mirror and correcting lens.
PROJECTION EFFICIENCY
The projection efficiency of any optical system will be defined as
the percentage of the total light flux, in lumens, emitted in a forward
direction by an axial element of a perfectly diffusing source, such as
a luminescent screen of a cathode-ray tube, which the optical system
accepts and focuses on the corresponding image element, assuming 100
percent mirror reflection and 100 percent lens transmission.
The efficiency, e, in percent as defined above is given by
e =100 sin' U, where U is the semi -apex angle shown on Figure 6.
Hence, to determine the efficiency of a lens for a perfectly diffusing
source, it is merely necessary to know the angle that the lens, or
entrance pupil, subtends at the source. As may be seen from Figure 6,
:
www.americanradiohistory.com
Fig.
PROJECTION OPTICS
259
6- Efficiency of a simple lens decreases as the
lens is moved away from
its source to decrease the magnification.
the farther a given lens is from a source, i.e., the small the magnification, the lower is the efficiency of the lens.
It has become customary to rate a lens by its f/number for infinite
magnification, i.e., object located at the focal point of the lens. The
f/number is defined as
f /number
= 0.5 sin
U
= 0.5
V e,
where eb is the efficiency (a fraction, not percent) for infinite magnification. The smallest f/number possible is 0.5, since at 0.5 the efficiency is unity and all the light emitted by the object element in a
forward direction is concentrated in the image element. Figure 7
shows the efficiency e, of a lens as a function of f /number. It is seen
that the efficiency of most lenses is very low.
As already mentioned, the efficiency of a given lens decreases when
Fig.
7-Variation
of lens efficiency with its f /number
www.americanradiohistory.com
260
TELEVISION, Volume IV
the magnification or throw decreases. This factor becomes of importance in the case of home projection, where magnifications as low as 5
may be used. Thus an ordinary f/2 lens having an e. of 6.25 percent
will have an efficiency of 4.6 percent when used for a magnification of 6.
Since the reflective optical systems under consideration are designed for a specific magnification and since the central part of the
system is masked to maintain contrast, this part being blocked by the
cathode -ray tube, it seems preferable to rate such systems by their
efficiencies rather than f /number.
Let e0 be the efficiency of the system with no masking and el the
efficiency of the central part of the system that is masked. The efficiency e of the masked system is then simply
e= (e0- el)
100%.
Here e0 and el (fractions, not percent) may be calculated approximately from the equations
h,2
h02
eo=-=
f2
p2
el
m2
(m.-1)2
ht2
m2
f2
m2-1
=
)
2
where h is the semiaperture of the correcting lens and ht is half the
diameter of the tube face. For high -efficiency systems e0 will be above
40 percent and el approximately 10 percent so e, the efficiency of the
system with blocking, will be about 30 percent. Neglecting losses in
the system, about 30 percent of the light emitted by an axial point will
be focused into an image point. This corresponds to the efficiency of
an f/0.8 lens used at a magnification of 6.
ALIGNMENT REQUIREMENTS
The center of the correcting lens must be located at the center of
curvature of the mirror and, for uniform illumination over the field,
the axis of symmetry of the correcting lens should preferably coincide
with the axis of symmetry of the periphery (circle) of the mirror.
The tube face must be located so that the center of curvature of the
tube face lies on the axis of symmetry of the correcting lens. For
uniform illumination over the field, the axes of symmetry of periphery of tube face and correcting lens should preferably coincide.
www.americanradiohistory.com
PROJECTION OPTICS
261
The tube face should, of course, be located at the correct axial distance
from mirror or correcting lens for focusing. The viewing screen
should be normal to the axis and at the correct throw.
The most critical alignment items are: (1) Lateral displacement of
the center of the correcting lens from an axis of symmetry of the
mirror, i.e., a line passing through the center of curvature of the mirror; (2) Lateral displacement of the center of curvature of the tube
from the axis of symmetry of the system. For good resolution these
displacements should be kept within 0.001R, where R is the radius of
curvature of the mirror. The permissible tolerances on other alignments are about 10 times greater.
There are two distinct applications for projection television, namely,
in theater television equipment and in television receivers for home use.
PROJECTION RECEIVERS FOR HOME
In a self- contained projection television receivers the optical system
can be mounted near the floor with its axis vertical, projecting the
image straight up and onto a flat mirror inclined at 45 degrees to the
incoming beam of light, and throwing the image on a translucent
screen. Such an arrangement presents the advantages of compactness,
relatively small depth of the cabinet and can be styled along the
familiar lines of a radio console.
A number of such reflective projection systems suitable for home
receivers of the type described have been designed, built and operated
in actual receivers. The smallest of these was built for use with a
cathode-ray tube having face diameter of 3 inches, and consists of a
spherical mirror 9 inches in diameter and a correcting lens 6 inches
in diameter. The largest has tube, mirror and lens diameters of 5,
14 and 9.5 inches respectively. A number of systems in sizes intermediate between the two just described have been built. The throw
or distance between the correcting lens and the viewing screen varies
between 36 and 54 inches and the optical efficiencies are between 18
and 35 percent. In resolution and contrast these systems compare
favorably with well- corrected conventional projection lenses, and do
not limit the performance of present television systems.
SYSTEMS FOR THEATERS AND CAMERAS
A description of the RCA theater television system was published
several years ago.3 The optical system consists of a 30 -inch mirror,
22.5 -inch correcting lens and operates with a cathode -ray tube 7.5
(8)
Landis, D. O., U. S. Patent 2,273,801, filed Dec. 30, 1938.
www.americanradiohistory.com
262
TELEVISION, Volume
IT
inches in diameter. Figure 8 shows the optical system with the cathode ray tube in place. The control console may be seen in the background.
Reflective optical systems built for infinite throw find useful application in television pickup cameras under conditions of low illumination, such as during the last minutes of a football game or in direct
pickup from a theater stage. The great light- gathering power of these
Arrangement of optical system for a home television receiver employing
reflective optical principles. This design gives a large- screen picture with
a console cabinet no deeper than that of an ordinary home radio receiver.
optical systems is demonstrated in Figure 9. An optical system with
infinite throw was pointed from a window in Camden, N. J., toward
the Philadelphia skyline. The bright image of the skyline can be seen
inverted on the dummy tube face, undestroyed by the full daylight
illumination.
www.americanradiohistory.com
PROJECTION OPTICS
Fig.
Fig.
8
-RCA theater
26:
television projector, with control console
in background.
9- Image
of Philadelphia skyline as formed on the face of a
dummy tube by a reflective optical system.
www.americanradiohistory.com
TELEVISION, Volume IV
264
An interesting modification,' applicable to all systems described, is
shown in Figure 10. Here a flat mirror is inserted about half-way
between the cathode -ray tube and the spherical mirror. Since the center of the mirror is blacked out to increase contrast, the opaque back
of the flat mirror cuts very littlt of the useful light coming from the
tube facing the spherical mirror, but the flat mirror permits placing
another cathode -ray tube back of the spherical mirror and facing the
flat mirror. Such an arrangement may be used in theater work since
both tubes can operate singly with roughly the same optical efficiency.
If one tube goes bad the other may be turned on by a flip of the switch.
With some technical difficulty both tubes may be operated at the same
time, the problem arising in the exact super -position of two scanning
patterns.
COST FACTORS IN REFLECTIVE OPTICS
The major objection to the use of reflective optics in television
Fig.
10
-Bi- reflective
optical system, employing two projection
television cathode -ray tubes.
receivers has been the high cost of the aspherical correcting lens. The
spherical mirror, while quite large, is an old and familiar item to the
well-established optical industry, as most of the conventional optical
surfaces are spherical and are easily made. The aspherical correcting
lens, similar to a figure of revolution developed by rotating a shallow
letter S around one of its ends, presents an altogether different problem. Unlike the spherical mirror, such a figure is not a naturally generated surface and there are no machines on the market for
straightforward production of such surfaces. True enough, astronomers, with their traditional patience and lack of hurry, produced
excellent aspherical lenses on machines used for making astronomical
instruments, but only by tedious step -by-step methods.
In the early stages of the development, RCA used methods and
machines based upon astronomical technique. Exceedingly high cost
of experimental reflective optics resulted. The gain in light over the
(9)
Epstein, D. W., U. S. Patent 2,295,779, filed Aug. 17, 1940.
www.americanradiohistory.com
PROJECTION OPTICS
265
conventional projection lens was very attractive, but the cost of such
individually produced lenses was prohibitive. The apparent solution
to the cost problem was that of molding the aspherical lenses from a
suitable transparent material.
PLASTIC CORRECTING LENSES
special development project was undertaken and soon concentrated on investigation of a clear thermoplastic material known under
the name of methyl methacrylate, and sold under the registered trade
.
names of Lucite and Plexiglas.
A new set of difficult problems came to the foreground. The most
formidable of these was that of making molding surfaces of metal
in shapes of the negative replicas of aspherical lenses. Almost as
serious was the problem of obtaining optical finishes on metals. Both
of these problems have been successfully solved.
A flat disk of hardenable stainless steel is first profiled with the
aid of a template. The template itself is filed according to a theoretically calculated curve. The profiling machine has a five -to -one lever
action which calls for a template five times deeper than the final curve.
Profiling is done by diamond wheels. The resultant curve is tested
on a precision curvemeter, and final adjustments of the curve are done
by fine grinding and polishing on a precision polishing machine. The
final optical finish of the surface is the result of proper choice of metal,
proper hardening and tempering, proper choice of abrasives and polishing agents, and most of all, patience and perseverance.
The molding process is essentially that of applying very high
pressure to heated plastic material confined in a heated mold and cooling it. under pressure until it reaches room temperature. The mold
is then opened and the lens extracted. The only operation which
remains is that of boring a hole in the center of the lens for accommodating the protruding neck of the cathode -ray tube. The lens is
then ready for use, with no polishing or finishing of any sort required.
Molded correcting lenses for reflective optical systems possess
very good optical properties, including slightly better transmission
and slightly lesser scattering of light than glass. They do not possess
the surface hardness and scratch resistance of glass, but even without
any special care or protection they have stood up under laboratory
operation for more than three years. The cold flow under operating
conditions of three years was found to be negligible. The cold flow
depends on the operating temperature, which for the plastic lens of a
television receiver is not far from room temperature. Should design
A
www.americanradiohistory.com
266
TELEVISION, Volume IV
considerations call for higher operating temperatures, the new boilable methyl methacrylates can be used.
MOUNTING PROBLEMS
From a practical standpoint, the use of reflective optics in television receivers calls for careful consideration in the mechanical construction of the mounting which supports the optical system and the
cathode-ray tube. This mounting, combining "the barrel" and "tube
Molding press used for producing plastic correcting lenses
for projection television systems.
support," has to fulfill a number of requirements: (1) Since the position of the correcting lens with respect to the mirror is rather critical,
the mount must provide for positive and simple alignment at the
factory; (2) It must be dustproof, since accumulation of dust on the
mirror and correcting lens reduces both the contrast and the illumination, while frequent dusting would be detrimental to the plastic lens
and the front surface mirror; (3) It must be electrically shock -proof
since in some cases final optical focusing of the picture on the viewing
www.americanradiohistory.com
PROJECTION OPTICS
267
screen must be done with a picture and consequently with high voltage
on the cathode -ray tube; (4) The barrel should preferably be made of
metal, to cut off x -rays generated by the cathode -ray tube. These rays
are very soft and weak; nevertheless, they are measurable and should
be screened in; (5) It must provide for positive and convenient initial
adjustments of the tube face position along three rectangular coordinates, one of which coincides with the optical axis of the system.
These initial adjustments may be carried out by the factory and by
experienced servicemen; (6) It must provide for easy tube replacement
by people unfamiliar with optics, such as the average serviceman and
the customer himself; (7) It must provide for easy and safe focusing
after tube replacement; (8) It must be designed to lend itself to such
inexpensive manufacturing processes as stamping or die casting, involving a minimum of machining; (9) It must not deform in transportation and during years of service.
TYPICAL MOUNTING
A layout of a mounting satisfying the requirements discussed is
shown in Figure 11. Here the correcting lens fits into a recess on
the top of a metal barrel, this recess being counterbored for a snug fit
with the correcting lens. The spherical mirror is mounted on the
bottom cover of the barrel by means of a collar and nut through the
center hole of the mirror.
The tube support consists of an arm of insulating material anchored
on the side of the barrel and a metal ring supporting the face edges
of the cathode -ray tube. The tube face is held tight against this ring
by suitable springs. The high voltage is brought to the second anode
of the tube through a dust-tight hole in the wall of the barrel. The
metal ring holding the tube is at high potential, and several inches of
Micalex insulate it from ground. The high -voltage cable has a grounded
shield on the outside and the barrel itself is grounded.
The tube support arm is arranged to slide back and forth, providing for tube adjustment in a direction perpendicular to the optical
axis of the system, say, along a rectangular coordinate x. The support
of the arm is arranged to slide along an intermediate supporting plate
in a direction y, perpendicular to both the x coordinate and the axis
of the optical system. The intermediate supporting plate is made to
slide up and dowii the barrel by means of a screw, providing a focusing
means along the axis of the optical system or coordinate z.
The deflecting yoke is supported by the neck of the cathode -ray
tube and is equipped with dust-proof gaskets. The top of the barrel
www.americanradiohistory.com
TELEVISION, Volume IV
268
may be equipped with a cardboard shield reaching to the upper part
of the television cabinet and preventing dust from settling on the
upper side of the correcting lens.
The arrangement described satisfies the requirements enumerated
more or less completely and allows for variations governed by the
Deflecting yoke-
Correcting
/
gal
/i, ¡
II.
\nu
Direction
of Axis
\/
Y
High - voltage
cable
Spherica I_ <
mirror
Fig. 11- Method of mounting optical components of a projection television
system to give rigidity while permitting adjustments when required.
individual preference of the designer.
APPARENT DETAIL
If one wants to place an enlargement of a given picture on the wall
of a room of a given size, he can find by experiment a size of enlarge-
www.americanradiohistory.com
PROJECTION OPTICS
269
ment that will give an "optimum effect." This size will give a picture that is not unduly blurred and does not require squinting to see
the detail. In television with its intrinsic or absolute detail governed
by the bandwidth of the channel of the transmitter or the wire channel, the subject of optimum size for a given application is of major
importance.
The amount of apparent detail needed for a pleasing television picture will determine how much magnification the picture will stand in
any particular application. For a given amount of absolute detail the
picture size will be larger for hotel lobby applications than it will be
for home use, still larger for auditorium use and much larger for
theater use. The exact sizes may vary somewhat but it is believed
that the buying public will soon find out what value of apparent detail
is the most acceptable for a given use. Consequently, the apparent
detail will determine the size of the projected television picture to be
preferred for each application.
www.americanradiohistory.com
CATHODE -COUPLED WIDE -BAND AMPLIFIERS *¡
BY
G. C. SZIKLAI AND A. C. SCHROEDER
Research Department, RCA Laboratories Division,
Princeton, N. J.
Summary -A general analysis indicates that, in wide -band amplifiers,
stable operation is possible with triodes in circuits using the cathode as a
signal terminal. The amplification, however, is approximately equal only to
the square root of that available with grounded-cathode amplifier, and therefore twice as many tube units are required to obtain the same amplification.
In certain applications, however, the utility of such circuits outweighs the
loss of gain.
A simple radio -frequency amplifier was designed for television receivers, using a cathode -input circuit. By combining a cathode -output and a
cathode -input stage using one single twin -triode tube, a circuit was devised
which compares favorably with pentode stages with respect to gain, stability, and economy, while it has far superior noise characteristics. The new
circuit, called the "cathode- coupled twin- triode" amplifier, provides greater
flexibility than conventional amplifier circuits, and can be used for radio frequency, intermediate- frequency, video, converter, or detector services.
Since the same tube type can also be used for synchronizing and deflection
circuits, the number of tube types can be materially reduced, and greater
standardization with further economical advantages may be obtained. An
interesting application of the new circuit is a novel bidirectional amplifier.
I.
INTRODUCTION
FOR approximately two decades, screen -grid tubes were used
almost exclusively for amplification of high- frequency signals.
The screen grid acting as a shield reduced the effect of the output circuit on the signal circuit, and provided a high- impedance output.
When, with the advent of the video art, in the case of extremely wide band amplifiers, the external circuits had lower impedances, the advantage of the high plate impedance became less significant. In the particular case of the cathode-output (cathode -follower) circuit, for instance,
multigrid tubes were used as triodes purely for the reason that they
had higher transconductance than the commercially available triodes.
As the operating frequencies of radio communication increased, the
transit -time effect became more and more significant. In order to
reduce the effect, the spacing of the tube electrodes was reduced, and
it became increasingly difficult to align several grids in extremely close
*
Decimal classification R583 x R363.4
Reprinted from Proc. I.R.E., October, 1945.
:
V70
www.americanradiohistory.com
.
WIDE -BAND AMPLIFIERS
271
proximity. Thus, in lieu of the screen grid, the grid was used as a
shield between the input and output, and in certain cases the cathodeinput (grounded-grid or inverted) amplifiers provided superior results
in performance and economy. As engineering knowledge about noise
sources expanded, the multigrid tubes were avoided in stages where
the signal was small.
It is the purpose of this report to give a comparative analysis of
vacuum -tube circuits using multigrid and triode tubes in wide -band
circuits. In the case of the triodes, circuits using the cathode as a
signal electrode are emphasized, and a new cathode -coupled circuit is
introduced. This circuit surpasses the advantages of pentode circuits
with respect to economy and stability, and possibly permits a broader
tube -standardization program.
II. DEFINITIONS
As it appeared above, the nomenclature applied td the various amplifier circuits is not too well standardized. As compared with the conventional amplifier, in which the cathode is substantially grounded
with respect to high- frequency current, two other configurations are
possible when the cathode is not grounded. In the first, the cathode
serves as output terminal, and is called the cathode follower, or
grounded -plate amplifier. The second uses the cathode as the input
terminal, and is called the inverted, or grounded -grid amplifier. In the
present paper we propose to regard the cathode as the reference point,
since it is the primary electrode of a vacuum tube (the source of electrons), and we shall use the terms of grounded-cathode (Figure 1),
cathode-output (Figure 2), and cathode -input circuits (Figure 3).
For the circuit shown in Figure 4, we adapted the term of "cathode coupled twin -triode" stage. All circuits in which the cathode is not at
ground, but serves as an input or output terminal, will be designated
as cathode -coupled circuits against the conventional grounded- cathode
amplifier.
III. WIDE -BAND GROUNDED -CATHODE AMPLIFIERS
The basic circuit, and its equivalent network, are shown in Figure
1. This familiar circuit is designed such that, at frequencies of
fo ± pf/2, the amplification is 0.707 times that of the amplification at
is the bandwidth. If this
f0, where fo is the resonant frequency and
stage is preceded by a similar stage, the amplification is then
Age = g,,,/ (Ao
C1Co) (see Appendix I (7)) where gm is the trans conductance, ¿w = 2aAf, C1 is the input, and Co is the output capaci-
/f
www.americanradiohistory.com
TELEVISION, Volume IV
272
rP
°
A41CorP-I
n2
Fig.
1-Grounded -cathode
amplifier circuit and equivalent network.
tance. The grid -to -plate capacitance and other sources of feedback are
assumed to be negligible. Since the last assumption is generally untrue,
in order to reduce the grid -to -plate capacitance a screen is placed
between the grid and the plate.' In this and the following equations,
the value of pu = 2rAf may be taken around any center frequency,
and, accordingly, they are equally valid for video, intermediate-fre quency, or radio -frequency amplifiers (see Appendix II). The formula
given above is for a simple tuned circuit, as shown in Figure 1. With
a coupling circuit tof more complex nature, greater gains may be
obtained, as was shown by Wheeler.' For purposes of simple comparison, only the simple coupling circuit is considered here, but the same
factors of improvement apply in all the cases when more complex
coupling circuits are used.
IV. WIDE -BAND CATHODE- OUTPUT AMPLIFIERS
The basic circuit and its equivalent network are shown in Figure 2.
This circuit is shown in a form to work into a high- impedance circuit,
such as the input of another similar stage. A circuit of this type was
proposed in 1925 in order to reduce feedback in radio -frequency amplifiers.3 A more important application of this circuit became popular in
recent years when it was applied to output loads of low impedance,
such as transmission lines.' This latter type of operation of this cir-
T,
Fig.
=Z
To
2- Cathode- output amplifier
and equivalent network.
' W. Shottky, United States Patent No. 1,537,708.
' H. A. Wheeler, "Wide-band Amplifiers for Television," Proc. I.R.E.,
vol. 27, pp. 429 -438; July, 1939.
'
A. Winther, United States Patent No. 1,700,393.
A. D. Blumlein, United States Patent No. 2,178,985.
www.americanradiohistory.com
WIDE -BAND AMPLIFIERS
273
cuit has been frequently analyzed in the literature.' "6 It was shown
that the input capacitance of the stage is reduced by a factor of (1amplification from grid to cathode) providing greater permissible
impedance for the previous circuit. In general, the circuit behaves as
if the tube had an amplification factor and plate resistance divided
by (µ -I- 1) . For our particular case the amplification is Ago =
V g,p /(po)C1) (see Appendix I (16)). This is (provided the stage is
preceded by a similar stage) the square root of the amplification obtainable from a pentode with the same g,,, and capacitances, thus indicating
that two cathode -output stages in cascade are required to provide gain
in the same order as that of one pentode stage.
V. WIDE -BAND CATHODE -INPUT AMPLIFIERS
The basic circuit and the equivalent network are shown in
Figure 3. In this circuit, the input and output circuits are shielded
by the grid. This method of shielding although disclosed' as early
í5T2
) Egk
ri,
rP
AtJCOrP-1
T =
Fig.
3- Cathode -input amplifier
and equivalent network.
as 1927, became popular only during the last few years.8-10 If the
stage operates from a source of matching impedance, as is the
case when the source is predominantly resistive, the amplification is
(see Appendix I (23)). It
A
(1/2) V µ /[1\0C°rs(AwC0rp
may be observed again that the amplification is less than the square
root of the amplification of the grounded- cathode triode amplifier, thus
requiring twice as many stages for the same over -all gain. When the
signal sources is predominantly reactive, the expression changes to
=
-1)]
5
A. Preisman, "Some Notes on Video Amplifier Design, RCA REVIEW, vol. 2, pp. 430 -432; April, 1938.
6 A. A. Barco, "An Iconoscope Preamplifier," RCA REVIEW, Vol. IV,
pp. 102 -107; July, 1939.
E. F. Alexanderson, United States Patent No. 1,896,534.
8 C. E. Strong, "The Inverted Amplifier," Electronics, vol. 13, pp. 14 -56;
July, 1940.
9 M.
Dishal, "Theoretical Gain and Signal -to-Noise Ratio of the
Grounded -Grid Amplifier at Ultra -High Frequencies," Proc. I.R.E., vol. 32,
pp. 276 -284; May, 1944.
7° M. C. Jones, "Grounded -Grid Radio -Frequency Voltage Amplifiers,"
Proc. I.R.E., vol. 32, pp. 423 -429; July, 1944.
www.americanradiohistory.com
TELEVISION, Volume IV
274
2 rP
2rPGWCo-
I
EST
T=
Z
Fig. 4-Cathode-coupled twin amplifier and equivalent network.
A,= V g, / CL (0C0)
(see Appendix I (26) ) .
While this amplifier has lower gain than the grounded- cathode
amplifier, it finds its greatest utility as a radio -frequency stage between
the antenna and the converter stage because of the fact that it is stable
without the need of a screen grid or neutralization. This type of amplifier generates considerably lower noise currents than a pentode would
in the same service. It provides great improvement, for instance, in
receiving television signals. Since the impedance appearing across the
tube input for high g,,, is low, Z1- [r + (1 /AwC0) ]/ (1 + p.) the tuned
circuit provides an adequately flat response over the whole television
band, "and therefore no tuning means is required for the antenna circuit
for a six -channel receiver.
The circuit diagram of a simple cathode -input radio -frequency
amplifier to be used with television receivers is shown in Figure 5.
Figure 6 shows such an amplifier mounted in an RCA TRK -120 television receiver. Figure 7 shows the inside of the auxiliary chassis. This
amplifier affords an additional amplification of 2 to 4, and a significant
improvement of the signal -to -noise ratio. The heterodyne oscillator
signal is substantially reduced in the antenna, thereby reducing radiofrequency interference between two receivers. Several of these simple
amplifiers were made and attached to receivers in the Princeton area
(approximately 45 miles from New York and Philadelphia), and considerable improvements were obtained in every case.
Since the antenna circuit feeding into the cathode is untuned, some
thought has been given to the question of cross modulation in the
cathode -input radio -frequency amplifier due to two strong carrier
INPUT
2
Fig.
T
5- Television
O
v
radio-frequency amplifier circuit.
www.americanradiohistory.com
WIDE -BAND AMPLIFIERS
Fig.
6-Radio -frequency
Fig.
275
amplifier on RCA television -receiver chassis.
7- Internal view
of radio -frequency amplifier.
www.americanradiohistory.com
TELEVISION, Volume IV
276
signals. Cross modulation is a function of the strength of the signals
and the degree of curvature of the tube characteristic. In this amplifier,
the magnitude of signal voltages appearing between grid and cathode
is less than those in the antenna, since a 1:1 transformer is used for
coupling, and the circuit is highly degenerative. These voltages are
less than those appearing at the grid of a converter tube in television
sets using a step -up transformer for coupling the antenna to the grid.
The amplifier characteristics of a cathode-input amplifier are less
curved because of the high degeneration of the cathode circuit. One is
therefore led to the preliminary conclusion that cross modulation is
less serious in the cathode-input amplifier than in the converter even
though the grid of the latter is tuned. The tuned circuit in the converter is too broad to give sufficient adjacent -channel rejection.
In some cases the input loading is far in excess of that required to
obtain the desired bandwidth. In such cases a compromise between the
T,=
T2=
Fig.
8- Tapped
3z
cathode-input amplifier and equivalent network.
grounded- cathode and cathode -input amplifier may be obtained by
moving the ground from the grid toward the cathode on the secondary
of the input transformer, shown in Figure 8. The limiting factor in
this case will be found in the stability since the grid is a less effective
shield, but with proper tapping of the coil, stable operation may be
obtained. The circuit will behave as if the amplification factor of the
tube were increased by the transformer ratio of the cathode part of
the coil to the total secondary.
These circuits are particularly suitable for antenna -plex systems
as a consequence of the inherently good noise and wide -band characteristics.
VI. WIDE -BAND CATHODE- COUPLED AMPLIFIERS
As has been shown, the cathode -output circuit provides a comparatively high input- impedance circuit, with the additional advantage
that this impedance is not changed materially by external potentials,
www.americanradiohistory.com
WIDE -BAND AMPLIFIERS
277
such as grid bias, plate voltage, etc. It has been proposed to use such
a stage in conjunction with a grounded- cathode amplifier," but such
a circuit, besides requiring the same number of circuit elements as
two stages, uses a pentode tube as the grounded- cathode amplifier. In
some cases this dual stage does not provide adequate stability, and
also does not provide better noise characteristic than a pentode. By
connecting a cathode -output and cathode-input stage together, as shown
in Figure 4, we obtain a high -gain wide -band amplifier stage.
The amplification of a wide -band amplifier of this type is
ACCTT = gm/ (2pwVC1C0) (See Appendix I (41)) which is favorably
comparable to the gain obtained for grounded-cathode amplifiers, particularly since the input capacitance (C1) is reduced by a factor of
(1 + 2AwCory) / (4iwCory) (see Appendix I (46) ) .
The circuit is economical since a coil and a resistor (the coil preferably wound on the resistor) are the only coupling elements required
between the two tube units. The resistor and by -pass capacitor customarily required in a screen supply are eliminated. Since the plate
currents in the two triode units swing in opposite directions, subsequent similar stages have little influence on each other, due to varying
load on the plate supply. By examining the circuit we may notice that
the input and output signals are of the same polarity. Hence, when a
cathode -coupled amplifier is used for video amplification, no attention
need be given to the number of stages in order to obtain the proper
polarity.
A twin- triode tube with a common cathode may be manufactured
more economically than a pentode of the same transconductance. While
this point may be debatable at present, it can be shown that receivers
could be designed in which twin triodes were used in nearly all stages,
and by reducing the tube types, the cost of the preferred -type tube
could be reduced still further. Figure 9 is a block diagram of a
16 -tube television receiver in which 12 tubes are of the twin- triode type.
Table I shows the amplification obtainable from conventional high
g pentodes in grounded- cathode circuits and from twin triodes in
coupled- cathode wide -band amplifier circuits. To allow for the capacitances of tube sockets, wiring, etc., 2 micromicrofarads was added to
the tube capacitances of each terminal, given in the tube handbook,
for miniature tubes. Similarly, for octal metal tubes, 5 micromicrofarads was added. The bandwidth was assumed to be 4 megacycles,
and the gain formulas given above were used. The input capacitance
1-
,
11 P. Selgin, "The Cathode Driver as an R -F Coupling
vol. 28, pp. 26 -28; March, 1944.
www.americanradiohistory.com
Stage, Radio,
TELEVISION, Volume IV
278
ANT
R.E
IF
I.E
AMP
CONY.
ea
CI
=N2
I.F
AMP
AMP
AMP
DET
VIDEO
CO
COI
CC
CO
11311
VERT
DEE
A P
OSC.
CN.
SOUND I
TO
F M CHASSIS
VERT
SEP
6J6
NOR.
POWER
NOR.
NOR.
DAMP
OSC.L
OUTPUT
KINE.
LJ
Fig.
9
-Block diagram
of television receiver using CJG tubes.
of the coupled stages was corrected for degeneration. The tube -cost
figures were also taken from the RCA Tube Handbook.
VII. SPECIAL
CATHODE -COUPLED STAGES
On examination of Figure 4, we may observe
that the two grounded
electrodes correspond to the input and output electrodes in the reverse
direction. By injecting a signal of different frequency through a
resonant circuit that appears substantially as a short circuit for the
signal applied in the original direction to the grid of the second tube
T_,, and taking it off the plate of the first tube T1 in the same manner,
we obtain a bidirectional amplifier as shown in Figure 10.
The terminals of the two signals are completely independent, and
the capacitance of each will determine the bandwidth and gain of its
own signals. An amplifier of this type may be useful for bidirectional
relay stations, reflex circuits, etc. Approximate calculations and experimental results indicate that with simple resonant circuits the two
signals must be approximately twice their bandwidths apart. With a
TABLE I
COMPARATIVE WIDE -BAND AMPLIFIER DATA
Tube Types
Bandwidth
4
Circuit
No.
Base
6AC7
Octal
6í,B7
Octal
6AGS
6J6
Mini-
ature
Mini-
ature
GroundedCathode
Grounded Cathode
GroundedCathode
CathodeCoupled
Equivalent
Root MeanList
Square
Amplifi- Grid Noise Price
$
cation
AV
megacycles
Cm
AV
Cgk
Nµf
55f
9000
11.0
6.8
8.0.
5.0
5.0
14.2
5000
8.8
12.6
1.15
5000
6.5
2.2
1.8
17.6
10.4
2.15
1.6
10.4
7.5
1.85
5300
Co
www.americanradiohistory.com
1.75
WIDE-BAND AMPLIFIERS
279
smaller frequency separation, the electrodes, which are supposed to be
grounded, do not provide constant potentials, and, due to the regeneration, both pass bands are reduced. Further work on more complex
circuits may permit the choice of closer signal frequencies.
The experimental chassis containing a bidirectional cathode-coupled
stage is shown in Figure 11. The signals applied were the frequency
bands 8.5 to 13 megacycles and 24 to 28.5 megacycles. A gain of
approximately 12 was obtained in both directions with a 6J6 tube.
Figure 12 shows a simple intermediate -frequency transformer construction for the frequency band 8.5 to 13 megacycles with a 6J6 tube.
The advantage of the bidirectional amplifier could be summed up by
claiming a total amplification equal to the square of that of the unidirectional stage, or by claiming twice the bandwidth with the same
gain.
Fig.
10- Bidirectional -amplifier
circuit.
The cathode -coupled stage can be used also as a frequency converter
as shown in Figure 13. The grid of T2 is substantially grounded for
all frequencies except for the frequency of the tank circuit of the local
oscillator. The local oscillator varies the transconductance of the tube,
and therefore provides an intermediate -frequency output across the
tuned circuit connected to the plate. The second tube T2 acts as a
cathode -output stage for the oscillator signal, and attenuates it by 6
decibels toward the antenna since it works into an impedance like its
own. The first tube T1 further attenuates this signal by providing a
divider through its grid-cathode capacitance and the input impedance.
A simple cathode -coupled two -terminal oscillator circuit12 is shown
in Figure 14. This is merely the twin- triode cathode -coupled amplifier
described above, in which the output plate is coupled back to the input
grid through some coupling impedance. The grid of the cathode-input
12
M. G. Crosby,
United States Patent No. 2,269,417.
www.americanradiohistory.com
TELEVISION, Volume IV
280
is normally returned to ground. However, by properly
grid, it is possible to obtain a frequency variation in
this
biasing
excess of plus or minus 75 kilocycles about a mean frequency of 50
section
T2
Fig.
11- Experimental
bidirectional -amplifier stage.
megacycles with a 6J6 tube, with a bias variation of plus or minus one
volt. In a television or frequency-modulation receiver this feature can
be used to good advantage to provide vernier tuning or automatic frequency control without adding a reactance tube. En a frequency-modu-
www.americanradiohistory.com
WIDE -BAND AMPLIFIERS
Fig.
12- Miniature
intermediate -frequency transformer and tube.
R.F
Fig.
13- Cathode -coupled
frequency converter.
To
A.F.C.
Fig.
14- Two -terminal
oscillator.
www.americanradiohistory.com
281
282
TELEVISION, Volume IV
lation transmitter it may be possible to use this property to obtain
direct frequency modulation of the oscillator.
Further applications of the cathode -coupled stage may include
lock -in oscillators, reactance tubes, self -oscillating converters, etc. The
economy and the standardization possibilities of the circuit may well
suit it for a large number of different applications.
APPENDIX I
DERIVATION OF GAIN FORMULAS FOR `VIDE-BAND AMPLIFIERS
The grounded- cathode amplifier, with its equivalent network, is
shown in Figure 1. If we desire to maintain an amplification at a frequency f = [(do ± 1/(2.6m)] /27r that is approximately equal to 71 per
cent of the amplification at resonance (see Appendix II)
7'8T2
= 1/ (AwC1)
(1)
provided we have unity coupling in our transformer. C1 includes the
capacitance in the primary divided by the square of the transformation
ratio. The source resistance r$ may be a loading resistor, the surge
impedance of a transmission line, the radiation resistance of an
antenna, etc. The voltage applied to the grid is according to Thevenin's
theorem
E
ERT.
(2)
The output of the tube is
E= E,4
I
('LZL)
/(r, +Z1)]
(3)
but since ZL is determined by an external loading resistance which is
in shunt with the plate resistance rp according to the relation
(rPZL)
/ (rP -f- ZL) = 1/1/(C0)
(4)
or
ZL
= rP/ ( pwCorp
-
1)
(5)
if (1) and (5) are substituted into (3), and both sides are divided by
Eq,
the amplification A is
A-Ea/Es= [1/
AwClrs)] [1/(pwCarp)].
www.americanradiohistory.com
(6)
WIDE -BAND AMPLIFIERS
283
If the stage under consideration is preceded by a similar stage, we
in which case we obtain an over-all
may set r8 equal to 1 /(c C
response of 0.5 at fo ± (Af /2), and by replacing µ /rp by g,,,, (6) will
take the convenient form of
),
A = gm/ (D,,,C1Co)
(7)
a formula equally useful for tiodes or pentodes if feedback can be
neglected.
The equivalent noise resistance of the grounded- cathode amplifier
is given by
Requ
= 2.2 /gm
(8)
while for the pentodes
Ruequ= [2.2 /(g,)0(1
+a))]
[1
+ qa
(Iti /gnu)]
(9)
a = Ic2 /Ia
I. is the plate current, and
Ice is the screen current. The root-meansquare grid -noise input may be calculated then with the aid of the
equation
= 1.3
X 10-10
VR equ of
(10)
The cathode -input amplifier, with its equivalent network, is shown
in Figure 2. Again for bandwidth considerations we make the assumption that R8T12= 1 /(AaC1) (see (1)) and
[(rpT,,) /(µ +1) +RL] /[(7'1,TO2) /(µ +1)
RL] =Ae,C0
(11)
where the input capacitance C1 is equal to the sum of the reduced grid cathode capacitance due to degenerations and the incidental capacitance
to ground, while RL is the equivalent parallel resistance of the losses in
the output circuit. By rearranging (11),
.
R1,(µ
+1) /T02= rp(RLAÚC -1).
The amplification is
ACO
= TO
-µ+1
RL/T02
l
ro
µ
+1
RL
+
T02
www.americanradiohistory.com
(12)
TELEVISION, Volume IV
284
RL
(µ + 1)
T2
= To
µ+1
rfl
+
RL
(13)
(µ+1)
T2
If we substitute from (12)
Aoo
If
µ»
1
To
1
(
and we substitute for
To
I
RRL
, -1) r
(RLp Co
RLpwCo
fl
rfl (N/pCo
if VRL is high we may replace p /rr by
.o
-11
\ RLpC
1
µ,
(14)
I
from (12)
µRL
e1
1
/I
o
l
/RLpCo /
g,,,,
(15)
(15) will take the form
=Vg /(poC,).
(16)
Comparing (7) and (16), we may notice that the latter is in the
order of the square root of the former, thus two cascade stages are
required for amplification of the same order of magnitude. The equivalent noise resistance in this case is equal to that of the groundedcathode amplifier.
The cathode-input amplifier, with its equivalent network, is shown
in Figure 3. This circuit may be analyzed in two ways. In one instance,
the input impedance of the tube, which is usually very low, is matched
to a predominantly resistive input, such as an antenna, a transmission
line, etc. In the second case, the transformation ratio is reversed and
the input impedance loads a tuned circuit to provide the required
bandwidth.
In the first case, for optimum power transfer
+1) (r7, +Z1) /p.
r8T2= (r7, +ZL) /(µ
From the equivalent network, it may be seen that
Ekµ
+E8T= I(r8T2 +ro +ZL)
www.americanradiohistory.com
(17)
(18)
WIDE -BAND AMPLIFIERS
285
and
Ek = E8T
-
(19)
Ir3T2
then
I= [(µ +1) EsT] /[(µ +1)
rOT2
If we multiply (20) with the plate load
E8, we
ZL
+rP+ZL].
(20)
and divide through with
obtain
Act
= EO/E8 =[(µ+
1)TZ,,I /Iµ +1)r,,T2
If we substitute from (17) for
p»
1,
1
= (1/2) V p./ [L
(21)
T, we obtain
equation (22), after substitution for
210,
+Z,,].
\
(N/TAr.
1.
r(µ+l - I(rP-f-ZL)
Aci=[ (µ +1)ZL]
If
the form
+r
oC0r8 (AwCori,
Z1,
(22)
from (5), takes
- 1) ].
(23)
For the second case, when the cathode-input amplifier operates from
a tap on a tuned circuit fed by a comparatively high impedance, such
as another stage of amplifier,
T2
=
(p.
+
/
1) (rPAwC0)
(24)
µ/ (rPpwC0).
If this value is substituted in the gain equation
A =[(p +1)Z1í /I
if
pi»
1
(rP
yields the equation after substitution for
Aei
(25)
+Z1,)T]
ZL
= Vgni/ (OwCp).
from (5)
(26)
The equivalent noise input may be calculated from (8) with the aid
of the equation
ekN" =
1.3 X
ZL
10- 10VR efh,pf
+r
+R1(µ +1)
µR1
(27)
A compromise between the grounded-cathode and cathode-input
amplifier may be obtained by connecting the ground to a tap on the
input transformer, as shown in Figure 8. From the equivalent network
we may see that
www.americanradiohistory.com
1'F;IJ;'I l.tilU:C,
286
E8T1
+
µE7,T2
l'lm,
I17
= I(rçT1' + rp
-}-
(28)
Z,.)
and
Ek = E8T 1= Ir8T12
(29)
If we solve for I, we obtain
I=
[(µT2 +1)E3T1] /[(µT2 +1)r8T12 +rp +Z1j
(30)
which is the same as (20) except that in place of p we have 1,,T2, and,
accordingly, we increased the amplification factor by T2.
The cathode -coupled twin -triode amplifier is shown in Figure 4,
with its equivalent network. From the equivalent network we may
observe that
Z1P+ 0:1-z2)Zk=µE,rr=i4(E1-Ek)
=µE1-
(i1-i2)µZa:
(31)
and
(i2-i1)Zk+i2
+Z,,)=pE,,
(32)
from (31)
i1[rp+Zn(p.+1)] -i2[Zk(µ+1)] =µE1
(33)
and from (32)
- i1[Z,(!1 +1)] +i2[r
If (33) is divided through with
rp
+ZL +Zk(µ
-}-
Zk
+1)] =0.
(34)
(µ + 1), we have
i1- i2[Zk(µ +l)] /[rP +Zk(µ +1)] =.(µE,) /[rP +Zk(µ +1)]
(35)
and if (34) is divided through with rP + Zk(µ + 1), we have
-i1+ i2[rP +ZL +Zk(µ +1)] /[Zk(,t +1)] =0.
(361
If we add (35) and (36) we have
CrP+ZL+Zk(µ+1)
Zk
(µ + 1)
µE1
.(37)
i2
Zk(µ + 1)
rP
+ Z,;(µ +
Thus the solution for the plate current of
1)J
T2
1'p
is
www.americanradiohistory.com
+ Z,
+
1)
WIDE-BAND AMPLIFIERS
+ 1)
rPY -I-
Z,,rP
+ 2ZrP(
+
1)
- ZZ Iµ + 1)
287
(38)
If we multiply through with Z,, and divide with E,, we obtain the
amplification from grid number 1 to plate number 2
ACC;,/,T
_
µZLZx(µ + 1)
rP2
+rPZr+2Z,rp (µ+1) +Z,,Zi.(µ+1)
(39)
If Z is much larger than (rP + ZL) / (2 (µ + 1)) , which is an easy
condition to fulfill, (39) will take the form
ACCTT
= µZr./ (2?P + Z
(40)
If we substitute L wC0 for (2r7, + ZL) / (2r,,ZL) and multiply by the
input -circuit transfer VC /Cti, where Co is the output capacitance of
the preceding stage factor (assumed to be equal to that of the stage
under analysis) and Ci is the input capacitance corrected for degeneration, we have
AocTT
= 9',/ (200)-VCr).
(41)
The grids of both tubes are at equal gain points with respect to
their cathodes (in other words, the same gain is obtained from either
grid to the output of the plate circuit, when the signal is applied
between the grid and the cathode), and thus both tubes contribute
equally to the total noise. The apparent noise generating resistances
in either grid is equal to (8), and therefore the equivalent noise resistance between the input grid and cathode is
R
eqU
= 4.4/g,
(42)
where g,, is the transconductance of one tube unit. The root -meansquare grid -noise equivalent may be determined with the aid of (10).
The equivalent noise resistance of the cathode -coupled twin -triode
amplifier is considerably better than that of a pentode amplifier, and
therefore a great improvement can be obtained in the noise factor by
using cathode -coupled amplifiers in the early stages of amplification.
A particularly useful instance is when cathode- coupled intermediate frequency amplifiers are used after low -gain frequency converters.
To evaluate C.,, we calculate from the equation
www.americanradiohistory.com
TELEVISION, Volume IV
288
C;=CI,+CyP+
Cok
(1-A1)
(43)
where CD is the incidental (socket, wiring, coil, etc.) capacitance, C7,
is the grid -to-plate capacitance, and Cuk is the cathode -grid capacitance.
A, is the amplification of the first tube only.
Ai= µZk/ [rp + Zk (p +
If µ
»
_ [µ/(i +1)1
1
1) 1
[(rP +Zp) /(2r1, +ZP)1.
(44)
and we substitute
Zp
= 2r, /(2).5
u,C0
-
1)
(45)
into (44) we have
A1
= (2AmCrp+
1)/(40u0CorN).
(46)
z
I
00
.707
Ca,,
o
ew
Fig.
15- Impedance
characteristic of low -pass filter.
APPENDIX
II
The bandwidth in the present paper is considered as the frequency,
or the separation between the frequencies, at which the amplification
is reduced by a factor of 1/x/2 of the value at the frequency of maximum amplification. The gain is a direct function of the impedance of
the output circuit; therefore we may examine the impedance, and
particularly its absolute value, directly.
In the case of a simple resistance -capacitance circuit as shown in
Figure 15, the absolute value of the admittance at
YI=
If we multiply by
R and
V2
R
(47)
_
rationalize
V2 = V1 +
0,12C2R2
www.americanradiohistory.com
(48)
WIDE -BAND AMPLIFIERS
since Aw
= (di
-
289
0,
Ow
= 1/ (RC).
(49)
In the case of the band -pass analogy of this circuit, shown in Figure
the admittance at the resonant frequency wo is 1 /R, and at the
frequencies col and w_ the absolute value of the admittance is
16,
lYl=
1
-+7w,C-R
R
(50)
w
If we multiply by R and rationalize, (50) becomes
v2 = v/1 + wn2C'R "[1
w02 =1 /LC and we = wl, or
If (51) is squared, it yields
where
-
(0,02/.(:2)] 2
(51)
w...
I
0
.707
o
?ig.
16- Impedance
w,.RC11-
or
w,.RC [
1-
characteristic of band -pass filter.
=1
(w02/w(.2)
] =
(0)02/(0,.2) ]
-
1
if
w,.
>
wo
if
w,.
<
wo
(52)
If we rearrange (52), and solve the quadratic
wr2± [1/RC]
w,
_
we- wo2=0
- [1/(RC)] ± v[1/(R2C2)]
w_
= (02- (01,
+4wo2
2
[1/(RC)] ± v[1/(R2C2)]
2
Since Aw
(53)
Aw
= 1 /RC.
www.americanradiohistory.com
(54)
+4wo2
IMPROVED CATHODE -RAY TUBES
WITH METAL -BACKED LUMINESCENT SCREENS*1
L'y
D. W. EPSTEIN AND L. PENSAK
Research Department, RCA Laboratories Division,
Princeton, N. J.
Summary- Considerably improved cathode -ray tubes result from the
application of a light-reflecting, electron -pervious, thin metallic layer on the
beam side of the luminescent screen. Although this has been realized for
some time, it is only recently that practical methods for applying such a
metallic layer in kinescopes have been developed.
Observations and measurements on such tubes, using aluminum for
backing, show that under appropriate conditions such tubes possess many
advantages over similar conventional tubes. These are:
1. Improved efficiency of conversion of electron beam energy into useful
light -in other words, more useful light output for a given beam
power input.
2. Elimination of ion spot-thus making other, generally less direct,
means for eliminating the ion spot unnecessary.
3. Improved contrast.
4. Elimination of secondary emission restrictions -thus permitting
the use of high voltages and screen materials with poor secondary
emission.
ONE of the outstanding quests in the cathode-ray tube field has
been the search for means of increasing the brightness of the
pictures on the face of the tube. Previous methods consisted
primarily of efforts to increase beam power -that is, raising the voltage
well
and increasing the current by improvements in electron optics
in
convertas a search for luminescent materials with greater efficiency
ing beam power into light. The most recent step in increasing light output is the application of a light -reflecting metallic layer on the beam
side of the fluorescent screen.
Many practical tubes with metallic layers on the screens were built
and used six and seven years ago. However, these tubes were limited
to high -voltage operation and the metallic layers did not possess the
light- reflecting properties which characterize the new metal films. The
advantages of having a thin reflecting layer have long been anticipated
and, to a limited extent, observed in the laboratory. It is only recently,
however, that methods have been developed which will make such tubes
possible and practical.
Before showing how this is accomplished, it is worthwhile to review
briefly the pertinent part of the state of affairs at the luminescent
-as
Decimal Classification: R583.6 X R138.313.
Presented at the I.R.E. Winter Technical Meeting, January 24, 1946,
in New York, N. Y. Reprinted from RCA REVIEW, March, 1946.
*
290
www.americanradiohistory.com
IMPROVED CATHODE -RAY TUBES
291
LUMINESCENT
MATERIAL
CONVENTIONAL KINESCOPE SHOWING TYPICAL
DISTRIBUTION OF LIGHT FROM A SPOT
GLASS
FACE
Fig.
1.
LUMINESCENT
MATERIAL
GLASS
FACE
--
ALUMINUM
REFLECTING
BACKING
KINESCOPE WITH REFLECTING METALLIC FILM
SHOWING GAIN DUE TO IMPROVED DISTRIBUTION OF
LIGHT
NEW
Fig. 2.
www.americanradiohistory.com
292
TELEVISION, Volume IV
screen in a conventional kinescope. This is shown schematically in Figure 1. The region in the circle is a greatly magnified and somewhat distorted diagram of a small section of the tube face of which one element
is fluorescing. Generally, at least 50% of the light generated in the
screen is emitted towards the electron gun in the tube. Another 15 -25%
is lost by total internal reflection in the glass of the tube face. Thus
only about 25-35% of the total light generated is emitted in the forward direction to constitute the useful light output of the tube. It
should also be pointed out here that some of the wasted light is harmful
in that it is scattered back onto the screen to set a limit on possible
contrast in the picture. There are several mechanisms for this. One is
the back-scattering of the light which strikes the inside walls of the
tube; although the light is largely absorbed by the blackening on the
wall, some of it comes back to the screen. Another is the light from one
portion of the screen which can illuminate other regions directly
because of the curvature of the face. Some of the totally reflected
trapped light in the glass is reflected back onto the screen and is scattered causing what is known as halation.
Figure 2 shows a tube whose screen is covered with an electron -pervious, but light -reflecting, metallic layer. Now it is seen that the light
which previously would go towards the rear (electron gun) is reflected
forward into the direction of viewing. Thus without an increase of
light generated, the efficiency of conversion of electron beam power into
useful light has been increased. At the same time, some of the limitations on contrast have been removed, that is, the back -reflected light
and the effects due to curvature of the face. Experiments show that the
large area contrast is considerably improved by a factor of three to ten
times; the detail contrast, being primarily limited by halation, is only
slightly improved.
The properties which this metallic layer should possess are: (1) it
should be thin enough and of the right kind of metal to cause negligible
absorption of the electron beam at the desired operating voltages; (2)
it should be opaque, relatively smooth, and highly light- reflecting, so
as to act as a mirror; (3) it should have sufficient conductivity to conduct the full beam current; (4) it should be strong enough to withstand
the stresses due to effect of the focussed electron beam; (5) it should be
durable enough to be able to withstand the necessary subsequent processing of the tube; and (6) it should be of a metal that will not chemically react with the luminescent screen material.
The metal chosen to work with is aluminum, because it combines
properties which provide the best compromise in meeting the above
conditions. It is easily applied by evaporation and does not affect
www.americanradiohistory.com
IMPROVED CATHODE-RAY TUBES
293
luminescent screens. Its ability to meet condition (1) is indicated in
Figure 3 where are shown a group of calculated curves giving the fraction of electron beam power that is passed by films of various thicknesses, as a function of the initial beam voltage. It will be noticed that a
10,000 -volt beam will retain only 15% of its incident energy on passing
through an aluminum film 5,000 A thick. A film 2,000 A thick will pass
FRACTION OF ELECTRON BEAM POWER
PASSED BY ALUMINUM FILM AS A FUNCTION
OF BEAM VOLTAGE AND FILM THICKNESS.
IA
6
0.9
O
O
w
}
ti
0.8-
V
a
II-
0.7Z
CC
OA
-
CC
w
0
o_
î
0.5
w
°°
OA
Ó
z
2
0.3
I-
v
CC
LL
0.2
a1á
0.1
0.0
0
10
20
40
30
BEAM VOLTAGE IN
Fig.
50
KILOVOLTS
3.
fifty -seven per cent and a film 1,000 A will pass about 77% of the
energy. If we assume that the effect of the mirror is to double the
apparent brightness, then it is evident that a tube operating at 10,000
volts and with a film about 3,000 A thick will show no difference from
an unaluminized standard tube at the same voltage. It should be
noticed, however, that a moderate increase in voltage causes a rapid
www.americanradiohistory.com
294
TELEVISION, Volume IV
decrease in percentage loss of beam energy in the film. Experience has
indicated that the most useful range of film thickness is between 500 A
and 5,000 A.
In order for the film to meet condition (2), "that it be relatively
smooth and mirror like ", it has been found possible to cover the fluorescent screen with a thin film of organic material stretched over the
crystals like a blanket. This provides a smooth surface upon which the
aluminum can be evaporated. If such an intermediate film is not present, the aluminum will be broken up on evaporation so that it will
not have its reflecting properties nor will it be continuous and conducting in the thicknesses necessary for low voltage operation. In
order to obtain conductivity without the organic film, it would be necessary to evaporate five to ten times as much aluminum as is now necessary. This is why previous metallized screen tubes were restricted to
high voltage operation.
These earlier tubes were aluminized in order to avoid undesirable
effects due to poor secondary emission from the screen. It can readily
be shown that if one tries to operate tubes at a voltage such that the
secondary emission ratio from the screen is less than one to one, the
screen will accumulate sufficient charge to slow up approaching electrons to a velocity at which the secondary emission is unity. This means
that the screen is effectively operating at a voltage that may be considerably less than that applied to the tube. This is known as the "sticking" effect and is almost entirely corrected by providing a conducting
layer over the screen. The new method of providing an aluminum film
makes possible the correction of the effects due to secondary emission
dfficulties in tubes operating in the voltage range in which kinescopes
are now operated. Thus the choice of luminescent materials for the
screen is enlarged and improved techniques for applying these screens
to the tube face are made available.
This aluminum film also provides a new line of attack on the old
television tube problem of ion spot. Ions can be completely stopped by
a film of aluminum that will readily pass electrons. Experience with
tubes in the laboratory has shown that, with the right set of conditions
-such as proper aluminum thickness and reasonably low gas pressure
-tubes can be made which will show no ion spot at normal operating
voltages.
Among other advantages of the aluminum film are the protection of
the phosphor during processing and life and the improvement of the
stability of the pattern with regard to disi,lacement due to surface leakages on the face of the tube such as are produced if one touches the face
of an operating kinescope.
www.americanradiohistory.com
IMPROVED CATHODE -RAY TUBES
'
295
Figure 4 gives the efficiency in candle power per watt as a function
of applied voltage and at a fixed beam current obtained for two laboratory-made 12 -inch tubes, identical except that one was aluminized. These
curves are typical of measurements made on a number of tubes. It is
to be noted that at the lower voltages the unaluminized tube has the
higher efficiency whereas above the cross -over voltage the aluminized
tube has the higher efficiency. The cross -over voltage which is cons.s
s.
VARIATION OF EFFICIENCY WITH
APPLIED VOLTAGE
5
N
40
35
\)0
Z_
25
UNALUMINIZEO
Z
U
20
W
/I
15
/
1.0
V
I
I
I
O
1
2
5
4
5
6
7
8
10
9
IN
APPLIED VOLTAGE
11
12
17
14
15
KILOVOLTS
Fig. 4.
trolled by the aluminum thickness is primarily dictated by such considerations as operating voltage and ion spot elimination. As seen from
Figure 4, the increase in efficiency above the cross-over voltage is quite
considerable; for luminescent screens with poor secondary emission
characteristics, the gain may be considerably greater than that shown
on the figure.
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION AND ITS EFFECT
ON TELEVISION PICTURE CONTRAST *t
BY
E. W. HEROLD
Research Department, RCA Laboratories Division
Princeton, N. J.
Summary -The objects of this paper are (1) to investigate the effect
on a television receiver of a c -w .r interfering signal which lies in the high end
of the picture band, (2) to set up a maximum permissible interference level,
and (3) to correlate this level with radiation from the local oscillator of
superheterodyne receivers.
It was observed that the chief annoying effect of interference at the
high end of the video band was a loss in contrast. A strong interference,
in fact, caused a complete loss in contrast or even a negative picture. Overall contrast gradation curves were computed theoretically which checked
the experimental observations; the observations and computations indicated
that a 20 decibels signal -to- interference field strength ratio at the antenna is
a minimum satisfactory value. To maintain this ratio in a 500 microvolt -permeter region of a desired transmitter, nearby receivers must have a radiation below 0.01 microwatts. Pre-war receivers, which used no radio frequency stage, radiated 100,000 times as much as this and were extremely
unsatisfactory. A grounded -grid triode radio frequency stage may give a
reduction of about 30 decibels or more and a pentode radio frequency stage
may be made even better. Other remedies are also discussed but all increase
receiver cost somewhat. However, it is made clear that an adequate television service will require suppression of radiation if the frequency assignments are such as to make interference possible.
Throughout this paper "c -w" indicates "continuous wave ".
I. INTRODUCTION
HE interference caused by local -oscillator radiation from superheterodynes has long been recognized as an important problem
in receiver design. In spite of this, very few published papers
indicate quantitatively how much radiation is present from various
receiver circuits or how much radiation might be considered tolerable.
In the sound broadcast field, even with the commonly used multi -grid
mixers and converters which give partial separation of the local oscillator from the antenna, the radiation problem is serious in the shortwave bands." 2, 3 In television reception, it has been common practice
*
Decimal Classification: R583.15.
f Reprinted from RCA REVIEW, March, 1946.
R. Moebes, "The Superheterodyne Receiver as a Source
H -F Interference," Telégr.- Fernspr.- Funk -u. Fernschtech., Vol. 29, pp. of
199 -201, July,
1
1940.
2 R. Moebes, "On the Permissible Value of Local
Oscillator Voltage at
the Antenna of Superheterodyne Receivers," Telégr.- Fernspr.
Funk -u.
Fernschtech., Vol. 31, pp. 217 -222, August, 1942.
3 G. S. Wickizer, "Radiation
from Superheterodyne All -Wave Receivers," unpublished report of RCA Communications. Inc., April
7, 1937.
296
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION
297
to use triode or pentode mixer tubes because of their high signal -tonoise ratio4 and, when no radio frequency stage is used, the radiation is
high. In the New York area, the channel assignments throughout the
war were such that, with the usual 12.75 -megacycle intermediate frequency, a receiver tuned to channel 1, 50 to 56 megacycles, radiated a
local -oscillator frequency of 64 megacycles which lay in the upper video
frequencies of channel 2, 60 to 66 megacycles. Post -war frequency
assignments and choice of intermediate frequencies will undoubtedly be
different but the problem remains and is the reason for the writing of
this paper. The work to be described is also applicable to other types
of c -w interference (such as sound carriers in the picture channel and
harmonic radiation from amateur and other services) and, to a lesser
extent, to certain types of noise interference.
To the writer's knowledge, only one study of the interference
problem in television has been published to date.' This study provided
an excellent start, but was made using British television standards,
with viewing tubes and picture pick -ups in common use at the time,
and was entirely subjective. Furthermore, when c -w interference was
studied, the interference was introduced into the video circuit so that
an interference pattern might be observed at all light levels. Practically, when the interference comes through a receiver antenna circuit
and with light levels such that the picture carrier is of very small
amplitude, the interference pattern may not be observable on a kinescope or viewing tube. The U. S. standards, which incorporate negative
modulation and vestigial sideband operation, require other special
consideration. The present report is intended to treat the problem
when U. S. standards are used; the conclusions will be based on
objective analysis supported by a subjective study.
In the reception of a television picture, a small interfering signal
will give rise to a pattern which can sometimes be observed on the
viewing screen at certain light levels. If the interfering frequency is
close to the picture carrier, the "beat" interference is of low frequency
and gives rise to relatively large vertical or horizontal bars (i.e., large
detail patterns). Jarvis and Seamans showed that such a condition is
the most annoying to the viewer, particularly when the bars are stationary, i.e., the beat frequency is synchronized with the scanning
system. With present U. S. standards, the video channel is about 4
in
4 E. W. Herold, "Superheterodyne Converter System Considerations
Television Receivers," RCA REVIEW, Vol. 4, No. 3, pp. 324 -337, January,
1940.
5
R. F. J. Jarvis and E. C. H. Seaman, "The Effect of Noise and Interfering Signals on Television Transmission," P. O. E. E. Jour. (Brit.) ,
Vol. 32, pp. 193 -199, October, 1939.
www.americanradiohistory.com
TELEVISION, Volume IV
298
megacycles or more wide so that low-frequency beat interference
(under 1 megacycle) is not as probable as higher frequency beats (i.e.,
small detail patterns) Furthermore, it would be wise to choose an
intermediate frequency so that receiver local -oscillator radiation will
not produce the most annoying interference, namely, large detail patterns. In the present study, therefore, only higher frequency beats will
be considered and, since synchronization with the scanning system is
unlikely, it will be less important to consider the annoyance of the
possible small -detail fluctuating pattern, and more important to consider other effects due to the interference. The chief one of the other
effects is a degradation of picture quality due to a loss in contrast.
.
II. PICTURE
CONTRAST WITH SMALL CONTINUOUS WAVE INTERFERENCE
In order to obtain an understanding of how picture contrast is
affected by an interfering c -w signal, let us look at Figure 1. At (a) is
shown a typical black -white transition in a simulated television modulated signal with conventional negative polarity (i.e., decreasing carrier
for increasing light levels). During the black portion, the transmitter
sends out nearly maximum output, increasing to a peak only for the
synchronizing and blanking interval. During a white picture, the
carrier of an ideal transmitter is very low, substantially zero for the
brightest light values; only for the synchronizing pulses is peak carrier
amplitude attained.* When such a signal is received, the final detector
follows the carrier envelope and the direct current restorer system
operates so that black level is set at the point shown. Such a signal will produce maximum light output on the viewing tube (kinescope)
for the
.white part, and minimum light for the black part, the ultimate
contrast
range being set by the picture viewing tube capabilities, room lighting,
etc.
In Figure 1 (b) is shown a similar received carrier combined with
an interfering unmodulated carrier whose frequency is assumed
to be
such that the "beat" is in the high video range. Assuming the
black
level setting of the viewing tube is unchanged from Figure
1 (a), it
is seen that the originally black portion of the signal envelope
now has
small periodic excursions toward white. If the "beat" is
high,t the
eye will not observe the checkered nature of the pattern
so much as
the fact that the general black level illumination has been raised,
i.e.,
the blacks now look grey. With the idealized 100% modulated
carrier
In practical transmitters,
modulation is not always reached so
that white level may correspond 100%
to a larger carrier than shown on Figure 1.
This changes the effects described quantitatively
by a small amount but,
qualitatively, the idealized 100% modulation herein
treated is entirely
adequate to explain the behavior.
f Or if the viewing distance is sufficiently large.
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION
299
here assumed, in the white portion of the picture, there is no "beat"
since the picture carrier is zero.* The interference, however, causes
a spurious carrier to appear at the receiver so that, instead of a completely white output, the brightness is decreased again toward the grey.
Thus, both the black and the white portions of the picture are shifted
toward each other, i.e., toward a neutral grey. The picture contrast
has then suffered.
If the receiver controls are readjusted, of course, the over -simplified
case of a black to white transition, which we have been discussing, can
be corrected while the interference is present. With an actual picture,
however, no correction is possible without an almost complete loss of
BLACK
PICTURE
WHITE
PICTURE
(O RECEIVED CARR ER WITH
BLACK
LEVEL
WHITE
LEVEL
INTERFERENCE
NO
BLACK
LEVEL
WHITE
LEVEL
HYIIIIIIIIII IIIIIIIIIIIIIi
iIIIIIIIIIIIIiIlllllllllllllll
(b)
SMALL INTERFERING
Ii
C
111111111,11111111111111111111!
I
1IIIIIIIIIIIIIIIIIIIIIIIIIIIII1
-W ADDED
,l
BLACK
LEVEL
WHITE
LEVEL
11v11
1
(C) LARGE INTERFERING
EOUALLI NG
Fig.
C -W ADDED, APPROX.
PICTURE CARRIER
1- Received television signals
showing how an interfering c -w leads to
loss in contrast, even to the point of a negative picture (case c).
the picture tone values at the extremes of brightness ; the loss will be
particularly serious in the darker portions and can be interpreted as
a loss in gradation contrast or "gamma." If the direct current restoring
system of the receiver follows the peak values of the "beat" dgring synchronizing intervals, there is less tendency for the white part to become
* It should be noted that this condition cannot be duplicated by introducing the interference in the video frequency band as was done by Jarvis
and Seaman (Reference 2).
www.americanradiohistory.com
300
TELEVISION, Volume IV
grey but the raised brightness of the black becomes worse. Practically,
many present television receivers use a direct current restoring system
which will follow the average, not the peak, of the synchronizing pulse
whenever the "beat" lies above a megacycle or so. This is caused by the
relatively poor high -frequency video response of the restorer input,
since high -frequency response is not needed in this circuit. Thus, in
these receivers, the interference considered here has no effect on the
black -level setting and the observed effects will be substantially as
indicated on Figure 1, i.e., the black appears grey and the brightness
of the white parts is reduced. Similarly, receivers with a limiter ahead
of the peak-operated type of direct current restorer, will operate very
much like the average -operated type.
III. THE
NEGATIVE PICTURE PRODUCED BY STRONG INTERFERENCE
In conducting experiments on the effect of c -w interference, it was
9bserved that a strong interfering signal gave rise to a picture of
reversed contrast, i.e., the dark portions of the original became the
light portions of the reproduction and vice versa. Although this phenomena had been observed by others, for example when a strong sound
carrier was tuned into the picture channel, it had usually been assumed
that overloading occurred, or some other unusual behavior was present.
However, the writer's experiments showed that the effect existed when
there was no overloading and, indeed, was a straightforward extension
of the contrast loss phenomenon described above. In fact, the experiments showed that, as the interfering carrier was increased, the picture
contrast steadily decreased until, at a definite point, the picture was
substantially "washed out." Further increase of interference gave a
negative picture of rather poor contrast and, finally, with interference
signals far in excess of the black -level picture signal the picture again
disappeared. Synchronization was well maintained throughout, with
little or no apparent effect due to the interfering c -w. The scanning
return lines are, of course, visible in such a negative picture since there
is no blanking.
Figure 1 (c) shows how a large interfering c-w can lead to a negative picture. When the interference approximately equals the picture
carrier during black transmission, the "beats" produced alternately
raise the received signal to double amplitude and reduce it to substantially zero amplitude. Thus, with a black transmitted picture, the viewing tube has excursions extending to full white, leading to an average
brightness well up in the grey region. On the other hand, during white
transmission, the transmitted carrier is not present and no "beats"
occur. The interfering c -w simply replaces the normal picture carrier
and makes the picture appear black. To summarize this, the black
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION
301
transmissions now appear grey and the white transmissions appear
black, leading to a complete reversal of contrast (i.e., a negative
gamma) .
It was here assumed that the direct current restorer of the receiver
is unaffected by the interference and, as already indicated, this is typical of the many receiver circuits in which either direct current restoring
follows the average of the high video- frequency beat or in which
limiters are uscd. A direct current restorer whose input contained all
video components and whose output followed peak amplitude would not
lead to a complete washout of the picture or a negative picture, although
PEAKING COILS
VIDEO
KINESCOPE
/ 7- INPT
ELECTRODES
AMPLIFIER
+100
VIDEO
BRIGHTNESS
CONTROL
SIGNAL
INPUT
+300
D
-C RESTORER
-- PEAKING COILS
KINESCOPE
INPUT
VIDEO AMP. AND
D-C RESTORER
E
+300
VIDEO
I
SIGNAL
I
N
LECTRODES
PUT
+
BRIGHTNESS
CONTROL
300
-D -C RESTORING
IS
DUE TO GRID CURRENT
OF VIDEO AMPLIFIER
Fig. 2 -(A) A direct current restoring circuit which does not follow high
video frequencies and so permits strong interference to produce a negative
picture; (B) a direct current restoring circuit which operates on the peaks
of a synchronizing wave and so does not give a negative picture (unless a
limiter precedes the circuit)
.
the loss in contrast is more serious than with the averaging type of
direct current restorer. In this connection, two typical direct current
restorer circuits are shown in Figure 2. One of these, Figure 2 (A) is
of the averaging type and can give rise to a negative picture while the
other, Figure 2 (B), is of the peak type and cannot (i.e., unless preceded
,
www.americanradiohistory.com
302
TELEVISION, Volume IV
by a limiter). In a later section of this paper calculations will be made
which show the contrast loss of receivers with each type of restorer as
a function of interfering amplitude.
IV. CONTRAST EVALUATION
To evaluate quantitatively the effect of interference on contrast, it
is necessary to consider the various ways in which the contrast of a
reproduced picture may be expressed. The simplest expression for the
contrast is simply the over -all maximum brightness ratio, i.e., the ratio
of the light from the brightest portion of the reproduced picture to the
light from the darkest portion. In a perfect system, this ratio can be
infinite since the darkest portion can be completely black. Maximum
contrast ratio has been used to discuss kinescope performance' although
it is then necessary to distinguish between halation effects and normal
large -area contrast ratio. It is often stated' that a ratio of 35:1 or
more is desirable in a television picture and such ratios are attainable
with the best reproducing systems. We shall use degradation in maximum contrast ratio as one criterion for estimating the effect of c -w
interference.
In photography, it has long been well known that, even if the maximum contrast ratio is fixed, startling changes in appearance are made
possible by difference in the contrast gradation, i.e., the way in which
various brightness values of an original are interpreted in the reproduction. The same is, of course, true in television. If a curve is drawn
of reproduced light values as a function of original light values, complete information on contrast gradation is shown. Furthermore, such
a curve may also indicate maximum contrast ratio by the ratio of the
maximum to minimum light value at the ends of the curve. The over -all
contrast gradation curve is, therefore, an even more significant measure of the degradation caused by an interfering signal; such curves
will also be used in this paper.
Practically, whether or not an interfering signal is noticed depends
upon the quality of the over -all system when free from interference.
There are many grounds for believing that future television pictures
will be far superior in their contrast range and low -light tone renditions than those which are presently called high in quality. In considering interference, therefore, it is well to concentrate on the effect
which is obtained when the received picture is more nearly ideal, since
this will have most value for the future. In this respect, an objective
6 R. R. Law, "Contrast
in Kinescopes," Proc. I.R.E., Vol. 27, pp. 511 -524,
August, 1939.
7 P. C.
Goldmark and J. N. Dyer, "Quality in Television Pictures ",
Jour. Soc. Mot. Pic. Eng., Vol. 35, pp. 234 -253, September, 1940.
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION
30:;
study is at present more valuable than a subjective one made with less than -ideal viewing tubes, etc.
V.
COMPUTED EFFECTS OF INTERFERENCE
This section is concerned with the computation of the over -all con-
trast gradation curve when interference is present, and assuming an
idealized kinescope. The video wave which results from envelope
detection of a television signal which includes c -w interference is
derivable as follows. If we call the picture carrier, as it arrives at the
second detector, A sin wt and the interfering carrier is B sin (w + p) t,
then the second detector receives an over -all signal of
A
sin
wt
+ B sin
(w
+ p) t = [ V A2 + B2 + 2AB cos pt] sin
(wt
+ ß)
where ß is a time -variable phase angle which is of no concern here.
After detection only the envelope is of interest. It may be written
Ve= VA'
+B'+2AB cos pt=
(A
+ B)
1
pt
4AB
sin' - (A+B)2
2
=
(A + B) V1
-
Ic'
sin '
(,;>
(1)
where
L°'
=
4AB
(A+B)2
and
ci5
_
pt
2
The picture carrier amplitude, A, has a maximum value, Amas,
during the synchronizing interval, and a value 3/4 Amas at the black
level.* With maximum brightness of the original picture, A is reduced
to zero when the modulation is complete (100%). For intermediate
brightness, and a constant transmitter gamma, A follows the relation
A=3/4Amaæ
E1
(2)
Lmnx
J
where LT. is the instantaneous original picture brightness, L,na.c is the
maximum brightness and Yr is the transmitter "gamma," or slope of
the modulation characteristic when corrected for negative polarity of
modulation and plotted on log -log paper.
*
According to U. S. television standards.
www.americanradiohistory.com
TELEVISION, Volume IV
304
A sufficiently close approximation to an idealized kinescope characteristic is a power law over the range from cut -off to zero bias. We
shall assume that the output light is constant for inputs beyond zero
bias and zero beyond cut-off. Thus, the kinescope characteristic is
represented by Figure 3. Mathematically, the light output is
L= K,V,.+V)ry' when O<
L=0 when (V,., +
L
=K Vm yR
when
(V,.,
V)
+V) <V,,,
<O
(3)
(V + V) > V
V
is the magnitude of the voltage needed to cut off the tube, V
where
is the applied bias and signal, and yH is the exponent of the power law.
BIAS,
V
Yco
Fig.
3- Characteristic
of an assumed power-law kinescope following the
equation L = K (V.. + V) yR
The manner with which the video signal is applied to the kinescope is
shown in Figure 4. In Figure 4 (a) is shown the normal, interferencefree case. It is seen that the receiver gain control is so adjusted that
equals the
the range of black to white video signal, which is 3/4
assumed kinescope cut-off V,,. Furthermore, the direct current restorer,
which operates from the synchronizing pulse, together with the kinescope bias control comprise a net bias, Vd., which sets the black level at
A,
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION
305
cut -off. The difference between Amax and V, 0 is shown as 1/4 A,, on the
figure and is adjusted to this value by the bias control (often called the
brightness control). The instantaneous video signal is shown as V,.
(equation 1).
Figure 4 (b) shows the video signal when a high beat -frequency
interference is applied and the direct current restorer is of the type
shown in Figure 2 (a), i.e., it operates on the average of the synchronizing video pulse during the fluctuating beat. From equation (1) we
find the average to be
2
Va
(Ama.a+B)a
/2
1-k"- sin'
N/
d¢
(4)
0
which can be evaluated for various values of B by the usual tables for
the complete elliptic integral.'
4 A MAX.
T
A
r\
MAX
KINESCOPE
CHARACTERISTIC
1
Vd
LIGHT OUTPUT
/-
,I I
4
A MAX.
KINESCOPE
CHARACTERISTIC
1111111111
L
--
I
14_.I
LIGHT OUTPUT
-^
AMAX.
KINESCOPE
CHARACTE RISTIC
LIGHT OUT PUT-W
Fig.
-This
figure shows the video signal placement on the kinescope
characteristic: (a) with no interference; (b) with high video beat interference and an average- operated direct current restorer; (c) with high video
beat interference and a peak- operated direct current restorer.
4
Figure 4 (c) shows the case when a peak- operated d -c restorer is
used without a limiter, such as the one of Figure 2 (b). In this figure,
the direct current restorer operates from the peak signal during synchronizing, amas + B. When a limiter is used, conditions will be sub8 B. O. Peirce, A SHORT TABLE OF INTEGRALS,
p. 121, Third Edition, Ginn and Co.
www.americanradiohistory.com
TELEVISION, Volume IV
306
stantially the same as in Figure 4 (b) since the increased peak values
are clipped by the limiter. It is, of course, clear that a manual change
of the kinescope brightness control can change the effect of one of the
direct current restorers to that of the other. In this analysis, the controls will be assumed to remain at their normal setting when no interference is present.
Putting equation (1) in equation (3) and including the effect of
d -c restorer and kinescope bias, V
(Figure 4) we see
L =K
(Vd_a
=K
[Vd_a
- (A + B) V 17 0)7R
(5)
sin= CYR
1
where we must remember the limitations imposed on the equation by
the cut -off and zero bias points (see equation 3). These limitations
are that
[Vd_a
- (A
-I-
B) V
1
-
k2
sin2 0]
>
<
-
O
and
[Vd_a
- (A
-f-
B) V
1
-
k2
sin2 0]
3
Ama
4
Each of these limiting conditions may be solved for a, value of 0 which
will be needed as an integration limit when finding the average light
output. Calling these sb1 and 02 respectively, we find
(V4 -c)2
Y'2
=
(6)
4AB
k2
(Vd -c
Sin-1
-34
Amax)
2
(7)
4AB
These angles have limiting values of 0 and 7r/2 respectively and these
limits are used when the arguments of (6) and (7) are greater than
unity or imaginary.
The average light output over the fluctuating beats is then
L
(1- 02/K4Amaan+K
/
\
.Vd-a
i
V
1
-
www.americanradiohistory.com
- (A + B)
k2
sin2
R
d 4,
(8)
LOCAL OSCILLATOR RADIATION
307
where the first term gives the light output when the instantaneous bias
on the kinescope exceeds zero, and the integral gives the total light
output averaged over the normal kinescope range. The integration is
1 and yR
2 although the result involves
straightforward for
the incomplete elliptic integral. Since tables for these are available',
a numerical answer may be obtained, although the calculations are very
laborious.
The square law relation, y,i = 2, is a far better approximation to
an actual kinescope than the linear one. The writer has carried through
the calculation of equation (8) to find the over -all contrast gradation
curves for different interference levels, using yR = 2 and assuming, in
turn, each of the two types of direct current restorer which give
y=
.
where
Vc
=V
=
-4
1
A,,,a1,
(average- operated type)
is found from equation (4), and
-1
Vd-r
=
_
(Auiar.
3
+ B)
Arnaa
4
Am, + B (peak -operated type)
4
The calculations were made by assuming a complementary gamma at
the transmitter of yT =1/2 (equation 2). The curves can be corrected
for other transmitter gammas by an appropriate compression or expansion of the abscissa scale.
Figure 5 shows the calculated reproduced light as a function of
original light at the transmitter, using the average- operated direct current restorer. The curves are largely self-explanatory and show the
marked decrease in contrast as the interference level is increased.
Because vestigial sideband operation was assumed, it should be remembered that the interfering c -w receives 6 decibels more gain in the
receiver than the desired picture carrier. Furthermore, with U. S. television standards, the black -level carrier is 2.5 decibels less than the peak
carrier which is used to rate transmitters and field strengths. Thus the
curve labeled "interference 8.5 decibels down" means an interfering c -w
whose antenna field strength is 8.5 decibels less than the peak of the
picture carrier field strength; at the second detector of the receiver,
because of the increased amplification for the interference, the inter9
H. Hancock, ELLIPTIC INTEGRALS, John Wiley and Sons, New
York.
www.americanradiohistory.com
TELEVISION, Volume IV
308
ference is just equal to the black -level picture carrier. From the point
of view of interference calculation, of course, it is the value at the
antenna which matters, so that the curves are significantly labeled.
The negative picture for the stronger interference levels is clearly
indicated by the reversed slope or negative gamma. One of the more
striking features shown by Figure 5, is the rapidity with which contrast is lost as the interference level reaches a point 15 decibels below
the picture carrier. Between the 14.5 decibels curve shown, which still
gives a positive picture, and the 8.5 decibels curve, which gives a nega-
INTERFEPENCE
aEµrE_a
INTERF_
.
yC/
¡J
GE
'
/
/
2,i
8.5 db DOWN
/,7
i
2
,DOT
--- -
``.
`.
¡
i
\ V/.
`
****.
Co
...-
A
P
O
0.001
0.01
0.1
LO
ORIGINAL PICTURE LIGHT VALUES
5 -Effect of c -w interference on television picture contrast using a
direct current restorer which follows the average of the synchronizing pulse.
Kinescope gamma = 2, transmitter gamma = %, and vestigial side -band
operation using U. S. standards.
Fig.
tive picture, the original contrast is substantially wiped out. In Figure
7 will be shown curves of maximum contrast ratio as a function of
interference level which show that this rapid loss of contrast holds for
other kinescope gammas as well.
Figure 6 shows a set of contrast gradation curves for the other type
of direct current restorer (such as that of Figure 2b). It is here found
that no reversal of the picture takes place at any interference level, as
had been expected. It should be remembered that the difference between
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION
309
the results of Figure 5 and those of Figure 6 lie only in the kinescope
bias provided by the direct current restorer. Thus a manual adjustment of the bias control (brightness control) will change either set of
curves into the other.
Since the gamma (y,t) of existing kinescopes is often in excess of
3
two, it is of interest to examine the contrast degradation for yR
in
are
given
and yR = 4. Over -all contrast gradation curves, such as
Figures 5 and 6 for the square -law kinescope, are extremely tedious to
compute for the higher -power laws. However, there is a simplification
Amax). If
in equation 8 when the interference level is small (B
=
«
1.0
/
---DOWN---_=_.- i
i
INTERFERENCE B.Sdb
ÏNTERFÉRENCE
dbÓ
2
if
CS
r/
0044/.
D/
--P/_..'''
\'F,P
N
'
i
i
/,'
OO
ab
..-
c.(c..
4.C
4.Q.
P4
,S4'
\v
O
0.001
0.01
0.1
io
ORIGINAL PICTURE LIGHT VALUES
Fig. 6 -Effect of c -w interference on television picture contrast using a
peak- operated direct current restorer and no limiter. Kinescope gamma = 2,
transmitter gamma = 1/z, vestigial side -band operation using U. S. standards.
only the end points are desired, i.e., the output light level for a completely black and a completely white transmission, the small- interference approximation is readily usable. Furthermore, for one case of
large interference, namely, when B =3/4 Amax, the end points can also
be simply evaluated for higher gammas. If only the light outputs for
white and also for black transmission are given, their ratio is the most
easily understood evaluation of contrast degradation and, as discussed
www.americanradiohistory.com
310
TELEVISION, Volume IV
in Part IV above, is called maximum contrast ratio. Figure 7 shows
curves of the contrast ratio as a function of interference level for
kinescope gammas of from 1 to 4, assuming the average- operated type
of direct current restorer.
Examining Figure 7, we notice that the interference level at which
the picture washes out (contrast ratio of unity) lies between -10 and
-13 decibels for all the kinescope power laws, and that the loss in contrast is quite rapid as this point is approached. If we assume a transmitter gamma, y2., which is the reciprocal of the kinescope gamma, yR,
the interference -free picture for each of the assumed kinescopes will be
the same. Figure 7 shows, however, that small interference has far less
effect on the higher-gamma kinescopes. This illustrates the well -known
10,000-
ìS
X4,
1000-
Sq_
'15R=
I
`
100
\\
\.
10
A
IS
NEGATIVE PICTURE
INDICATE D BELOW
THIS LINE
/
1
01
-50
-40
-30
-20
INTERFERENCE LEVEL
-10
,
0
db
Fig. 7 -Effect of c -w interference on ratio of light output during white transmission to light output during black transmission. An average- operated'
direct current restorer is assumed and the curves show the effect of different
kinescope gammas.
advantage of gamma compression at the transmitter, with corresponding expansion at the receiver, in improving the signal -to-noise ratio.
On the other hand, with larger interference, the curves eventually cross
and the advantage is no longer present.
VI. EXPERIMENTAL RESULTS AND ESTIMATED
TOLERABLE INTERFERENCE
A television receiver in the writer's home was used to obtain
experimental subjective data on the effect of interference. This receiver included the type of direct current restorer shown in Figure 2a
except for a modification originally made to reduce possible effects of
kinescope grid leakage. Normal program material and the test pattern
from the National Broadcasting Company's New York Station, WNBT,
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION
311
was used. A Ferris Microvolter was used as the calibrated source of c -w
interference and was connected across the receiver antenna transmission line (100 ohms impedance) through two 500 -ohm resistors, so as
not to interfere with the input alignment or impedance values. With no
television signal present, a response curve of the receiver was measured
and the diode second detector current calibrated in terms of the signal
generator voltage. In this way, when the television signal came on, its
relative magnitude with respect to the signal generator readings could
be determined by its second detector current. It was necessary, of
course, to use a substantially black picture to calibrate the black-level
carrier of the received signal. It was then assumed that the peak
carrier was 2.5 decibels higher, corresponding to U. S. standards. A
check on this relative calibration of the received television signal was
made by using a low-frequency "beat" interference and observing the
magnitude of the beats on the synchronizing pulses as viewed on an
oscillograph across the kinescope grid. The two methods checked very
well.
Although many observations were made using various interfering
frequencies, giving beats with the picture carrier from some tenths of
megacycles to around 4 megacycles, most attention was given to a beat
at 3.7 megacycles, well within the video band but at such a high frequency that the predominant effect was loss in contrast, rather than
the very fine -grained pattern. In fact the interference pattern could
hardly be observed with the kinescope and viewing distances used and
might even pass completely unnoticed if attention was not called to it.
The contrast changes, on the other hand, were very marked. In such
subjective tests, it is not possible to obtain accurate data as to loss of
contrast; fortunately, however, the transition point beween the positive
and negative picture was quite clearly defined since it led to an almost
complete wash -out of contrast values. The data are presented in
tabular form in Table I and represent an average over a number of
observations. In every case the receiver controls were set as for an
interference -free picture and were left untouched for the observation.
Table I
Interference
Beat Frequency
3.7 Mc
3.7 Mc
3.7 Mc
3.7 Mc
3.7 Mc
Interference
Level, decibels
-- 22
-greater than
28
16
10
5
Observed
Results
Barely perceptible loss in contrast
Substantial but tolerable loss in
contrast
Intolerable loss in contrast
Completely washed -out picture
Negative picture of poor contrast,
return lines visible
www.americanradiohistory.com
TELEVISION, Volume IV
312
From the computed data as presented in Figure 7 it is seen that a
completely washed -out picture (contrast ratio
1) was predicted with
an interference of -10 to -13 decibels, depending on the kinescope
gamma. Thus the computation and the observed value of Table I are in
=
close agreement.
A comparison with the results of Jarvis and Seaman,5 in England,
is of some interest, in spite of substantial differences between their
technique (which introduced the interference in the video channel
instead of at the antenna)
was the annoyance value of
When 8.5 decibels is added
with those used here for U.
shown in Table II.
and even though the criterion they used
the beat pattern, rather than contrast loss.
to their figures, to make them comparable
S. standards, the various results appear as
Table II
Interference Level
Using 3.7 Mc Beat
Observed Result
and 525 Line System
Just visible change
Just tolerable change
Seaman and Jarvis
Results With 2.0 Mc Beat
and 405 Line System
-28db
-22dó
-26db
-19db
On the basis of the calculations, as supported by the experiments,
it is clear that relatively small differences in interference level near
the critical point will cause rapid deterioration of the picture (see
Figure 7). There can be no denying that the permissible interference
limit must be below the point at which a complete wash -out of the
picture occurs. If the interference is 10 decibels below this limit, a reasonable safety factor is allowed, although a noticeable deterioration of
picture quality is still present. On these grounds we may say that the
type of interference here considered, i.e., in the high -video range,
should be at least 20 decibels below the picture carrier at the antenna.
This number will be used as a criterion in the discussion of local oscillator radiation below.
VII.
QUANTITATIVE LIMITS ON RECEIVER LOCAL
OSCILLATOR RADIATION
Since the superheterodyne receiver is here acting as an interfering
transmitter, it is logical to measure its radiation in terms of the power
which the local oscillator delivers to the antenna. With long transmission lines having appreciable loss, the radiated antenna power may
be appreciably less than would be measured on a bench test with a
receiver connected directly to a calibrated measuring receiver. Although this loss should be considered in special cases, the more general
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION
313
approach should assume no loss in the connection to the antenna, since
negligible loss is readily obtained by use of good lines.
The type of interference considered here will be most serious between nearby antennas; thus complications introduced by propagation
phenomena need not be considered. It can be assumed that "free-space"
propagation will occur.* Thus a receiver radiating W watts into a
half -wave dipole will give a field strength at a distance, d, of
E --
V45W
d
..
10,000
FEET FROM INTERFER'G
RECEIVER
100 50
500
INIINIFAML.LIMIIN
1000
ßïíí=_
MIMFAIMMIAlir
100
on
10
RPM
-
.01
001
10
lo
102
3
IÓ
lob
IÓ
FIELD STRENGTH OF DESIRED STATION, µV/m.
8-
Fig.
Permissible radiation from television receiver to give an interference level of -20 decibels. A half -wave dipole is assumed on the receiver
causing the interference.
If we consider a nearby receiver, with an antenna so situated as to
receive both the interfering radiation and a desired signal from a
television transmitter whose field strength is Ea, then, using the 20
decibels criterion,
Ea
Ead
Ei
V45>!V
--
> 10
Thus the radiated receiver power should be
This equation has been plotted in Figure
8
E82d'
(9)
W
4500
and shows that, to protect
* The limiting distance for antennas 30 feet high, within which free
space propagation may be achieved (on the average) is some 700 feet at 60
megacycles and greater than this at higher frequencies.
www.americanradiohistory.com
TELEVISION, Volume IV
314
the 500 -microvolt /meter area, the radiating receiver should radiate
less than 0.01 microwatts if receivers are to be as close as 50 feet.
This may be contrasted with prewar receivers using a 6AC7 mixer
and no radio frequency stage which radiated in the order of 1,000 microwatts; to avoid annoyance to other receivers at a distance of 50 feet, it
is necessary to have a signal of over 100,000 microvolts /meter. Fortunately, channel assignments, station hours, and the small number of
existing receivers have been such as not to bring this problem into
prominence except under special circumstances.* However, this situation cannot continue and measures should be taken to reduce radiation
on all future receivers.
VIII.
RECEIVER DESIGN CONSIDERATIONS
It
is, of course, one thing to suggest 10 -8 watts as a maximum
permissible radiation and another to achieve it. A carefully designed
PE NTODE
MIXER
ANTENNA
RADIATION
RESISTANCE
T
C2
eo
TO
LOCAL
OSCIL
9-
Fig,
Typical pentode mixer with double -tuned input circuit. The required
local oscillator voltage is indicated as eLo, whereas the resulting antenna
voltage is shown as e,.
pentode radio frequency stage between the antenna and a pentode
mixer, together with a reasonably high intermediate frequency (so as
to tune the oscillator far away from the band-pass of the amplifier), can
be expected to provide enough attenuation.
To consider further, let us calculate the local oscillator radiation
to be expected from the simplest television receiver with good performance, which uses a double -tuned circuit to couple the antenna to a
pentode mixer. This type of input was commonly used in prewar
receivers and is shown in simplified form in Figure 9. The local
oscillator injection must be sufficient to give good results with the
mixer, so that the local oscillator voltage across the input circuit, ezo
in the figure is fixed. Thus it is clear that the selectivity of the sec* In the New York area, as already mentioned, prewar receivers tuned
to WNBT (50 -56 megacycles) radiated at 64 megacycles in the WCBW
band (60-66 megacycles).
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION
315
ondary of the double -tuned circuit is not effective in reducing the
radiation. It can be inferred, on the same grounds, that a single -tuned
circuit would have no selective effect at all on the local -oscillator radiation. In either case, however, the secondary capacitance does affect the
radiation since its value determines the antenna -to -grid step-up for a
given band width. With the double -tuned circuit, it can be shown that,
to a fair approximation, the primary voltage, el of Figure 9, induced
by the local oscillator, is
eLO2 C2
1
e12
2
C1
1
a
+2
(10)
Po)
where (ai_t is the angular, mid -band intermediate frequency, and poi is
the 3 decibels down, angular, input circuit band width. This approximation assumes adjustment for a flat -top response curve and a high
secondary Q (i.e., the damping is provided entirely by the antenna).
Under these conditions the primary Q determines the band width according to the relation
Q1=o)C1Ra=
Solving this for C1 and putting it into the previous expression, it is
seen that the radiated watts are
e12
W
R
praC2
eLO2
/
2
1+2
(-
z
A.,
which is conveniently independent of antenna radiation resistance.
Assuming an 8- megacycle circuit band òidth, which is suitable for
the 6-megacycle channel of present U. S. transmissions, and an oscillator excitation of 2 volts, which is satisfactory for the high- transconductance pentodes, Table.III was calculated for two currently available
tubes.
Table
Intermediate
Frequency
10 Mc
20 Mc
30 Mc
100 Mc
III
Microwatts Radiated
6AC7
680
210
96
9
www.americanradiohistory.com
6AK5
340
105
48
5
316
TELEVISION, Volume IV
Even with an intermediate frequency of 100 megacycles, the local oscillator radiation is far from the 0.01 microwatts desired.
A grounded -grid triode radio frequency stage, such as the one recently described10 using a 6J4 or 6J6 tube, requires a slightly different
approach. A double -tuned circuit between the triode and the pentode
mixer will give slightly less radiation than a single -tuned circuit. With
the double -tuned circuit, the local oscillator voltage at the triode plate
(a loading resistor for primary damping is assumed) is given by equation (10) above. This voltage reacts on the input antenna circuit only
through the plate-cathode capacitance and through the plate resistance,
if we may assume good grounding of the grid. Assuming an antenna
directly connected to the cathode-grid input circuit, we find an antenna
radiation around 30 decibels less than the figures in Table III. Although
this is a substantial improvement, it may be insufficient unless a high
intermediate frequency is chosen. A pentode radio frequency stage, on
the other hand, will have much less output -to-input coupling and niáy
have an additional pair of selective circuits In its input. This should
allow such a radio frequency stage to reduce oscillator radiation sufficiently to give satisfactory performance even for lower intermediate
frequencies. The use of a mixer with less inherent radiation, such as
the cathode -coupled, double- triode mixer (see Figure 13 of reference
10), will decrease the isolation requirements on the radio frequency
stage.
Still other arrangements which are possible make use of balanced
triode or pentode mixers with the local oscillator driving the tubes in
parallel, while the antenna signal drives them in push pull. These
circuits must be carefully balanced to give effective reduction of radiation. If combined with a grounded -grid triode radio frequency stage,
however, it may be possible to attain the desired 50 decibels or so of
radiation reduction. Neutralization of the radiation from a single
mixer tube is possible but again, to be effective, is achieved by a rather
critical adjustment. Partiollar methods of operating balanced mixers
will give conversion using an oscillator of half of normal frequency.
This places the oscillator so far away from the normal signal channel
that, when combined with the neutralization dne to a balanced oscillator
feed and unbalanced signal feed, adequate reduction of radiation may
be achieved.
The one solution, which unfortunately cannot be proposed with
those tubes which are commercially available on the open market, is
the separation of oscillator and signal circuits of the mixer by the
10 G. C. Sziklai and A. C. Schroeder, "CathodeCoupled Wide -Band
Amplifier," Proc. I. R. E., Vol. 33, pp. 701 -709, October, 1945.
www.americanradiohistory.com
LOCAL OSCILLATOR RADIATION
317
methods used in such low- frequency tubes as the 6L7.11 The signal -tonoise ratio of this illustrative type of mixer has not been adequate for
television- service. Thus, if a radio frequency stage cannot be used, the
only practicable remedy is the use of additional selective circuits
between antenna and mixer, possibly with a rejection band at local
oscillator frequency.
In all cases, local oscillator shielding should be employed to prevent
direct chassis radiation.
IX. CONCLUSIONS
The tolerable amount of local oscillator interference, or other c -w
interference, in a television picture is greater when the interfering
frequency is at the high end of the picture band. In this case, the
chief annoyance is loss of picture contrast which, for strong interference, can be very bad, even to the point of a negative picture. The
transition between a slight loss in contrast and a completely washed -out
picture is s.ufficiently sharp to make a choice of minimum interference
level not too difficult. A value of signal -to- interference field strength
ratio of 20 decibels may be considered a satisfactory minimum when the
interference is near the upper end of the picture band.
On the basis of an interference 20 decibels below a desired carrier, and assuming channel assignments and choice of intermediate frequency so that an interference does take place with another channel, it
is found necessary to reduce receiver radiation to 0.01 microwatts to
protect the 500 microvolt per meter field strength contour with receivers
50 feet apart. Prewar receivers radiated 105 times as much as this and
so were extremely unsatisfactory. A grounded -grid triode radio frequency stage is not a sufficient safeguard, though a carefully designed
pentode radio frequency stage may be. Other alternatives lie in the use
of balanced or radiation -neutralized mixers, or additional selectivity
with an oscillator rejection circuit between antenna and mixer.
Although none of the suggested remedies lend themselves to a
minimum -cost receiver design, an adequate television service will require substantial suppression of local oscillator radiation if frequency
allocations are such as to make interference possible.
11 C. F. Nesslage, E. W. Herold and W. A. Harris, "A New Tube for
Use in Superheterodyne Frequency Conversion Systems," Proc. I. R. E., Vol.
24, pp. 207 -218, February, 1936.
www.americanradiohistory.com
.
DEVELOPMENT OF AN ULTRA LOW LOSS TRANSMISSION LINE FOR TELEVISION'
BY
E. O. JOHNSON
Engineering Products Department, RCA Victor Division;
Camden. N. J.
Summary-The development of a low loss 300 -ohm parallel wire polyethylene dielectric transmission line is described. Loss curves, as well as a
photograph of a production run sample of the line, are included.
INTRODUCTION
VJ
RANSMISSION lines for use on home television receiver installations have been very unsatisfactory to date primarily
because of their very high losses. In addition to the high signal
attenuation these lines have had many other undesirable characteristics.
Twisted Pair Lines
Transmission line losses in twisted pair lines have been so great
that installation men have recommended installing television receivers
on top floors in buildings so as to shorten the transmission line and
thereby improve the signal intensity by eliminating as much of the
transmission line loss as possible. It was not uncommon to find a
receiver installation within a mile of the transmitting antenna that did
not have sufficient signal to override the local noise level. Two such
installations recently investigated had transmission lines 600 and 400
feet in length and attenuations of 400 to 1 and 275 to 1 respectively.
Each of these installations was less than a mile from the transmitting
antenna and there was not enough signal to operate the receivers satisfactorily due to the long length of high loss transmission line used.
Typical line losses varied from 5 to 12 decibels per 100 feet at 50 megacycles.
The manufacture of transmission lines has involved many different
operations with a resultant high cost. Typical construction is as
follows:
(a) Small gauge copper wire is tinned so as to prevent corrosion
from the various compounds in the wire covering.
Decimal Classification: R320.41 X R583.
Reprinted from RCA REVIEW, June, 1946.
318
www.americanradiohistory.com
TRANSMISSION LINE
319
(b) The small tinned copper conductors are twisted forming a flexible stranded conductor.
(c) The stranded conductor is rubber covered.
(d) Two of the stranded rubber covered conductors are twisted
together.
(e) The twisted pair is rubber covered.
lines this operation is omitted.)
(On the less expensive
(f) A cotton braid is placed over the line to prevent the rubber
from deteriorating in the sunlight.
(g) The line is given an impregnating dip in an asphaltic compound for weatherproofing.
Coaxial Lines
Coaxial cables and twin coaxial cables having medium loss characteristics have been available for some time. However, the cost of these
lines has been so high as to prohibit their general use on home television receiver installations. One example is a twin coaxial cable having
a loss of 1.4 decibels per 100 feet at 50 megacycles and selling for $1.25
per foot. The average home receiver installation requires 70 feet of
transmission line, thus making a total cost of $87.50 for the transmission line alone. Therefore, practically all of the home receiver installations have been made with one of the twisted pair lines.
DEVELOPMENTAL SPECIFICATIONS OF LINE
The developmental problem was to produce a line that did not have
the undesirable characteristics previously enumerated. The following
developmental specifications were established.
(a) Low Loss -Loss should be less than any twisted pair of coaxial
transmission line available in the pre -war period for the receiver installation. 1 decibel per hundred feet at 50 megacycles
was set as the goal.
(b) Low Cost-The manufacturing cost of the line should be very
low to permit the installation of both the receiver and antenna
in the most desirable locations. A goal of six cents a foot list
price was established.
(c) Weather Resistance -The average life of transmission lines
used in the pre -war period was very low. In a few months the
cotton braid failed and the sunlight hardened the rubber covering which cracked and permitted the absorption of moisture
www.americanradiohistory.com
TELEVISION, Volume IV
320
with a resultant increase in line losses: A minimum life of five
years was desired for the new line.
(d) Deterioration Due to Heat -The asphaltic impregnating compounds used on the cotton braided lines softened and came off
on the hands, clothing, woodwork, furniture and rugs in hot
weather. The elimination of this undesirable characteristic
was of primary importance.
(e) Flexibility at Low Temperatures -The asphaltic impregnating
compounds used on the cotton braided lines hardened in cold
weather and movement on the line during installation, or by
the wind after installation, caused the line to crack and break
the cotton fibers in the braid. This soon resulted in failure of
the line insulation. The new line should be flexible under all
temperature conditions.
DEVELOPMENT OF THE LINE
Work was begun to develop a transmission line meeting the specifications outlined above. A thorough check was made on all available
transmission lines to determine what might be done to reduce their
electrical loss.
With a parallel wire type of transmission line the loss varies inversely with impedance (See Figure 1). If such a transmission line
has an impedance of 72 ohms and a loss of 6 decibels per hundred feet
at 50 megacycles and the wires are separated far enough to produce an
impedance of 144 ohms, the loss will decrease to 3 decibels per hundred
feet at 50 megacycles or one half of its original value. It was thought
that a high impedance transmission line of the conventional twisted
pair type could be produced which would meet the specifications. How
ever, it was found that the maximum improvement that could be
obtained by increase in impedance was about 2:1.
.
(1) Dielectric
Along with the work on conventional transmission lines, one manufacturer developed a parallel wire transmission line with a spun glass
woven web. This line had excellent low loss characteristics which met
the loss specifications. However, the line was very hygroscopic and
required impregnation. Eventually several good weather -proofing compounds were found which did not increase the line loss appreciably. The
cost of this line however did not meet the tentative specifications.
Prior to the war, research work was done by this company on some
relatively high impedance parallel wire transmission lines insulated
www.americanradiohistory.com
TRANSMISSION LINE
321
with polystyrene. These lines had desirable electrical characteristics,
but were unsatisfactory mechanically due to the brittleness of the
polystyrene.
Polyethylene, suitable for transmission line insulation, was developed during the war for use at ultra- high -frequencies. This dielectric,
while expensive, has very excellent electrical and mechanical properties.
A full description of this material is beyond the scope of this paper
and the reader is referred to one of the excellent papers on poly ethylene." 2. 3 Polyethylene has a power factor of approximately .0003
A =LOSS DB PER
I
874T
O
F
IN MEGACYCLES
ZIMPEDANCE
OF
D DIAMETER
'
®
100 FEET
F FREOUENCY
WM
...
.'' /,/
IMOir-D
0
20
LINE IN OHMS
OF CONDUCTORS IN INCHES
.,,,',
--'.,,/
N::
,600
40
60
BO
100
140
200
FREQUENCY IN MEGACYCLES
Fig.
1- Computed loss of
open wire transmission lines
using No. 20, A.W.G. wires.
at frequencies as high as 1000 megacycles and a dielectric constant of
about 2.29. It is a very strong, tough and flexible material that is not
affected by acids, alkalis, ozone, sunlight and water. These properties
M. C. Crafton, Jr. and N. B. Slade, "A New Dielectric For Cables"
Modern Plastics, Vol. 21, No. 11, pp. 90 -93, pp. 168 -170, July, 1944.
Floats ", Modern Industry, Vol. 9, No. 1,
"Polyethylene Plastic
pp. 45 and 137 -140, January 15 1945.
3 "War Time Trends in Insulated Wire and Cable ", Publication C -56,
Anaconda Wire & Cable Company, 1944.
1
2
-It
www.americanradiohistory.com
3i?
TELEVISION, Volume IV
make it an outstanding dielectric for use at high- and ultra-high -frequencies. Several experimental lines were made of both twisted pair
and parallel wire types. Measurements and field tests on these samples
indicated a line could be constructed using polyethylene as the dielectric that would meet the specifications.
(2) Line Conductors
The conductor should be stranded to give the required flexibility
and prevent breakage during use. The choice of conductor size is a
compromise between mechanical strength, line loss and cost. Past
experience on transmission lines proved that seven strands of No. 28
AWG would meet the structural requirements. A conductor of this size
was also found to be satisfactory from the standpoint of line loss and
cost. In the past transmission lines have had the copper conductors
tinned because the bare copper wire was attacked by the various compounds in the insulating materials. Polyethylene is very inert and contains nothing that will react on copper, therefore bare copper conductors can be used. This is a fortunate condition because a lower loss
line is obtained at reduced cost. At 100 megacycles the radio frequency
currents are all on the outer surface of the conductors. As a matter of
fact, the skin depth at this frequency is only .00067 inches. This means
that with tinned wire most of the current is flowing in the tin surface
layer, and since tin is a poorer conductor than copper the line loss is
increased. The reduction in cost is an important item since the cost
of the tin is saved as well as the expense of the tinning operation.
(3) Line Impedance
The development of the transmission line departs radically from
past practices in connection with its surge impedance. Most of the
transmission lines used in conjunction with television receivers have
had surge impedances of between 70 and 125 ohms. A resonant dipole
in free space has an impedance of approximately 72 ohms, therefore a
72-ohm line gives the best transfer of power when it is desired to
receive signals on but one frequency. The problem of receiving television signals on a number of television bands is an entirely different
problem. If a one-half wave dipole is designed to be resonant at 50
megacycles and used as an antenna, it will have an impedance of approximately 72 ohms at 50 megacycles and an impedance of about 2000 ohms
at 100 megacycles. If a reflector is used in connection with this dipole
the antenna's impedance can be as low as 200 ohms. Therefore the
antenna's impedance may vary from some 20 ohms to 2000 ohms.
www.americanradiohistory.com
TRANSMISSION LINE
323
If a fixed antenna is to be used to cover a two -to -one or greater
frequency range, it is desirable to use a line having an impedance such
as to provide a maximum amount of energy over the desired frequency
band. The line impedance would then be something less than one half
the difference between the lowest and highest value of impedance,
probably near 600 ohms. The actual value of optimum impedance is a
complex affair and dependent upon many things such as the frequency
response of the antenna, type of antenna load, impedance of antenna
load, loss characteristics of the trasmission line, and other factors. It
is sufficient to say here that the value of impedance would in all cases
be many times higher than 72 ohms or the impedance of transmission
lines used in the past. From an electrical viewpoint the line impedance
should be high and probably between 300 to 600 ohms with the higher
value of impedance favored, because the line loss is inversely proportional to its impedance for any given set of conditions.
There are, however, other considerations which have a bearing on
line impedance. The line should be of such a size that standard hardware equipment can be used for the installation of the line. Standard
bakelite screw eyes have a 9/16" hole. This type of screw eye has been
produced by various manufacturers for years and represents a standard
transmission line support. The outside dimensions of the line should
not be greater than 9/16" if standard hardware equipment is to be
used. With seven strands No. 28 AWG conductors this will limit the
line impedance to a maximum value of about 400 ohms.
The amount of polyethylene used in a web line construction is approximately proportional to the square of the conductor spacing. If
the web spacing is doubled the thickness of the web must also be
doubled to maintain good mechanical design and have the web of sufficient thickness that the line cannot be crushed by the hands during
installation and use. The use of a minimum amount of polyethylene
favors lower line impedance.
The line impedance therefore should be approximately 300 to 400
ohms, with the cost favoring the lower value and the line losses favoring the higher value. A folded dipole antenna has an impedance of 288
ohms and was a deciding factor in choosing a line impedance of 300
ohms as the best value to give an ultra low loss transmission line for
a minimum cost. The folded dipole is useful in receiving signals in a
relatively narrow frequency .band and provides a higher signal level
than is obtained from simple wide band antennas.
The 300 -ohm line used in connection with a half wave dipole gives
a broad frequency response and permits multi -channel reception with.
out cutting the antenna elements or changing their spacing as has been
www.americanradiohistory.com
TELEVISION, Volume IV
324
2.0
1.8
1.6
1.4
1.2
1.0
W
W
o .8
O
K
6
m
o
.6
z
.2
20
40
60
FREQUENCY IN
Fig.
2
-Loss characteristics
Fig.
80
100
140
200
MEGACYCLES
of the new television transmission line.
3- Sample of production run transmission line.
www.americanradiohistory.com
TRANSMISSION LINE
325
required in the past. Figure 2 shows the loss characteristics; Figure 3
is a photograph of a sample of the production run transmission line.
Figure 4 (Curve A) shows the relative response with frequency of
a 44- megacycle half -wave dipole and reflector in conjunction with the
300 -ohm line. Curves Cl to C6 give the relative response frequency
characteristics of folded dipoles adjusted for each of the first six television channels and used with the 300 -ohm line.
A
OWE HALF WAVE DIPOLE
AND REFLECTOR WITH
300
140
C-I
OHM
LINE
C-I TO C-6 ONE HALF WAVE
FOLDED DIPOLES AND
REFLECTORS WITH 300
OHM
LINE
120
w
In
z
0
a
0
100
w
C-2
cr
C-3
BO
C-4
60
40
20
40
50
Fig.
4- Television
70
60
FREQUENCY
IN
80
90
MEGACYCLES
antenna characteristics using the new
300 ohm transmission line.
FIELD TESTS
quantity of the 300 -ohm transmission line was made on a developmental basis and installed, for field test purposes, in forty -seven test
locations in the New York and Philadelphia areas prior to April, 1945.
Some of the lines have been in service for over two years. Loss measurements on these lines show they have not changed by any measurable
A
amount.
www.americanradiohistory.com
326
TELEVISION, Volume IV
CONCLUSIONS
This developmental project resulted in an ultra low loss transmission line of unusual characteristics. It more than meets the specification requirements set forth. The line has a loss of less than 0.8 decibel
per hundred feet at 50 megacycles. Polyethylene is a very strong,
tough flexible material which is not affected by acids, alkalis, ozone,
sunlight or water. This produces a line which does not crack during
cold weather or soften during hot weather, and which give long trouble free service. The line can be used with folded dipoles or dipole antennas giving high gain single channel or medium gain multi -channel
reception respectively.
www.americanradiohistory.com
AN EXPERIMENTAL COLOR TELEVISION SYSTEM *t
BY
R. D. KELL, G. L. FREDENDALL, A. C. SCHROEDER, R. C. WEBB
Research Department, RCA Laboratories Division,
Princeton, N. J.
Summary -A description is given of color television apparatus using an
image orthicon in the color camera for direct pnckup of studio scenes. A
sequential three -color semi -mechanical system is used. Provision is made
for demonstration of color pictures in three dimensions. The associated
sound channel is transmitted on the edge of the picture during a portion of
the horizontal blanking period.
INTRODUCTION
and early in 1946, a series of demonstrations were
given of television in color and of television images in three
dimensions, also in color. These demonstrations were conducted
to show the status and to point out the problems remaining to be solved
before 'color television could be considered ready for development as a
service." The remaining problems are such as to require much additional research and development. However, this article is not concerned
with these aspects of the situation but rather with a description of the
system and apparatus used during the demonstrations.
Work on broadcast television was interrupted by the war, but
advances in electronic and radio techniques during the war period did
have a direct influence on television, particularly monochrome television.
In order to resume the studies of color television and to evaluate these
advances as they applied to television in color, laboratory facilities
for research on the various problems involved in the generation, transmission and reception of television images in color together with new
studio facilities, new circuits and apparatus were developed and put
in operation.
A TE in 1945
L
CAMERA STUDIO SETTING
A small studio set was constructed in the laboratory to make possible small scale productions of colorful program material. The set
is shown in Figure 1. Illumination is obtained from an overhead bank
of 36 100 -watt fluorescent lamps which provide an incident light of
about 200 foot -candles. Two auxiliary banks of 24 100 -watt lamps
Decimal Classification: R583
from RCA REVIEW, June, 1946.
t Reprinted
327
www.americanradiohistory.com
TELEVISION, Volume IV
328
Fig.
CAMERA STUDIO
1- Camera
Studio Set.
CONTROL ROOM
VIEWING ROOM
SYNCHRONIZ !NG GEN
'-'--
_
_
NEL
NNI
L
o ME N
EO U
Tr
1
NT
cSTN
RSERT
ED
NERE
II
T
STUDIO MICROPHONE
_JI
VIDEO AMP
THIS TPRU CONNEC
T ION REMOV F
w
SOUND DUPLE
EQUIPMENT IS USED
Fig.
2
-Block Diagram
ORUI
REDE
ER
LINEILMIXING
AMP
CONTROL ROO
O
1t:M-`:
SOUND
LjIDEMOOULATOq
LOUD SPEAKER
of Color Television System.
www.americanradiohistory.com
EXPERIMENTAL COLOR TELEVISION
329
each can be moved about as desired in front of the scene to bring the
combined illumination up to more than 400 foot -candles. In order to
obtain a more uniform light spectrum, half of the lamps are of the
white type and half are of the daylight type. By distributing the lamps
uniformly on the 3 -phase 60 -cycle power supply, no difficulty is experienced due to the lights operating on alternating current power.
CIRCUITS AND APPARATUS
A complete set of experimental equipment was constructed as
shown in the block diagram of Figure 2. The system employs the
latest and most suitable devices and circuits which have resulted from
many years of extensive research in the field of electronic monochrome
television. Added to these electronic components are two mechanically
rotated tri -color filters so arranged that when the observer is viewing
the picture on the kinescope through a red section of the filter in the
receiver, for example, the pickup tube is being exposed to the televised
scene through a red section of the filter in the camera. Similarly, when
the blue and green filter sections, in turn, are in front of fhe kinescope,
the blue and green sections are correspondingly in front of the pickup
tube. The red, blue, and green images are repeated frequently enough
so that thé three are superimposed by the "persistence of vision" of
the observer, to create the illusion of a single picture in multiple
colors.I.2,3.45
The operating standards used are: 120 fields per second, 60 frames
1 interlaced, 525 lines, 40 single -color fields or 20 interlaced full
color pictures per second. The color sequence is red, blue, green. With
these operating standards, the resolution obtained with the overall
sÿstem is about 250 lines.
The apparatus is designed so that by slight modification the transmission and reception of color pictures in three dimensions can be
demonstrated. For this operating condition, polarizing light filters are
incorporated with the rotating color filters at the camera and the kinescope. Special polaroid spectacles are provided for the observers to
enable them to separate the right and left images.
During public demonstrations of the color equipment, when it was
necessary to transmit the signal to a point several miles away by means
of a microwave relay link, it was found convenient to transmit the asso2 to
I
J. H. Hammond, U. S. Patent No. 1,725,710.
R. D. Kell, U. S. Patent No. 1,748,883.
J. L. Baird, British Patent No. 473,323.
4 P. C. Goldmark, J. N. Dyer. E. R. Piore and J. M. Hollywood, "Color
Television, Part I ", Proc. I.R.E., Vol. 30, No. 4, pp. 162 -182, April, 1942.
5 CBS En °infers, "Color Television on Ultra High Frequencies ", Electronics, Vol. 19, No. 4, pp. 109 -115, April, 1946.
2
3
www.americanradiohistory.com
TELEVISION, Volume IV
'330
ciated sound on the same radio carrier as the picture by means of a
time division duplexing circuit.6
CAMERA
One of the most outstanding new components incorporated in the
color system is a special form of the image orthicon.? This camera tube
is found to have sufficient sensitivity and a sufficiently uniform spectral
response to make possible direct pickup of studio and outdoor scenes
having an illumii,ation level of from 150 to 300 foot -candles.
Equalization of the sensitivity of the orthicon to the three colors is
accomplished by appropriately masking down the aperture on the filter
Fig.
3- Experimental Color Camera, Left Side.
disc for those colors to which the tube is most sensitive, thus reducing
the time devoted to storing charges during those particular color fields.
Computation of the degree of masking required is based upon measurements of photo- cathode current flowing when the rotating color disc is
6
G. L. Fredendall, Kurt Schlesinger, and A. C. Schroeder, "Transmission of Television Sound on the Picture Carrier ", Proc. I.R.E., Vol. 34, No. 2,
pp. 49 -61, Feb., 1946.
Albert Rose and P. K. Weimer, "The Image Orthicon, a Sensitive
Television Pickup Tube ", presented at the I.R.E. Winter Technical Meeting
on January 24, 1946 in New York, N. Y.
t
www.americanradiohistory.com
EXPERIMENTAL COLOR TELEVISION
331
temporarily replaced by individual color filters. The subject for this
test should be a white surface illuminated by the studio lights. After
this first color balance has been obtained with a given camera tube, it is
possible to operate the camera in much the same way as a conventional
photographic camera. For televising scenes under illumination of
different color temperature, the correction is made by the addition
of a correcting filter over the lens.
The camera is used with either a 90 millimeter, f :3.5, or a 50
millimeter, f :1.9, Eastman Ektar lens. Both are color corrected. A
lens aperture of about f :4.5 is required for the illumination present in
the studio.
Side views of the camera are shown in Figures 3 and 4.
Fig.
4-Experimental
Color Camera, Right Side.
The rotary color disc for the camera, Figure 5, is 6% inches in
diameter. It has twelve filter sectors clamped with the color balance
mask between two plates of glass. The filters used are the conventional
Wratten tri -color photographic filters, numbers 25 (red), 47 (blue),
and 58 (green). The disc is rotated at 600 revolutions per minute by
gearing from a small synchronous motor powered directly from the
60 -cycle mains. The phase position of the motor with respect to vertical
scanning is adjusted by manually rotating the motor frame. The disc
is placed as close as possible to the face of the image orthicon in
order to get it near the focal plane of the lens, thus minimizing any
www.americanradiohistory.com
332
TELEVISION, Volume IV
optical distortion that may be introduced. Since the image orthicon
must operate in a uniform magnetic field of from 60 to 90 gauss,
preferably extending beyond the image section of the tube, the focusing
coil is made in two sections, with the smaller forward section being
mounted in front of the color disc.
A multi- conductor cable connects the camera with the control room
equipment which is mounted on racks of the conventional type. All
electrical controls are on panels in the control room. The video amplifier and the deflection circuits are located inside the camera. All high
voltage and plate supply units are located in the control room.
To overcome the difficulties in the camera and terminal equipment
due to 60 -cycle hum and crosstalk, special power supplies are required.
The heaters of the various tubes in the camera are operated from the
Fig.
5-Camera
Color Filter Disc.
laboratory direct current power supply. The heaters of the tubes in the
control racks are operated on 120 -cycle power obtained by using
selenium rectifiers across the output of special 60 -cycle filament transformers. All plate voltage supplies are regulated and are operated
from a 400 -cycle source.
The video pre -amplifier used in the camera consists of five stages
employing a combination of series and shunt peaking to obtain adequate bandwidth with sufficient gain to raise the signal level to
approximately 1 volt peak -to-peak at the sending end of the camera
cable.
www.americanradiohistory.com
EXPERIMENTAL COLOR TELEVISION
333
CONTROL ROOM EQUIPMENT
A view of the control room equipment is shown in Figure 6. Here is
located the synchronizing signal generator, the main video amplifier,
the sync-mixing and output line amplifiers, the sound -on- picture terminal equipment, a black-and -white picture monitor, and the direct
current power supplies with all camera controls. Control of the video
signal amplitude is accomplished by a gain control circuit in the main
video amplifier.
Fig.
6
-View
of Control Room Racks.
Both the synchronizing signal generator and the video amplifiers
are standard black-and -white picture equipment modified to operate
on color standards. All of the amplifiers used in the system are equalized to beyond 9 megacycles. This would make possible the same degree
of horizontal resolution (at twice the scanning rate) ordinarily obtained with black -and-white standards using a 4.5 megacycle channel,
provided the other limitations imposed by the color system did not exist.
www.americanradiohistory.com
334
TELEVISION, Volume IV
CONTROL ROOM COLOR MONITOR
In addition to a standard 12 -inch black-and -white monitor which is
useful for checking camera focus, color phasing, scanning, etc., there
is provided a color monitor using a 9 -inch kinescope having an aluminized screen' and operating at a second anode voltage of 15 kilovolts.
Front and rear views of this unit are shown in Figures 7 and 8. A
213/4" diameter color disc is used in this receiver carrying six filter
sections and rotating at a speed of 1200 revolutions per minute. Power
to turn the disc is supplied from a 1/8 horsepower 60 -cycle induction
Fig. 7 -Color Monitor, Front View.
motor through a belt and pulley drive. This permits the motor with its
disturbing magnetic fields to be located at a considerable distance from
the cathode ray tube. Synchronization of the color disc speed with the
color field repetition rate is accomplished by a magnetic brake. Proper
color frame phasing is obtained by momentarily releasing the brake by
manually switching off its controlling current.
DRUM COLOR RECEIVER
For best viewing by a large number of observers a demonstration
s D. W. Epstein and L. Pensak, "Improved Cathode -Ray Tubes with
Metal-Backed Luminescent Screens ", RCA REVIEW, Vol. VII, No. 1, pp.
5 -10, March, 1946.
www.americanradiohistory.com
EXPERIMENTAL COLOR TELEVISION
335
receiver was built using a 12 -inch, short persistence, aluminized-screen
kinescope operating at a second anode potential of 17 kilovolts. Photographs of this receiver are shown in Figures 9 and 10.
The rotary color filter is in the form of a large drum, one end of
which is open to allow the kinescope to be supported inside the drum
and at right angles to the axis of the drum by a stationary bracket.
The other end of the drum is closed to provide attachment to the drive
shaft. The periphery of the drum consists of 12 rectangular red, blue,
and green color filter sections clamped in a suitable framework. The
drum is rotated at 600 revolutions per minute, in the direction of ver-
Fig. 8 -Color Monitor, Rear View.
tical scanning and in synchronism with the picture field repetition rate.
The mechanical drive for the drum is similar to that used for the disc
of the control room color monitor.
To overcome the difficulties due to 60 -cycle interference in various
components of the receiver, several precautions are taken. The kinescope is placed in a large mu -metal shield to protect it from the magnetic fields of the motor and power transformers. The power supply
for the tube plates is obtained from a regulated source. All heater
power is obtained from a full -wave selenium rectifier.
www.americanradiohistory.com
TELEVISION, Volume IV
336
High voltage for the kinescope accelerating electrodes is obtained
from a pulse power supply in which the high voltage pulses developed
in an auxiliary winding on the horizontal deflection transformer during
the "fly back" time are used in a voltage -quadrupling rectifier to obtain
17 kilovolts at a current of several hundred microamperes. With this
high anode voltage and the advantage of the aluminized kinescope
screen, a screen brightness of 4.5 foot lamberts is obtained in the highlights of the pictures. Since a light loss of approximately 90 per cent
is introduced by the color filters, the actual kinescope brightness is 45
Fig.
9
-Drum
Receiver, Front View.
foot lamberts. At present a minimum of 10 foot lamberts is considered
satisfactory for monochrome television.
TELEVISION PICTURES IN THREE DIMENSIONS
In natural stereoscopic vision the distance to any object (and hence
the sense of depth) in the scene is determined by three different prop
erties of the views seen by the eyes. The first is the difference between
the two images resulting from the different points of view of the two
eyes ; the second is the focusing of the individual eyes ; and the third
www.americanradiohistory.com
EXPERIMENTAL COLOR TELEVISION
337
is the amount of convergence or toe -in of the two eyes to see a given
object in the scene.
In the stereoscopic television system described here, the images intended for the right and left eyes, respectively, are reproduced on the
kinescope screen in time sequence. The two images are separated by
polarizing the light from them in planes at right angles to one another
by means of sheets of polaroid filter material associated with the color
filters on the rotating drum of the receiver as described previously.
The observer wears a pair of special polaroid glasses, in which the plane
of polarization of each lens is set to agree with the plane of polarization
of the picture intended for the corresponding eye.
Fig. 10
-Drum Receiver, Rear View.
At the camera, a light splitter is mounted in front of the lens. This
attachment, shown in Figures 11 and 12, consists of a system of mirrors
set at 45 degrees behind each of two windows which are spaced horizontally on centers 31/2 inches apart. This spacing is a function of the
normal interpupillary distance and the overall magnification of the
system both optical and electrical. The factors are related by Rule9 in
the equationT = wed /sf
9 John J. Rule, "The Geometry of Stereoscopic Projection ", Jour. Opt.
Soc. Amer., Vol. 31, No. 4, pp. 325 -334, April, 1941.
www.americanradiohistory.com
TELEVISION, Volume IV
338
where w = width of image on photo- cathode of camera ; s = width of
image on viewing screen ; T = lens separation of the camera; e=
human interocular distance; f = focal length of camera; d = distance
of camera lens to plane of object which is intended to appear coincident
with the plane of the viewing screen. Sequential separation of the two
Fig.
11- Stereo Attachment
for the Camera (Light Splitter).
images is achieved by means of polaroid filters which are placed over
each of the windows so that the light coming from the scene as viewed
through the left "eye" is horizontally polarized, while that through
the right "eye" is vertically polarized. Selection of the particular
®
-
_II
WINDOWS
I
%
I
`
Yo
FI XED
ADJUSTABLE
MIRROR
MIRRORS
CAMERA LENS
Fig.
12
-Light Splitter.
image to be transmitted during a given field is made possible by
means of additional polaroid filters mounted on the rotating color disc
with their planes of transmission arranged in quadrature and in alternate fashion. Thus the vertically polarized image is transmitted during
www.americanradiohistory.com
EXPERIMENTAL COLOR TELEVISION
339
one field and the horizontally polarized image during the next, the
unwanted image being suppressed by crossed polarization.
The angular setting of one of the mirrors in the light splitter is
adjustable so that the convergence of the camera "eyes" can be set to
bring into register some object near the center of the useful depth of
field. Observation of this point on the screen of the kinescope corresponds to the situation illustrated in Figure 13(A). In this case the
object appears to be in the plane of the viewing screen since the angle
of toe -in of the eyes is commensurate with the focal distance. The
image of objects farther away from the camera will not fall at the
same place on the screen, but will be separated horizontally a small
amount, depending upon their position. Thus the horizontally polarized
image intended for the left eye is displaced to the left as indicated in
APPARENT IMAGE
POSITION.
II
VIEWING
SCREEN
01-0/
/\
/
\
POLAROID
J
LENSES-ILH
7-."11)
OBSERVER'S
I
\
'
A
I
V1
L
8
A
APPARENT
\ IMAGE POS-
i
\
I
I
\
\
\
\
EYES
0H
V
H
I
/
I
B
\
I
\
1
H1
1
®
b
B
\
HI__
®
\
/
\
R
C
Fig. 13-The Geometry of Stereo Viewing.
Figure
13 (B). This condition gives the impression that the object is
behind the screen as a result of the angle that the eyes must take in
order to obtain fusion of the two images. Conversely, objects nearer
to the camera are displaced horizontally in the opposite direction on
the screen as shown in Figure 13 (C) and thus appear to the observer
to be located in front of the screen.
The optical adjustments of the system can be made to give orthostereoscopic pictures only for one viewing distance and screen size.
When more than one receiver is to be operated from a given camera,
and if different amounts of magnification are to be used, some correction for the depth and perspective distortion can be obtained by keying the horizontal positioning circuit field by field thus changing the
horizontal displacement of a given object on the screen, and hence its
apparent position relative to the observer.
roP
www.americanradiohistory.com
340
TELEVISION, Volume IV
SOUND -ON- PICTURE EQUIPMENT
During certain public demonstrations of the color television apparatus the receiving equipment was in another building several miles
from the studio. A microwave relay link operating on a frequency of
10,000 megacycles was employed on these occasions for transmission.
The associated sound signal was transmitted on the picture carrier by
means of a time division duplexing system.'
The essential elements of the sound -on-picture equipment are shown
in dotted lines on the diagram of Figure 2. The sound modulator is
inserted between the video amplifier and the output line and mixing
amplifier and keys a rectangular pulse into the "back porch" of the
horizontal blanking. This pulse is of constant amplitude extending
down to white level and is adjustable in width in accordance with the
audio modulation amplitude. In fact, the variable width modulation
system is similar in many respects to the variable area sound track as
commonly used for sound motion pictures.
Demodulation of the duplexed sound channel is accomplished at the
receiver by means of a synchronized electronic switch which serves to
exclude all but the sound carrier pulses. An amplitude limiter removes
amplitude noise. This sound system has the basic theoretical limitation
that the maximum audio frequency that can be transmitted is one -half
the horizontal line scanning rate, or 15,750 cycles for the color system
employed. A further minor reduction in the available bandwidth is due
to the cutoff characteristic of the low-pass filter which is required to
exclude frequencies above 15,750 cycles.
www.americanradiohistory.com
SIMULTANEOUS ALL -ELECTRONIC COLOR
TELEVISION *f
A
Progress Report
BY
RCA Laboratories Division, Princeton, N.
J.
Summary -This paper presents the latest progress report on
Television and the first on the new simultaneous all -electronic system.
design and operating characteristics are reviewed. The apparatus for
ning color slides and color motion picture film together with the
television receivers are described.
Color
Basic
scancolor
N OCTOBER and November, 1946, the Radio Corporation of
America gave several demonstrations of color television to press,
industry and Government groups. These demonstrations constituted a progress report on the work done in color television, which
follows the program announced at the time of earlier demonstrations
in December, 1945.1 In the current demonstration, important advances
in color television were shown. The new system is all -electronic, having
the potential flexibility inherent in electronic arrangements, and simultaneous, all three color images being transmitted continuously. This
system has many operating and performance advantages and is compatible with the present black- and -white television. Since each of the
three color channels employs the same standards as those now in use
for black- and-white transmission, the green channel is suitable for
monochrome presentation. Color television of this type can be introduced at any time it is made ready and can be operated interchangeably
with black -and -white television; undesirable obsolescence is not created.
The recent demonstrations included television pictures in natural
color scanned from kodachrome slides and from 16-millimeter color
motion picture film. In order to demonstrate interchangeability, pictures in monochrome using signals of present black- and -white standards
were shown on the color receivers ; pictures in monochrome, using
signals of the simultaneous color transmission, were then demonstrated
on a current model black- and -white receiver.
Research work is under way and progress is being made in the
radio transmission and reception of simùltaneous all- electronic color
*
Decimal Classification: R583.
REVIEW,
t Reprinted fromL.RCA
Fredendall, A.
R. D. Kell, G.
1
December, 1946.
C. Schroeder, and R. C. Webb, "An
Experimental Color Television System ", RCA REVIEW, Vol. VII, No. 2,
pp. 141 -154, June, 1946.
341
www.americanradiohistory.com
TELEVISION, Volume IV
342
television and in the building of television cameras for studio and
outdoor pickup of this system. This work, together with propagation
tests and field surveys, is a part of the over -all schedule yet to be fully
worked out, but already well along.
Since simultaneous all- electronic color television is of far-reaching
importance, the experimental equipment used during the recent demonstrations is described herein. This includes the apparatus for scanning
color slides and color motion picture film together with the television
receivers for color. Some of the basic design and operating characteristics are also reviewed. Figure 1 is a block diagram of the system.
COLOR FILM
SCANNING UNIT
RED
IPHOTOCEILL
1
COLOR
SELECTIVE MIRRORS
GREEN
LENS
PHOTOCELLS
SCANNING
LENS
CATHODE- RAY
TUBE
L
COLOR
FILM
BLUE
COLOR
CHANNELS
SIMULTANEOUS COLOR PICTURE
PROJECTION
BED
GREEN
COLOR
PICTURE
BLUE
COLOR
CHANNELS
-RA-
CATHODE
PROJECTION
TUBES
44111111144.,
PROJECTION
LENSES
SCREEN
Fig. 1 -Block diagram of the simultaneous all- electronic color television
system. (Each of the color channels, shown at the right of the upper figure, have the same operating standards as the current black- and -white
system. At the receiving end, shown at the left of the lower figure, each
color channel is associated with its separate cathode -ray projection tube.)
STATIONARY PICTURE SIGNAL GENERATOR
One of the primary needs for the development of a simultaneous
color television system is a standard source of tricolor video signals
on which one may rely for good resolution, good registration, high
signal -to -noise ratio, freedom from spurious signals, and good color
www.americanradiohistory.com
SIMULTANEOUS COLOR TELEVISION
343
fidelity. A special slide scanner utilizing a cathode -ray tube as a flying spot scanner, a beam splitter, and three photoelectric tubes, were developed for this purpose.
A photograph of this apparatus with a superimposed phantom view
of the kinescope and the light paths is shown in Figure 2. The raster
formed on the screen of the kinescope is imaged on the slide by means
of a lens. The light rays transmitted by the slide are condensed and
Fig.
2-Stationary picture signal generator.
then divided by dichroic mirrors which pass one color of light and
reflect the other colors. The use of dichroic mirrors for a light splitter
instead of half -silvered mirrors and color filters, reduces light losses
and therefore provides a signal with higher signal -to -noise ratio. The
divided light beams are further filtered by color absorption filters, then
www.americanradiohistory.com
344
TELEVISION, Volume IV
collected by multiplier type phototubes which convert the varying light
intensity of the spot as transmitted by the slide into video signals
corresponding to the three primary colors of the slide. The use of
multiplier phototubes provides a high video input to the amplifier.
The amplifiers are equalized to correct for the decay characteristic of
the phosphor used in the flying -spot kinescope.
On the bottom of the rack in the photograph (Figure 2) is the
chassis containing the synchronizing, blanking, and deflection circuits
for the flying -spot kinescope. Behind the kinescope is the high -voltage
power supply which provides approximately 30 kilovolts for the kinescope and a first anode focusing potential variable between four and
seven kilovolts.' The flying -spot kinescope has a special short persistence
phosphor whose light intensity drops to less than 1 per cent of its
original value in one microsecond. Facing the kinescope screen is the
slide holder with an F:2 objective lens. The whole optical assembly
is mounted on the same chassis with the three video amplifiers. The
beam splitter is at the lower end of this chassis and a condensing lens
system for each channel reduces the beam diameter to that of the photocell aperture after the beam is divided. The photocells are enclosed in
shield cans, which also support the color filters.
The voltage for the phototube multipliers may be controlled by the
variable power supply directly under the beam splitter chassis, and by
this means the video levels of all three channels may be varied simultaneously. The supply voltage of the phototube multipliers of the
individual channels may be varied individually by the potentiometers
visible at the top of the chassis, to provide the desired color balance.
Each of the three video amplifiers contains three stages having a flat
frequency response to approximately 5.5 megacycles. Included in each
of the amplifiers are the equalizing circuits to compensate for the
various phosphor persistences. The output level of the amplifiers is
approximately 1 volt peak -to -peak. The small chassis above the beam
splitter is for the insertion of the synchronizing signals in the green
video signals.
The quality of the signal from this generator is highly satisfactory
not only because of the high resolution, but also because the blacks have
the unusual characteristic of being practically free from noise. Noise
in the picture therefore has the general appearance of the equivalent
effect found in motion pictures from photographic grain and dirt. The
registration of the three signals is inherently correct.
MOTION PICTURE FILM SCANNER
The motion picture film scanner was built with no attempt at
www.americanradiohistory.com
SIMULTANEOUS COLOR TELEVISION
345
refinement or optimum design, in order to hasten preliminary tests of
reception with moving subjects. Its general scheme is the same as
that of the slide scanner, with the film gate replacing the slide holder.
A photograph of the apparatus is shown in Figure 3. A standard
16- millimeter home sound film projector was modified by substituting
a synchronous motor drive so that the film speed was changed to 30
frames per second (instead of 24). Each frame was then scanned
Fig.
3-Motion
picture film scanner.
twice to give 60 fields per second. The pull-down mechanism, which
was unchanged, is so slow that it was necessary to blank approximately
30 per cent of the field time to avoid showing the distorted picture produced during the film pull -down time. The picture therefore actually
contained only about 370 lines, although the nominal number of lines
was retained at 525.
www.americanradiohistory.com
TELEVISION, Volume IV
346
The picture quality was judged to be good, particularly when allowance is made for the fact that part of the picture area was missing, due
to the compromise in design of the film projector. The sound was
usable, but was not very satisfactory due to the improper film speed.
REPRODUCING EQUIPMENT
The picture reproducer, as shown in Figure 4, 5, 6, contains three
Fig.
4- Laboratory
model simultaneous all- electronic
color television receiver.
three -inch kinescopes arranged side -by -side in an equilateral -triangular
group, each having an associated projection lens and deflection yoke.
The kinescopes are identical except that phosphors selected for producing red, green, and blue light, respectively, are used. The kinescopes and lenses are mounted in an assembly frame which also holds
the yokes in such a manner that each yoke may be adjusted in rotation
and height without disturbing the kinescope mounting position. Each
www.americanradiohistory.com
SIMULTANEOUS COLOR TELEVISION
347
kinescope is provided with the video signal corresponding to its particular primary color, and has a scanning raster which produces light
for its primary color image in the completed picture. The lenses
project these three pictures simultaneously to the translucent viewing
Fig.
5- Receiver
with three cathode -ray projection tubes, lenses removed.
screen by way of a 45- degree mirror, as shown by dotted lines drawn
on the photograph (Figure 6). The kinescopes are operated at a second
anode potential of 25 kilovolts.
www.americanradiohistory.com
TELEVISION, Volume IV
348
The optical system serves to focus and combine the three pictures
on the translucent screen. In so doing, the images must not differ from
one another in geometric distortion or location. Prevention of such
difference is accomplished by placing the three kinescope faces in the
same plane and mounting the three lenses above this plane, with their
Fig.
6-Receiver
with lenses in position, showing projection paths.
axes perpendicular to it. The axis of each lens is offset from the center
of the kinescope face toward the center of the assembly by an amount
sufficient to bring the three pictures into approximate register on the
www.americanradiohistory.com
SIMULTANEOUS COLOR TELEVISION
349
screen. Exact registry is then obtained by moving the raster on the
kinescope face electrically. This offset, which is similar to a rising
front on a photographic camera, causes no distortion, but requires
extra covering power in the lens. The lenses used are F :2 projection
lenses. They are threaded into the lens plate at the calculated positions
and the threads serve as the focus adjustment.
The registration requirements are similar to those existing in color
printing and color photography. Ideally, the three rasters should be
identical and properly positioned within a fraction of the width of a
scanning line. Practically, a considerable amount of misregistration
may be present without being objectionable.
The scanning rasters are made substantially identical by using three
similar yokes, and supplying them with power from the same deflecting
circuit. The three yokes are connected in parallel rather than in series.
This permits a simple individual centering or positioning arrangement
and also insures more nearly identical deflection fields. In a series
arrangement, one yoke would operate at a higher alternating- current
potential with respect to ground, and would thus be shunted to a greater
extent by the stray capacitances.
The positioning or centering arrangements are the usual television
centering circuits, the only requirements being that the centering supply voltage must be stable, and that adjustment of the potentiometers
must not alter the current waveshape through one yoke with respect
to that through the others. To insure the latter, the horizontal centering potentiometers are by-passed very lightly and the vertical ones are
unby -passed. Enormous capacities would be required to by -pass the
vertical centering potentiometer properly. However, since the vertical
circuit is essentially resistive, the addition of more resistance would
simply change the amplitude, which is easily corrected.
The procedure used in registering the kinescope assembly is as
follows. The kinescopes are adjusted for the proper height and clamped,
then the lenses are focused optically. The yokes are then rotated until
the edges of the three rasters are parallel. One of the three rasters is
then considered standard for horizontal size, and one of the others is
adjusted to it by moving the deflecting yoke up or down. Although
this, of course, varies the raster size both horizontally and vertically,
only the horizontal size is considered during this adjustment. This
adjustment is then repeated on the remaining raster until the three
horizontal sizes are alike. The vertical sizes are then adjusted by varying the value of small resistors in series with the vertical deflecting
coils. The positioning or centering controls are then set to register
properly the three rasters which are now of the same size.
www.americanradiohistory.com
350
TELEVISION, Volume IV
Successful registry of the three rasters has been greatly facilitated
by the use of aluminized kinescope tubes.' The aluminum film insures
that the phosphor screen and the glass wall are at second -anode potential and hence do not collect charges that will divert the electron beam
erratically.
2 D. W. Epstein and L. Pensak, "Improved Cathode -Ray Tubes With
Metal- Backed Luminescent Screens", RCA REVIEW, Vol. VII, No. 1, pp.
5 -9, March, 1946.
www.americanradiohistory.com
MILITARY TELEVISION t
BY
GEORGE M. K. BAKER$
Staff Assistant to the Executive Vice President in charge,
RCA Laboratories Division. Princeton, N. J.
Summary-A general discussion and short history of television for
military uses are presented, followed by a brief description of military
television developments of World War II. Operational and other uses of
these developments are discussed. The future of television for military and
other purposes is considered in the concluding portion of the paper.
ANUMBER of excellent military developments completed during
World War II saw only limited combat service or were never
introduced into action at all. The reasons underlying such
limited employment furnish ample material for an exhaustive treatise,
but it is sufficient here to state that, whatever the reasons, television
for military use was one such outstanding development.
Why, then, spend time reviewing and studying the subject? There
are several valid reasons for so doing. First, in one sense, military
television did see extensive war service in the forms of radar, radar
countermeasures, loran, altimeters, shoran, advanced communication
aids, and other devices whose war records need no discussion. Practical television resulted from two major developments which were of
extreme importance because they not only made television possible but
also laid the foundations for other important war services -such as
radar. Because of the greater immediate importance of radar, it
received more intensive development and much wider use in World
War II than did its parent, television. The two major developments
which made possible television (and radar, shoran, loran, etc.) were:
(1) improved cathode-ray tubes for converting light waves to electrical
signals and for reconverting these signals into visible light; and (2)
electrical circuits and components associated with cathode -ray tubes
capable of controlling, timing and utilizing these tubes in an extremely
precise manner. The new tubes and circuits provided revolutionary
means of effecting time control and measurement with a previously unknown accuracy. In furnishing the basic ground work for such tubes
Decimal Classification: R583 X R560.
Detailed technical papers on military television equipment are included herein following this introductory paper.
$ Lieutenant Commander, United States Navy (Retired).
t
351
www.americanradiohistory.com
TELEVISION, Volume IV
352
and circuits, television saw war service in many forms and consequently
provides material which merits close study.
A second valid reason for examining military television is found in
the latest commercial television equipment and techniques. Television's
war service is now being repaid; war experience has given added
impetus to commercial television and this, in turn, is responsible for
the much -improved cameras, receivers, relay -links and tubes which
today are being incorporated into a growing domestic television service.
A background knowledge of war results and a detailed study of military
developments are important for understanding and ability to work with
advanced commercial television equipment.
There is another reason for reviewing wartime television developments. To the difficulties attending the introduction of any new military development, another may be added in the case of television for
military uses. Pre -war commercial television equipment required complete and radical redesign, and operation of television equipment in
planes and missiles raised some of the most troublesome engineering
problems of the war.' A study of these problems and a thorough understanding of their implications is a prerequisite for advanced television
development.
Finally, where military television did see action, it proved beyond
question that here was still another example of the effectiveness of
applied modern electronics. The future role of military television is
now clearer and more easily understood; heretofore unthought -of military applications have become apparent with passing time and coincidentally with the development of other new military devices.
Concurrently, it has become possible to perceive new and startling
peacetime uses for television as a direct result of the experience gained
in war.
It is well, then, to review very briefly the history of military television' and mention the operational and other uses to which it was
assigned.
In 1934, Dr. V. K. Zworykin first advanced a concrete proposal for
using television for military purposes to substitute for human eyes
where it was not advisable for men to go. This first system is described
'
These problems are fully discussed in other papers:
Charles J. Marshall and Leonard Katz, "Television Equipment for
Guided Missiles ", Proc. I.R.E., Vol. 34, No. 6, pp. 375 -401, June, 1946.
M. A. Trainer and W. J. Poch, "Television Equipment for Aircraft ",
RCA REVIEW, Vol. VII, No. 4, pp. 469-502, December, 1946.
A more detailed history of airborne television is included in another
paper: Henry E. Rhea, "Airborne Television ", Broadcast News, No. 43,
pp. 24 -41, June, 1946.
2
www.americanradiohistory.com
MILITARY TELEVISION
353
in detail in a memorandum by Dr. Zworykin.3 In 1935, work began on
a lightweight airborne reconnaissance television system which was
successfully demonstrated in 1937. Two years later intensive development work was commenced on still smaller and lighter equipment.
This new equipment was flight tested in 1941. The next three years
were devoted to overcoming a multitude of technical difficulties attending the installation of the new television equipment in aircraft and
guided missiles,' and in the design of the missiles themselves. Finally,
in 1944 military television came into actual combat use.
Three basic military television systems were developed during
World War II. They are known by the code names BLOCK, MIMO,
and RING. These systems are described in detail in other papers';
only a general discussion of each is included herewith.
The BLOCK system consists of several designs built to operate on
different frequencies. This equipment, manufactured in quantity, was
used by both the Army and Navy-by the Army in the GB -4 radio controlled glide bomb and in old "war-weary" B -17's, and by the Navy
in TDR -1 drones and in the GLOMB radio -controlled glider bomb.
BLOCK equipment, employing the iconoscope or image orthicon, is
lightweight and compact-the camera unit weighing but 33 pounds and
the transmitter unit 26 pounds. The entire weight of all television
equipment in the drone or bomb is 100 pounds. It was designed to
operate unattended and is, in the main, expendable. This equipment is
also used, however, for certain other non -expendable applications.
The MIMO system is similar in most respects to BLOCK. It is,
however, lighter and more compact, employing a new developmental
MIMO-miniature image orthicon. The entire system, mounted in the
Army ROC high -angle radio -controlled bomb, for which it was specifically designed, weighs but 50 pounds.
The RING equipment provides a more elaborate, high -resolution,
airborne television system for reconnaissance. It was designed for
attended operation with two or more cameras and is not considered to
be expendable.
K. Zworykin, "Flying Torpedo with an Electric Eye ", RCA
8 V.
REVIEW, Vol. VII, No. 3, pp. 293 -302, September, 1946.
4 M. A. Trainer and W. J. Poch, "Television Equipment for Aircraft ",
RCA REVIEW, Vol. VII, No. 4, pp. 469 -502, December, 1946.
R. D. Kell and G. C. Sziklai, "Miniature Airborne Television Equipment", RCA REVIEW, Vol. VII, No. 3, pp. 338 -357, September, 1946.
Paul K. Weimer, Harold B. Law and Stanley V. Forgue, "MIMOMiniature Image Orthicon ", RCA REVIEW, Vol. VII, No. 3, pp. 358 -366,
September, 1946.
R. E. Shelby, F. J. Somers, and L. R. Moffett, "Naval Airborne Television Reconnaissance System ", RCA REVIEW, Vol. VII, No. 3, pp. 303 -337,
Sept., 1946.
www.americanradiohistory.com
354
TELEVISION, Volume IV
BLOCK and MIMO transmitted television pictures to the controlling planes, which pictures gave the necessary control information.
RING transmitted the television reconnaissance pictures to receivers at
the base, in ships, or in other planes. All three systems provided information which fulfilled their design requirements. In the case of the
systems used in drones or bombs, the detail in the picture improved as
the need for more accurate control increased -i.e. as the drone or bomb
neared the target-and was adequate for control purposes. In the case
of the reconnaissance system, the detail, of course, varied with altitude
and visibility conditions but was generally sufficient to distinguish all
important details of the terrain under observation and to differentiate,
for example, between the types of various motor vehicles and parked
aircraft. The distance a RING -equipped plane could transmit clearly
to its base varied between 100 and 200 miles, with values of over 200
miles having been recorded.
In August, 1944, the Navy used television- equipped TDR -1 drones
against Japanese shipping in the Northern Solomons. A few months
later, a radar- equipped lighthouse at Rabaul Harbor was destroyed by
the same means after having successfully withstood repeated bombing
attacks of the ordinary type. Previous to these events, in the summer
of 1943, the Army began using old "war-weary" B -17's, television equipped and radio -controlled, to destroy submarine pens in Helgoland. Later, GB -4 glide bombs equipped with television were sent
against the V-1 and V-2 launching sites on the French coast.
BLOCK equipment was also installed in reconnaissance planes after
the capture of the Philippines and used for patrol work and battle
damage survey.
Within this country, military television equipment played its part
in the Manhattan project and in other wartime manufacturing processes which required constant surveillance of operations too dangerous
to be approached by men themselves. Military television also participated in Operations Crossroad, where modified BLOCK equipment was
used for telemetering and also for viewing the blasts from towers on
Bikini Atoll and from drones flying through and around the atomic
clouds -providing a close -range picture of these significant events to
observers who otherwise would have had to be contented with film
records and long -range views.
*
*
*
All people hope that wartime developments have seen their last
combat action. It does no harm, however, to consider how this elec-
www.americanradiohistory.com
MILITARY TELEVISION
:iC)Ci
tronic weapon can help in our nation's defense should ever there again
be cause to do so.
Military television will find continued usefulness in guided missiles.
In fact, it is probable that most radio-controlled weapons of the future
will include television equipment to make it unnecessary for the controlling organization to have either the missile or the target in direct
view. This is obviously a definite requirement in the case of very -high
altitude bombs or stratosphere rockets, or for ground- control of any
type of guided missile.
The range of daylight reconnaissance television equipment will be
extended materially to provide detailed information on distant points
as readily as it now does on fairly close-range targets. The ultimate
extension of this system is interesting and has already been the subject
of some conjecture. The central war rooms of the future will not have
ordinary walls; they will have huge television screens. The personnel
within will be able to see their actual surroundings in all directions,
with various range segments brought on the screens at will, and with
closeup studies available on any section as desired.
Command posts behind battle lines will see their entire sector as
one continuous panorama, the television signals coming from the
"walkie-lookie" transmitters in the foremost positions, and from
reconnaissance planes overhead. Officers controlling tank units will
have similar panoramic television information from leading tanks and
planes. Artillery observers will no longer receive "spots "; they will
see the shots landing with relation to the target and can make their
own corrections -an extension of the current high- resolution radars.
The flag bridge of ships will be smaller models of the larger land based war rooms, with the entire fleet spread out before the Admirals'
eyes. Enemy forces can be viewed as a whole or in part, either separate
from or dubbed into their true location with reference to land positions,
as in the case of contested amphibious operations. Another use for
television aboard ships will be in maintaining a display of CIC and
other data boards in gunnery plot, flag plot, on the bridge, and in other
important stations.
Night will not necessarily stop military television's activities of
the future. Tubes, sensitive to infrared (as are certain image orthicons
today) will make it possible to view scenes so illuminated almost as if
by daylight.
The principles of military television and the experience gained in
tests and combat operations are destined to -day a large part in the
www.americanradiohistory.com
356
TELEVISION, Volume IV
development of systems for air navigation, such as Teleran. This same
experience and these same principles will also greatly expedite the use
of television in connection with personal communication, air coverage
of news events, control of laboratory and production processes, air
supervision at scenes of disasters, etc., use by police in traffic and
crime control, and the development of "walkie-lookie" television. As
previously pointed out, many wartime television developments have
already been incorporated into postwar commercial equipment.
Eventually, means will probably be devised which will overcome,
through the use of radar techniques, the restrictions that are currently
imposed on television by weather and visibility conditions or the short range of infrared illumination. Until then, and after this occurs,
military television will be a part of all operations -an extension of
human vision. Even in this enlightened age, humans, with the
limitations of their eyes, will have to fight the wars, if wars must be
fought, and this new electronic weapon will assist them.
www.americanradiohistory.com
INTRODUCTION
to
TECHNICAL PAPERS ON AIRBORNE TELEVISION *t
-
THIS issue of RCA REVIEW contains the first of a series of
technical articles on airborne television a system of sight
transmission having momentous military and civilian applications. Prepared and written by scientists and engineers of Radio
Corporation of America, they are presented to readers of RCA
REVIEW as an historic record of pioneering and scientific progress.
The idea behind airborne television and its development originated
in RCA more than twelve years ago. It was in the spring of 1934 that
Dr. V. K. Zworykin formulated plans and submitted to me a memorandum suggesting the creation of such a system to serve as "electronic
eyes" in guiding radio-controlled aerial torpedoes. At that early date,
Dr. Zworykin foresaw the threat of Japan's "Kamikaze" or Suicide
Corps, and sought to achieve by technological means what the Japanese
hoped to attain by psychological training. I was so impressed that, accompanied by Dr. Zworykin, I went to Washington and presented his
plans to the War and Navy Departments. Some time elapsed before the
armed services became actively interested in airborne television, but
our scientists, meanwhile, continued to experiment and pioneer with
this revolutionary method of extending human sight. First Ray D. Kell
and Waldemar Poch developed light -weight cameras and associated
equipment. Then Henry Kozanouski joined in and produced research
equipment which was field tested in an airplane. When the war emergency arrived, the entire organization was ready to meet the challenge.
designated "Block", "Ring"
Three airborne television systems
evolved for secret warpseudonyms
as
security
and "Mimo" projects
equipment that
and
transmitting
-up
pick
time purposes. Television
and built
modified
redesigned,
room
was
once might have filled a large
system,
which
in
Block
the
uses
military
to "suitcase" compactness for
The
and
Navy.
Army
both
the
war
by
was employed effectively in the
final
during
the
was
developed
heavier, longer -range Ring system
stages of the conflict by engineers of the National Broadcasting Company, Inc., in conjunction with the U. S. Navy. The Mimo equipment
was the midget of the three systems, being even smaller than the Block
- -
*
Decimal Classification: R583 X R560.
from RCA REVIEW, September, 1946.
t Reprinted
357
www.americanradiohistory.com
358
TELEVISION, Volume IV
apparatus. It was developed primarily for use in guided missiles where
space was insufficient to accommodate the Block equipment.
The articles in RCA REVIEW will trace the development of the
three airborne television systems and relate in technical terms how
the special requirements were met for equipment that would operate
satisfactorily under the unusual handicaps of aerial warfare. These
reports will tell of the design of small antennas practicable for airplanes, the use of the airplane's power supply, and the overcoming of
the problems of noise and vibration. Special emphasis will be given to
the development of the now celebrated image orthicon tube. Accounts
likewise will be printed of technical aspects in the development of other
vital electronic tubes and equipment. The full text of Dr. Zworykin's
original memorandum of 1934 is published in this issue.
Great praise is due the scientists and engineers whose research and
pioneering, technical knowledge and ingenuity made possible airborne
television as a successful weapon of war and opened the way for monumental progress in widening television's scope of service in peace.
DAVID SARNOFF, President
Radio Corporation of America
www.americanradiohistory.com
FLYING TORPEDO WITH AN ELECTRIC EYE''t
BY
V. K. ZWORYKIN
Director of Electronic Research Laboratory, RCA Laboratories Division
Princeton, N. J.
A memorandum sent to DAVID SARNOFF, President, Radio Corporation of America, on April 25, 1934.
Summary-This paper, written in April, 1934, presents a detailed suggestion for the control of guided missiles using information obtained by
television. Shortcomings of previous systems of guided missile control are
briefly mentioned and a general description is given of the television apparatus for use in the new method of control. Approximate weight composition of such a television- controlled aerial bomb or torpedo is included. The
suggestion envisions that the torpedo or bomb (or standard airplane) should
be equipped both with automatic pilot control and remote radio control with
the instrument and target data supplied by an iconoscope camera and transmitter in the piloted weapon.
Television information furnished would be of two kinds, and would be
given simultaneously: (1) an actual view of the target which could be
sighted upon by means of crosshairs; (2) accurate information on the readings of instruments in the piloted weapon, given by the position of bright
spots on the edges of the picture and read on scales attached to the receiving
tube in the control ship. This latter feature is designed to facilitate the
checking of instruments in the torpedo prior to release and also while in
flight when actual target view is obscured and the automatic pilot is in
control.
The particular significance of this paper lies in the large time interval
which has elapsed since its preparation. This time element gives adequate
proof of the author's foresight and ingenuity, particularly when the details
of the system outlined are compared with those of systems in use today, 12
years later.
THERE have been quite a number of attempts to devise an efficient flying weapon. The aerial bomb is the simplest form, and
the recent improvements in aerial ballistics make these bombs a
most formidable modern weapon. The use of such a bomb usually requires a close approach of the bombing airplane to the target, thereby
subjecting the plane to the barrage of the anti -aircraft batteries. It
follows that, simultaneous with the development of aerial bombing,
there has been improvement in anti -aircraft artillery which has considerably lessened the effectiveness of the aerial bomb.
Considerable work has been done also on the development of radio controlled and automatic program -controlled airplanes having in mind
*
Decimal Classification R583 X 560.
from RCA REVIEW, September, 1946.
:
t Reprinted
359
www.americanradiohistory.com
360
TELEVISION, Volume IV
their use as flying torpedoes. The possibilities of such airplanes were
demonstrated repeatedly in various countries. during the past few
years. Both these methods, however, have the same fundamental difficulty, viz., that they can be used efficiently only by trained personnel at
a comparatively close range, thereby being subjected to anti -aircraft
gun -fire. Both radio and automatic -controlled planes lose their efficiency
as soon as they are beyond visual contact with the directing base. The
solution of the problem evidently was found by the Japanese, who,
according to newspaper reports, organized a Suicide Corps to control
surface and aerial torpedoes. The efficiency of this method, of course,
is yet to be proven but if such a psychological training of personnel is
possible, this weapon will be of the most dangerous nature. We hardly
can expect to introduce such methods in this country, and therefore
have to rely on our technical superiority bo meet the problem.
GENERAL DESCRIPTION
One possible means of obtaining practically the same results as the
suicide pilot is to provide a radio -controlled torpedo with an electric
eye. This torpedo will be in the form of a small steep angle glider,
without an engine, and equipped with radio controls and an iconoscope
camera. One or several such torpedoes can be carried on an airplane to
the proximity of where they are to be used and there released. After
it has been released the torpedo can be guided to its target by shortwave radio control, the operator being able to see the target through
the "eye" of the torpedo as it approaches.
The carrier airplane receives the picture viewed by the torpedo
while remaining at an altitude beyond artillery range. It is not even
necessary to have direct visibility of the target from this plane, as the
information is supplied by the torpedo from a much closer range. The
distance between the plane and torpedo will always be short; therefore
the power of the short -wave radio transmitter on the torpedo can be
very low. A transmitter of 5 or a maximum of 10 watts, operating
between 3 and 10 meters, will be sufficient for this purpose. Since the
image of the target increases in size as the torpedo approaches, it is
not necessary to provide an electric eye with great resolution or with
highly efficient optics. Therefore, an iconoscope camera operating with
a 90 -line picture and with a wide -angle lens will be sufficient for this
purpose. When the torpedo is first launched from the plane it may not
immediately supply any useful information to the control due to the
excess height or intervening clouds, but when it begins to approach the
ground the visibility will gradually increase and the accuracy of the
aiming will be improved. At close range the target will be sufficienbly
www.americanradiohistory.com
FLYING TORPEDO
361
large to provide good visibility even at 90 lines and the accuracy of the
aiming will be the greatest just before the moment of contact of the
torpedo with the target. This introduces an entirely new principle in
ballistics, since in all existing methods the operator has no way of controlling a projectile once it has been released.
The radio receiving equipment of the torpedo can also be simplified
by using a directional and more powerful radio transmitter on the
mother plane and also by a decrease in the width of the communication
channel, due to the necessity of transmitting from the controlling plane
only three or four sets of signals. This can be accomplished by a shortwave carrier modulated with widely separated frequencies. The necessary electrical supply for the torpedo is easily obtainable from a
propeller- driven generator with several commutators supplying all
necessary direct -current potentials.
The radio control of the torpedo can be accomplished by using one
of the already -developed schemes, but can be considerably simplified by
using, for control purposes, a circuit and tubes which were developed
during the past couple of years in connection with radio communication.
It is very difficult to specify at present the probable weight of the
total electrical equipment without actually building a model. A preliminary estimate shows that the total weight of the equipment with
automatic pilot, including wind- driven generator, will be below 150
pounds, or less than the weight of one pilot. This weight is composed
of the following items:
(1) Iconoscope camera for 90 lines with deflection and
tilting arrangement, and short -wave radio transmitter with modulation up to 100,000 cycles and
10 watt power
(2) Wind-driven generator for 1000, 300 and 6 volts
125 watts, and three control drums
(3) Short -wave radio receiver for 3 audio tuned channels with relays
(4) Automatic pilot with controls and accessories
-
45 pounds
15 pouilds
40 pounds
140 pounds
Total
-
40 pounds
-
The weight of the torpedo can be composed, for instance, of : Control
300
120 pounds; explosives
equipment
140 pounds; fuselage
pounds, making the total weight 560 pounds per unit.
Due to the fact that the torpedo has no landing speed, the load per
square foot of the wing area can be increased several times in comparison with that of an airplane, therefore the whole torpedo can be made
very compact. Four such torpedoes can be packed under the wings of a
-
www.americanradiohistory.com
TELEVISION, Volume IV
362
normal sized bomber. If necessary, the amount of explosives indicated
above can be increased by increasing the size of the torpedo.
An approximate idea of the appearance of such a torpedo is shown
in Figure 1, which, of course, is probably very far from the actual
shape that such a torpedo will have after its final development.
.
ICONOSCOPE CAMERA AND TRANSMITTER
The iconoscope camera has already been developed by us for television purposes. However, due to the decrease of the required number
CONTROL DRUMS
TRANS M.
I
1
LJ
RECEIVER!
ICONOSCOPE
GENERATOR
ANTENNA
Fig.
1
-The
Flying Torpedo.
of lines from the present 340 to 90 lines, the whole apparatus will be
substantially simpler and smaller. The associated circuits for deflection of the electron beam and the amplifier will contain only a fraction
of the number of tubes used in the present system.
The camera mounting is provided with a tilting arrangement.
which points it always in such a manner that the center of the received
picture coincides with the point to which the torpedo is heading. The
www.americanradiohistory.com
FLYING TORPEDO
363
tilting is controlled by the same device which controls the level flight
of the torpedo and which will be described later. The optical lens of
the camera is provided with a sighting cross -wire which, on the reproduced picture, gives the point on which to sight the torpedo.
It appears that it is desirable to watch from the controlling plane
not only the picture viewed by the torpedo, but also the condition of
the controls of the torpedo, the acceptance of the controlling signals,
altitude, etc. This is particularly important if the torpedo is not
launched directly at the visible target, but has to pass first through
intervening clouds. Such an arrangement can easily be achieved prac-
ICONOSCOPE
-j'
pue
,
AMPLIFIER
CONTROL INDICATOR
TILTING
MECHANISM
LE NS
Fig.
2- Iconoscope camera
for the Flying Torpedo.
tically without introducing any additional complication in the camera
circuits. All the necessary information can be transmitted on the edges
of the picture by projecting the small light spots on the sensitive
mosaic of the iconoscope. These spots are reflected by mirrors attached
either to the controlling mechanism directly, or through the medium of
small sized electric motors. This arrangement is shown in Figure 2.
In this way, the information will be given by the position of bright
spots on the edges of the picture and can be read accurately on scales
attached to tilt receiving tube in the controlling ship. Since all instruments can be set in operation while the torpedo is still attached to the
controlling ship, the function of these instruments, and therefore the
www.americanradiohistory.com
TELEVISION, Volume IV
364
preparedness of the torpedo for action, can be checked all the time and
particularly just before its launching. It is easy to provide the adjustment connection which would enable the operator on the controlling
ship to reset the instruments in the torpedo according to the reading
of the accurate instruments of the controlling ship. Due to the fact
that the actual free flight of the torpedo will take from one to a maximum of 10 minutes, this initial setting will be kept by the instruments
of the torpedo, and the readings on the scale attached to the receiving
tube will be very accurate. The appearance of the picture on the
receiver with the control indicating spots is shown in Figure 3.
ACCEPTANCE OF SIGNALS
AILERONS
;
I
I
I
I1
I
RUDDER
l
1
1
I
I
I
I
,_y,
ELEVATOR
Tel
t
I
ó
w
O
o
o
I
Q
,
o
7
T
_
--
.
)
¡,:/
)
)
J
o
o
rl
-
O
O w
°w
a n.
N
i
>
)
o
o
)
)
)
O
O
C
Fig.
I
Al
3- Sketch of a
I
N//
1
I
I
`M
I
COMPASS
1
I
5
typical scene reproduced by the Flying Torpedo.
The radio transmitter for the 90 -line picture with 16 frames per
second requires a modulation band of approximately 100,000 cycles, or
only one-tenth of what we are using in our present television system.
Such a transmitter for 10 watts output is very simple and requires a
small number of tubes. The antenna will be located on top of the fuselage and combined with the receiving antenna.
POWER GENERATOR
The voltages necessary to operate the picture transmitter as well as
the radio receiver and most of the controls can be confined to three
www.americanradiohistory.com
FLYING TORPEDO
365
potentials of 1000, 300 and 6 volts. The total power requirement will be
about 125 watts. A conventional wind -driven generator with three
commutators and automatically adjustable propeller will answer the
purpose. The shaft carrying the generator also has three drums, which
supply the power for the controlling mechanisms.
CONTROLS
To guide the torpedo from the mother plane, the controlling signals
are supplied by the second radio channel, also on ultra -short waves,
from the transmitter located on the mother plane. The sensitivity of
the radio receiver on the torpedo can be very low, on the order of 10
millivolts. Signals operating the different controls can be separated
from one another by sharply tuned filters in the output of the radio
receiver.
In order that the torpedo will respond quickly to the controlling
signals, the power necessary to move the rudder, elevators and ailerons
is taken from strums rotated by the same propeller-driven shaft which
carries the generator. Each drum has two friction bands which can be
energized by the output of the amplifying tube. This energizing is
accomplished either by a relay which tightens the grip of the band
around the drum, or directly by an electrostatic or electromagnetic field
supplied by the output of the control tube to the band. These two friction bands are connected to a corresponding controlling device, for
instance, the rudder, moving it in either of two opposite directions.
The shaft operating the rudder carries two potentiometers varying the
biases of the corresponding controlling tubes in such a way that the
tube which is energized, and therefore moves the rudder in one direction, is biased gradually negatively and the opposing tube at the same
time is biased positively. This biasing serves two purposes: First, it
allows the control to operate fastest near the neutral position and slow
down according to the prescribed formula with the increase of the
angle of rotation. It also checks the maximum permissible controlling
angle. When the controlling impulse ceases then both tubes, the one
which just functioned and the opposing one, will be in an unbalanced
state and immediately start to move the controlling element toward the
neutral position.
The banking of the torpedo will not require a separate controlling
signal because it is possible to arrange the banking to follow automatically the controlling directional signal. This is accomplished by
coupling the tube of the rudder so that the movement of the rudder will
be automatically followed by the movement of the ailerons according to
the prescribed relation. If necessary, this movement can also be made
www.americanradiohistory.com
qt*
366
TELEVISION, Volume IV
a function of the speed of the torpedo by operating the bias of the
aileron control tube from the altimeter or the rate -of-descent indicator.
In this way, the banking can always be made exactly right regardless
of the speed at which the torpedo turns.
AUTOMATIC PILOT
Where the torpedo is to be used against an objective that is not
visible from the launching point, it will be necessary to equip it with
an automatic pilot as well as remote control. The additional apparatus
necessary to accomplish this does not add greatly to its weight or complication.
Stabilization of the torpedo when it is not under remote control is
accomplished by two gyros. The construction of these gyros can be the
same as the gyro employed in the "artificial horizon" or in the "directional gyro." They are operated either by air from the Venturi tube, or
by an electric motor supplied by the main generator. The gyros ought
to be fully stabilized, but can be made much smaller than the gyros
used in the above mentioned instruments. The control of these gyros is
accomplished by means of a mirror, attached to the gyros, which
reflects a beam of light into the photocells. The photocells are arranged
in pairs serving as two arms of a Wheatstone bridge. The cells, or
rather the openings through which the light falls on the cells, are of a
wedge shape, as shown in Figure 4.
By this arrangement, the reflected light at zero position of the gyro
produces two equal impulses in both cells and therefore balances the
bridge. When the body of the torpedo turns with respect to the gyro,
the line of light begins to turn with respect to the neutral axis and
i',creases the impulse in one of the cells, decreasing it in the other. At
the limiting angle, the impulse from one cell will be zero and maximum
in the second, giving a maximum unbalance. Of course, the shapes of
the openings can be made according to any prescribed condition so that
the change in impulses can follow a desired mathematical relation
between the turn of the bomb and the condition of the electrical circuit. By using an alternating-current amplifier tuned to the frequency.
of the rotation of the gyro, multiplied by the number of mirrors
attached to it, it is possible to operate the amplifier, not only from the
actual displacement of the torpedo, but also from its first or second
derivative, thereby increasing the sensitivity of the controlling circuit.
In order to set the control to a desired condition, it is necessary only to
turn the housing with the photocells and gyro to a certain angle in
respect to the axis of rotation of the stabilizing gyro. This will upset
the equilibrium condition of the circuit and will energize the controlling
www.americanradiohistory.com
FLYING TORPEDO
367
tubes operating the controlling bands, as mentioned in the description
of the remote control of the system. When the torpedo attains the new
prescribed flying condition the circuit is again balanced and the course
of the torpedo will coincide with the neutral position of the controlling
gyro. In this way, any outside influence such as a gust of wind, air
pockets, etc., which affect the initial course of the torpedo, will be
corrected immediately by the controlling gyro.
In order to keep the torpedo under control when the controlling
signals cease, the automatic control should be adjusted to a new set of
PHOTOCELLS
GYROSCOPE
MIRRORS
DIRECTION OF
FLIGHT
LIGHT BEAM
PHOTOCELLS
DRIVE
BRIDGE
Fig.
4- Control mechanism
of the Flying Torpedo.
conditions prescribed by the controlling signals. For this purpose,
when the controlling signal changes the initial course of the torpedo,
the position of the photocells with respect to the controlling gyro will
automatically reset itself to balance the bridge according to the new
set of conditions, and therefore will be automatically established for
this new course in the controlling gyro circuit. The motion for this
adjustment of photocells can be made either by friction from the same
drums, or can be provided by separate small electric motors.
www.americanradiohistory.com
368
TELEVISION, Volume IV
METHOD OF OPERATION
When the torpedo is attached to the mother plane, provision is made
to manually adjust all the necessary instruments aecording to the more
accurate instruments located in the mother plane. Also, all the apparatus and controls of the torpedo can be checked by starting the camera
and observing the position of control indicators in the receiver.
If the torpedo is to be launched directly at the visible target, the
control is very simple and all that is necessary is to keep the center of
the picture or the cross -wire sight on the target all the time. For this
kind of operation, the automatic pilot is unnecessary; therefore, if only
this type of operation is expected, the equipment for the automatic pilot
can be omitted. If, while the torpedo is approaching the target, it
becomes clear that due to the faulty steering, cross winds, etc., it may
miss the target, it is entirely possible to steer the torpedo through a
loop, gain altitude and repeat the attempt two or more times before its
speed is lost, preventing the further repetition of this maneuver.
It may be desirable to launch the torpedo while the mother plane is
at a very high altitude, or screened from the target by intervening
clouds, with only approximate information of the position of the target.
In this case, the torpedo is launched down in a spiral glide and kept
under control of the automatic pilot and also under manual control by
observing the controlling marks on the picture. When the torpedo
descends low enough to make the target visible, then the target can be
brought into coincidence with the cross-wire sight and the torpedo
started to glide directly to the target.
A more elaborate form of this torpedo is a regular airplane equipped
with an engine and the same controls as described above. This plane
can be launched either from a small surface craft or from the shore at
a very distant target, and then controlled according to the picture
received through the iconoscope camera. This makes this new weapon
very versatile since it can be used both on the sea and land.
www.americanradiohistory.com
NAVAL AIRBORNE TELEVISION RECONNAISSANCE
SYSTEM*t
BY
R. E. SHELBY, F. J. SOMERS AND L. R. MOFFETT
Engineering Department, National Broadcasting Company, Inc.
New York, N. Y.
Summary -A high fidelity long range television reconnaissance system
developed during World War II for the Navy Department is described. The
Project Ring equipment was designed for multi -camera attended operation
at 20 frames /second, 40 fields /second, 567 lines /frame interlaced and utilizes
a 5 megacycle video bandwidth. A high power (1400 watt peak) airborne
television transmitter is employed. The maximum plane -to-ground transmission range attained during tests was over 200 miles. Very consistent
operation with satisfactory signal -to -noise ratio has been obtained with
this equipment at ranges of 100 miles or more with the aircraft flying at
altitudes of 7000 to 10,000 feet. The equipment differs from the light weight
simplified television gear designed during World War II for unattended
operation in guided missiles.
INTRODUCTION
)URING World War II two general types of airborne television
equipments were developed for the U. S. Armed Forces. One
i
type, known by the code designation "Block" was a simplified,
light-weight system designed for unattended operation in drone aircraft and guided missiles. A second type, described herein, was developed for long -rang, high -altitude reconnaissance operations. This
equipment was designed for attended operation, with weight and complexity considerations secondary to the production of high definition
television pictures suitable for airborne military reconnaissance. The
project under which the development was carried out was known by
the code designation "Ring".
Work on Project Ring was initiated in November 1942 when standard broadcast type transportable television pickup equipment utilizing
the type 1840 orthicon camera tube was demonstrated to representatives of the Navy Department and Marine Corps. As a result of this
demonstration, in which the pickup equipment was installed on the 85th
floor of the Empire State Building to simulate an aircraft at 1000 foot
altitude, it was concluded that the information presented by the television screen probably was sufficiently detailed to have value for airborne military reconnaissance. In order to check this conclusion, it
*
Decimal Classification: R583 X R520.
f Reprinted from RCA REVIEW, September,
369
www.americanradiohistory.com
1946.
370
TELEVISION, Volume IV
was necessary to conduct a series of actual flight tests using television
pickup equipment installed in aircraft. Accordingly, arrangements
were made by the Navy to carry out such tests at the Naval Air Station,
Banana River, Florida using a PBY-4 "Catalina" flying boat.
In making these preliminary flight tests, the use of readily available
equipment, even though designed for a different service, was dictated
by two factors -the need for a quick answer on the possibilities of
television reconnaissance, and the lack of adequate previous experience
or data on which the design of specialized high -fidelity equipment
might be based. In order to make a start, therefore, standard commercial transportable television pickup gear utilizing the type 1840
orthicon tube, was mounted in the PBY -4 aircraft. The equipment
consisted of two cameras, one in the bow position and the other in the
waist gunner's position arranged to view from the starboard machine
gun blister with the latter open. The cameras were connected via
standard multi- conductor cables to the control position, which was
located amidships in the space normally used by the navigator. At the
control position were the camera control units with their self-contained
video monitors, the master switching unit with its monitor showing
the outgoing picture, the synchronizing generator and the electronically regulated power supply rectifiers. A photograph of the control position
installation is shown in Figure 1.
In order to obtain sufficient power at 115 volts 60 cycles alternating
current for operation of the television equipment, a special auxiliary
power unit had to be installed in the aircraft. Here again, it was
necessary to use equipment which was readily obtainable. An available
5- kilovolt-ampere, 115 -volt, 60-cycle, gas- engine -driven alternator was
mounted in the compartment aft of the control position. This unit
weighed several hundred pounds and had to be dismantled and then
reassembled inside the aircraft in order to place it in position.
A low power video transmitter (60 watts peak), developed for
another application, was adapted for the experiments to expedite the
work. The receiving equipment on the ground consisted of Navy type
Block 1 television receivers feeding viewing monitors equipped with
12 -inch kinescopes. A projection type viewing monitor, utilizing refractive optics and producing a picture 18" x 24" in size, was also installed at the ground station.
The ground station receivers, having been designed for airborne
use, were operated from 28 volts direct current. The viewing monitors
were modified commercial television receivers and were operated from
the local 60 -cycle mains. The transmission standards used were the
525 line, 30 frame, 60 field interlaced commercial broadcast standards
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
371
for which the equipment (except the Navy Block I receivers) was designed. The transmitter operated on a carrier frequency of 90 megacycles with an effective bandwidth of 4.5 megacycles on each side of the
carrier. Negative transmission (maximum carrier amplitude corresponding to tips of synchronizing pulses) was used. The antenna
polarization was vertical. This polarization was chosen principally
because it is simpler to obtain uniform azimuthal radiation from an
aircraft with a vertical antenna.
A comprehensive series of flight and ground tests using this hastily
assembled airborne television system was conducted during the spring
and summer of 1943 at the Naval Air Station, Banana River, Florida,
Fig.
1- Control
position in PBY -4 aircraft installation using commercial
television equipment.
culminating in a demonstration for Navy, Marine Corps and NDRC
personnel in September of that year.
RESULTS OF INITIAL AIRBORNE TESTS
These initial tests using available commercial television equipment
in an airborne system provided information and experience on which
to base the design of specialized airborne television equipment.
This experience and information may be classified under the follow-
t NDRC -National
Defense Research Committee.
www.americanradiohistory.com
372
TELEVISION, Volume IV
ing headings:
(a) General electrical and mechanical requirements for airborne
television equipment.
(b) Desirable operating and maintenance features in airborne
television equipment.
(c) Purely technical features, such as transmitter power output,
types of antennas, practical video bandwidths, etc.
A brief summary of the conclusions drawn from the results of these
tests is given in the following paragraphs.
In regard to the general electrical and mechanical design requirements for airborne television equipment, it may be said that they are
similar to the requirements for satisfactory airborne radar or radio
communication gear. Ruggedness, compactness and light weight, as
well as the ability to operate satisfactorily over wide ranges of altitude,
temperature and humidity are required of airborne television gear.
In addition, special precautions must be taken to eliminate microphonic disturbances, as television equipment is more susceptible to this
type of trouble than either radar or communication equipment. Micro phonic effects in communication equipment can be minimized through
the use of a restricted audio bandwidth, (200 to 2500 cycles, for example), special high -level microphones and reasonable care in the
application of mechanical vibration isolators. In the case of radar,
most systems do not deal with low -level signals below 500 cycles per
second. On the other hand, an important part of the energy and information in a television signal is included in the frequency range
from 20 or 30 cycles to 1000 cycles per second and these frequencies
must be dealt with in low-level video amplifier stages. The designer
of airborne television equipment must, therefore, observe special precautions to minimize the effects of vibration and acoustical noise if
microphonic effects are to be minimized.
The initial flight tests using commercial transportable television
equipment also served to emphasize the practical operating requirements for airborne gear. The equipment should be simple to operate.
All electrical adjustments should be made at the control position, leaving the camera operators free to direct the cameras, change lenses and
adjust lens iris stops as required. All electrical controls which may
have to be changed in flight should be equipped with adjustment knobs
rather than being left as "screwdriver adjustments" as the latter are
not practical when operating in bumpy air. Also, complete equipment
bench test facilities should be provided at the ground station for servicing the airborne equipment, as only a limited amount of servicing is
possible when the aircraft is in flight. Ground test facilities should
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
373
also include an auxiliary power source for pre -flight ground checks of
the operation of the equipment prior to take -off.
Definite conclusions on the technical requirements for a satisfactory
airborne television system were drawn from the initial tests. Some of
these are:
1. The television pickup equipment preferably should include two
cameras, one with a short focal length lens for wide angle views
and the other with a long focal length lens for telephoto views.
2. All electrical adjustments should be made by an operator at a
master control position. The master control position should
provide facilities for rapid switching of the output of either
camera to the video transmitter. The master control operator
should be provided with an oscilloscope, for checking video signal
levels, and two video monitors. Switching facilities should be
arranged so that either of the monitors or the oscilloscope can
be independently switched to camera No. 1 video output, camera
No. 2 video output, the input to the modulator of the video
transmitter, or to the output of a radio frequency detector
coupled to the transmitting antenna.
3. Means should be provided for rapid checking of the depth of
modulation of the video carrier during the operation of the
transmitter and pickup equipment.
4. Interlaced scanning should be used to gain the maximum picture resolution, within the limits of tolerable image flicker for
a given video bandwidth.
5. The video transmitter should have adequate power output to
provide reliable picture transmission up to 100 miles.
6. Special precautions should be observed to eliminate the effects
on interlaced scanning of power supply frequency differences
between the airborne power supply and the ground station
power supply.
7. The type 1840 orthicon was found to be unsatisfactory for airborne operation under illumination conditions combining both
high peak scene brightness and high scene contrast. The iconoscope type pickup tube, though satisfactory in this respect, is
lower in light sensitivity than is desirable for general airborne
use. It was therefore concluded that newer developmental types
of tubes, which gave promise of improved results, should be
considered in the design of specialized equipment for airborne
use.
As a result of these tests, a study of general military television
requirements, observations and reports on other military television
www.americanradiohistory.com
374
TELEVISION, Volume IV
equipments, and a study of commercial television systems, recommendations and specifications for a set of airborne television reconnaissance equipment were drawn up and submitted to the Bureau of Ships
of the Navy Department. It was considered that a practical airborne
television reconnaissance system should provide reliable picture transmission over a distance of at least 100 miles with the aircraft at an
altitude of 7000 feet. Having settled on the range requirements, the
other features of the proposed system -keeping in mind that maximum
image resolution was required -were largely determined by the practical limitations imposed by such factors as the maximum practical video
transmission bandwidth, the minimum frame repetition rate within
the limits of tolerable image flicker and blurring (due to the relative
motion between the aircraft and objects on the ground), attainable
signal -to-noise ratios, and the state of the television art at the time.
Determination of the system standards is discussed in the next section.
SYSTEM DESIGN
The radio propagation characteristics between an aircraft at various altitudes and distances and a receiving antenna 30 feet high is
shown for vertical polarization over sea water by the theoretical
curves'. 2 of Figure 2. Figure 3, which gives calculated' values for
horizontal polarization over land, is considered to be reasonably accurate also for vertical polarization. Supplementing the above propagation studies, it was determined by measurements on practical television
receivers and the results of the initial flight tests at Banana River,
that, in this application, the minimum useable signal at the receiver
input terminals would be on the order of 50 microvolts, (5- megacycle
video bandwidth). From the above data, it was estimated that for a
100 mile range (aircraft at 7000 ft.) with a vertical dipole receiving
antenna 30 feet above the ground, the average carrier power output
of the transmitter should be in excess of 200 watts (800 watts peak).
Vertical polarization of the antennas was chosen because of its non directional characteristic in the horizontal plane; also, cancellation of
the direct path and multi -path signals is not as severe,, especially at
relatively short distances with the aircraft at high altitudes, as when
horizonal polarization is used.
Consideration of various limiting factors, such as weight, size, the
2
' K. A. Norton, "The Effect of Frequency on the Signal Range of an
Ultra -High Frequency Radio Station with Particular Reference to a Television Broadcast Service ", Statement made before the Federal Communications
Commission, Television Hearing, Report No. 48466, March 20, 1941.
2 K. A. Norton, "The Calculation of Ground -Wave Field
Intensity Over
a Finitely Conducting Spherical Earth ", Proc. I.R.E., Vol. 29, No. 12, pp.
623-639, December, 1941.
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
375
required power output, available carrier frequencies and the requirement for maximum video bandwidth led to the design of an amplitude modulated video transmitter producing an average carrier
power output of 350 watts (1400 watts peak) with a video bandwidth
of 5 megacycles (i.e. 5 megacycles above and below the carrier frequency), operable at carrier frequencies of either 90 or 102 megacycles.
/o
-\\
iilili\\11
B
_- ,,,--,--, ,
NE
LOF
P P
GNT
1
1
/
2
3
3
G T B
9/0
20
Al/LES
30 40 .50
/00
200
300
2-
Fig.
Calculated field strength intensity versus distance for the indicated
altitudes of the aircraft. (Receiving antenna 30 feet high, 200 watts radiated
power at 100 megacycles, for vertical polarization over sea water with a
ground conductivity of 4.3 mhos/meter and a dielectric constant of 81.)
Figure 4, which is a plot of actual flight test data obtained with a
developmental model of this transmitter, shows that this power output
is adequate for the required transmitting range and shows good cor-
www.americanradiohistory.com
TELEVISION, Volume IV
376
relation with the calculated curves of Figures 2 and 3. While a video
bandwidth greater than 5 megacycles could have been obtained with
this design at a correspondingly lower power output, the 5 megacycles
was chosen as the best compromise value.
Having determined the video bandwidth (5 megacycles), the next
/o
9
B
,
G
5
.
,-,_"`-
II1-\\\\1\1111
4
3
:,
:'l'
I
..,..,.
k_11MMI.1I.IrI
_\_1_-,,S'
\\\11i1
M11151111111111
2
-
=.
:.
;
=
1
I1
V
L NE
OF
3/ONT
3
2
0/
Z
3
4
5
6
7
8 8/0
20
MILES
30 40 .50
/00
Z00
300
Fig. 3- Calculated field strength intensity versus distance for the indicated
altitudes of the aircraft. (Receiving antenna 30 feet high, 200 watts radiated
power at 105 megacycles, for horizontal polarization over land with a ground
conductivity of 5 X 10 -14 electromagnetic units and a dielectric constant
of 15.
step in the design was to select the system standards to attain the
greatest practical image resolution. To secure the maximum amount of
information in the television image, it is desirable to operate at the
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
37'1
lowest frame repetition rate practicable. However, the frame rate cannot be reduced below the point where blurring occurs due to the relative movement between the aircraft and objects on the ground or where
image flicker becomes troublesome. In addition, interlaced scanning
is dictated if maximum picture resolution with minimum flicker is to
Es= SIGNAL
GENERATOR (MODULATED
30Z AT 400
CYCLES)
OUTPUT VOLTAGE TO PRODUCE SAME PVC VOLTAGE
AS RECEIVED SIGNAL
/00
PI
\,
\
,i=
N
`1\'`--,
.r-...-\
blib...
mamvm\
11.
-MUMIIIIMM
,1-,____\.
M,-____`.
`\\-__,_GF
UM
\IM_IMI\--117_WWIIIIILMWM_
-la_,__,-__11111WW_1011=
EMI 4VNE-__,_,
-/=HT,
. -C.
_
....
Off
COWER L/M/T
USENBLE S/GNFL
(DUE 7b LOCAL N0/SE)
/O0,u
V.
30
60
90
/20
/50
/80
2/0
DISTANCE IN MILES BETWEEN GROUND STAT/ON AND AIRCRAFT
4-
Fig.
Measured values of relative signal input to the receiver versus distance, the aircraft flying at the altitudes indicated, receiving antenna height
30 feet, and 350 watts radiated power. (Multiply E, by 0.2 to obtain the
approximate field strength at the receiving antenna.)
be realized. As a result of a number of laboratory tests to determine
minimum frame frequency, it was found that a frame repetition rate
of 20 per second with a field frequency of 40 per second (2:1 interlaced) was the lowest practical value that could be used for this appli-
www.americanradiohistory.com
378
TELEVISION, Volume IV
The amount of flicker produced by this low frame rate is
greater than would be acceptable for entertainment television but is
tolerable for television reconnaissance. The aspect ratio of the image
was chosen to be 4:3, the same as commercial television standards,
after due consideration of military requirements. The Project Ring
standards and commercial television standards are compared in Table I.
The theoretical total numbers of picture elements per frame for the
two systems are included as a matter of interest. The theoretical resocation.
lution, however, generally cannot be fully realized in practice for various reasons, such as the relative movement between the object being
televised and the camera, aperture effects of the camera pickup tubes
and the receiver kinescopes, the effects of shot and thermal noise,
residual phase distortion in the overall system and other practical
limitations.
Table
1
Project RING
Video bandwidth-megacycles
Fields per second (2:1 interlaced)
Frames per second
FCC Standards
4.25
5
40
20
1.33
60
30
Aspect ratio
1.33
Vertical blanking -per cent of field
period
8 per cent
7 to 8 per cent
Horizontal blanking-per cent of
line period
18 per cent
16 to 18 per cent
Lines per frame
567
525
Line frequency
11,340
15,750
Vertical resolution (lines)
522
483
Horizontal resolution (lines)
500
311
Horizontal resolution X Aspect
ratio
Total picture elements /frame
667
348,174
415
200,445
In the above choice of standards, the horizontal resolution of the
system has been made approximately equal to the vertical resolution.
The horizontal resolution (lines) is:
Nh
=
2f (1
-
Tv)
(1- Th)
Nv (a) (r)
where f is the video bandwidth in cycles, Tv is the fraction of the vertical period devoted to blanking, Th is the fraction of the horizontal
period devoted to blanking, Nv is the number of lines per frame, a is
the aspect ratio and r is the number of frames per second.
To simplify the design of the synchronizing generator, as well as
to reduce its weight and size, a somewhat simplified type of syn-
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
379
chronizing signal was chosen for interlaced scanning. This type of
synchronizing signal, developed by RCA Laboratories Division,3 4 is
shown in Figure 5. This waveform is not as flexible as to the number
of different ways it may be utilized in a television receiver as is the
synchronizing signal prescribed by the Federal Communications Commission for commercial television broadcasting. The integration time
required for vertical synchronization, to secure a well-interlaced picture, is several times longer than that required by the standard commercial synchronizing waveform. However, this type of synchronizing
signal has proven satisfactory for this application and resulted in the
saving of at least ten tubes and associated components in the synchronizing generator.
J
F-
-- H--I
.EAT SYNCH EN
S
ru.a
E-II
W
2
InI
`I
L
g1
(H)
H.
(
.) oDN
()
f}amN
5"
-)
B
-®
-
(,)ODN I- -U-
4.$1
(n)DDN
Co)
.ma.
YOwS
A
V
wIETE
uwl or wcccvrvc
z"cá,OOAOi eclaoSOmO
-(Uttr
(y).oBN
L
ENLARGED DETAIL VEW OF SIGNAL IN VEW
FM
T100 room
T Of
OIL UIE
TO START OF NETT
=TARE FROM START Of ONE FELD TO STMT
or
«
NEST
BETWEEN
NE
S
1/11,340 SEC S
FELD. I /40 XL+ 3.35.
f
i (y)O1N w EES.
Fig.
oETnE .ETwEEN
.
ONO
5-Synchronizing
®
signal waveform used for Project Ring.
In addition to the above standards which have to do with scanning
and video bandwidth, it was necessary to standardize on several other
items. It was decided to use amplitude modulation and transmit both
side bands. The choice of amplitude rather than frequency modulation
was based upon the following considerations:
1. A great deal more experience was available in the use of amplitude modulation for video transmission than in the use of frequency modulation.
-
3 A.
V. Bedford, "Synchronizing in Television ", RCA Report prepared
for use by Panel 8 of the National Television Systems Committee, October,
1940.
4 D. G. Fink, TELEVISION STANDARDS AND PRACTICE,
McGrawHill Book Company, New York, N. Y., 1943.
www.americanradiohistory.com
TELEVISION, Volume IV
380
2.
3.
4.
Pre -war field tests employing FMj with vestigial side band
transmission and without limiting in the receivers gave results
that were unsatisfactory, since this system was considerably
more vulnerable to multipath transmission effects than was the
standard AM* system.
Theoretical studies and laboratory tests had been reported
showing that a FM± signal would be more vulnerable to multi path effects than an AM* signal.'
Under an NDRC contract covering work on the Block project,
comparative tests were made between FMt and AM* using airborne transmitters operating near 300 megacycles. This work
was not carried to conclusion but the preliminary results obtained indicated again that FMf was more vulnerable to multi path effects than was AM.*
Offsetting the above unfavorable experience with FM for television
transmission was the highly successful use of FM for a 500 megacycle
television relay link reported by RCA Communications, Inc.,' and other
work done utilizing FM for airborne television transmission.
In view of the above uncertainty, it was decided that the safest
course would be to employ amplitude modulation for this application.
It is emphasized, however, that this finding should not be construed as
necessarily applying in the future, particularly at higher frequencies
and with the application of more thorough development of FM for
television.
The decision to transmit both side bands was based purely upon
the desire to avoid the additional weight and complications of a side
band filter in the airborne transmitter. Under many conditions, vestigial side band operation of the receiver was utilized by proper detuning of the receiver circuits.
Negative transmission (i.e. the tips of the synchronizing pulses
corresponding to maximum carrier amplitude) was chosen for Project
Ring for the same reasons which led to its choice for commercial
television broadcasting- principally because interference produces, in
such operation, less objectionable effects in the received picture than
with positive modulation.
t FM- frequency
modulated or modulation.
or modulation.
Murlan S. Corrington, "Frequency- Modulation Distortion Caused by
Multipath Transmission", Proc. I.R.E., Vol. 33, No. 12, pp. 878 -891, December, 1945.
' F. H. Kroger, Bertram Trevor and J. Ernest Smith, "A 500 -megacycle
Radio -Relay Distribution System for Television ", RCA REVIEW, Vol. V,
No. 1, pp. 31 -50, July, 1940.
*
AM- Amplitude modulated
5
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
381
CHOICE OF POWER SUPPLY
The presence in a television receiver of power supply hum of a frequency different from the field repetition rate usually makes interlaced
scanning operation difficult. Considerable trouble from this source was
encountered in the early tests with the PBY-4 installation at Banana
River. In that case, the efficiency, and therefore the speed, of the gas
engine driving the airborne power supply would change with altitude
in such a way that the power supply frequency would be several cycles
per second lower than the nominal 60 -cycle value by the time the aircraft had reached an altitude of 5000 feet. At the same time, the local
ground station power supply, also nominally 60 cycles, would vary
above and below this value at different times of the day. This resulted
in a situation where the airborne and ground power supply frequencies
would often be different by several cycles per second. Under these
conditions, a good interlace was always obtained on the monitors in
the aircraft. The airborne synchronizing generator and therefore
the vertical scanning frequency was locked in with the airborne power
supply; consequently, any residual power supply hum in the pickup
or monitoring equipment did not show up in the airborne video monitors as it produced a steady and almost unnoticeable pattern on the
television picture. (The vertical scanning frequency automatically
remained synchronous with the airborne power supply alternator even
though the latter varied in frequency). Considerable trouble with
interlace was experienced at the ground station due to this power
supply frequency difference because the electrical filtering and shielding of the modified commercial television receivers used as monitors
was inadequate for operation from a power source not synchronous
with the vertical scanning frequency. Trouble from this source was
minimized by additional filtering and shielding of the monitor circuits
and kinescopes against power supply hum effects.
Obviously, use of pure direct current power supply sources for the
airborne and ground equipments would have been an ideal solution to
this problem. However, such a solution for the airborne equipment
would have been uneconomical because of the amounts of power required and the number of different direct current voltages needed. An
all-direct -current power supply system using individual dynamotor
operated from 28 volts also would weigh more and be less efficient than
an equivalent alternating current system.
A practical solution to the airborne power supply problem, in the
case of the equipment specially designed for Project Ring, was found
in the use of a combination of 28 -volt direct current and 115 -volt 400
to 2400 cycles alternating current power sources. The 28 -volt direct
www.americanradiohistory.com
382
TELEVISION, Volume IV
current power is used for the filaments and heaters of the tubes in the
pickup and transmitting equipment. The 115 -volt 400- to 2400 -cycle
alternating current is used for all plate and bias supply rectifiers. The
vertical scanning frequency is not locked in with the alternating current power supply, but is the 567th sub -multiple of a stable vacuum
tube oscillator operating at 22,680 cycles per second, (twice the line
scanning frequency). The vertical scanning frequency (field frequency)
therefore remains very close to 40 cycles at all times regardless of the
frequency of the airborne power supply. Adequate power supply filtering and careful magnetic shielding of the pickup and other cathoderay tubes is effective in reducing the residual power supply hum to a
sufficiently low value to allow satisfactory interlaced scanning operation. The use of 28 volts direct current for all filaments, except the
video monitors and the oscilloscope, was found to be very helpful in
reducing power supply hum trouble.
It was desirable to design the ground station receiving and monitoring equipment for transportable operation using a basic power supply
of 28 volts direct current. All ground station equipment filaments,
except those of the kinescope anode supply rectifiers, are operated
from 28 volts direct current. The television receiver has a built -in
dynamotor supplying plate and bias voltages, while the 4500 -volt direct
current kinescope anode supply is obtained from rectified pulses
derived from a step -up winding on the horizontal deflection output
transformer. The filament of the rectifier for the 4500 -volt supply is
fed with current at the horizontal deflection frequency supplied by a
separate winding on the horizontal deflection output transformer. Plate
supply for the viewing monitors may be obtained from a 28 -volt to 350 volt direct current dynamotor. The 8500 -volt direct current kinescope
anode voltage for each viewing monitor is obtained from a self-
contained single phase rectifier of conventional design operated from
116 volts 400 cycles, the 400 -cycle voltage being supplied by a 28 -volt
direct current inverter. Adequate filtering and shielding of the 400 cycle rectifier eliminates hum -troubles from this source. The ground
station equipment may also be operated from 60 cycle alternating current using an electronically -regulated plate supply rectifier for the
viewing monitor, a 28 -volt rectifier and storage battery for the filaments and a 28 -volt direct current to 115 -volt 400 -cycle alternating
current inverter for the 8500 -volt anode supply rectifier.
AIRBORNE PICKUP EQUIPMENT DESIGN
The division of the airborne equipment into
units was dictated by the following factors:
a
www.americanradiohistory.com
number of separate
TELEVISION RECONNAISSANCE
383
(a) The necessity of providing small packages of reasonable weight
and size that could be conveniently removed from the aircraft
for servicing and quickly re- installed.
(b) The desirability of designing monitors, power supplies, etc.,
as interchangeable identical units so that in case of failure of
a single monitor or power supply during flight, the rest of the
system would still be operable. (Also, a single spare of each
type unit could be carried in the aircraft for installation in
case of failure during flight.)
(c) The division of the system into a number of relatively small
units interconnected by suitable cables provides flexibility in installations in aircraft where space is at a premium, and makes
possible more uniform weight distribution.
rerrrom
`
fg,"fo
.
Fig. 6 -Block diagram of airborne television equipment.
The block diagram of Figure 6 shows the units comprising the
Ring airborne pickup and transmitting system. The video monitors
are interchangeable units as are the regulated power supplies (except
the multiplier orthicon camera power supply which provides two additional voltage outputs). Identical interconnecting plugs are used on
both types of cameras and on both types of regulated power supplies
for the video equipment. Plug connections are arranged so that any
combination of camera types can be plugged into the system. When
two image orthicon cameras are used, all video power supplies are
units of the same type.
The block diagram, Figure 6, also shows the video and synchronizing pulse interconnections between the various units. The synchronizing generator provides four pulse outputs with amplitudes and
www.americanradiohistory.com
384
TELEVISION, Volume IV
waveforms as indicated on the diagram: vertical driving, horizontal
driving, kinescope blanking, and synchronizing signals. The vertical
and horizontal driving pulses are fed via 72 -ohm coaxial cables to the
two cameras where they energize the vertical and horizontal beam
scanning circuits. The vertical driving pulses have a frequency of 40
cycles per second and the pulse duration is 6 per cent of one cycle. The
horizontal driving pulses have a frequency of 11,340 cycles per second
and a duration of 8 per cent of a horizontal scanning period. The
kinescope blanking signals are added to the picture signals in each
television camera and consist of a composite wave comprising 40 -cycle
pulses having a duration of 8 per cent of a field period and horizontal
blanking pulses (11,340 cycles) having a duration of 18 per cent of
one horizontal scanning period. The kinescope blanking pulses are
timed to start earlier and last longer than the corresponding driving
pulse components. Distribution of the blanking signals is via 72 -ohm
coaxial cables. The coaxial cable carrying the synchronizing signal
output loops through the monitors and oscilloscope which are bridged
across the line and is finally terminated in the master control unit.
The wave form of the synchronizing signal is as shown on Figure 5.
The feeding of synchronizing signals separately to the video monitors and the oscilloscope has the advantage that these units may be used
to monitor video signals which do not contain synchronizing signals.
This is the case when the monitors or oscilloscope are switched to the
camera video outputs directly, since synchronizing signals are added
to the video signals only in the line amplifier which feeds the video
transmitter. The monitors and oscilloscope are equipped with "internal external" synchronizing switches, however, and may be synchronized
from the incoming video signal by throwing the switch to the proper
position when the incoming signal contains synchronizing pulses. The
latter type of operation is desirable as a check on the proper video -tosynchronizing signal amplitude ratio when checking the transmitter
output by means of the radio frequency detector provided.
It will be noted in Figure 6 that the master control unit receives
video signals from the two cameras as well as from a radio frequency
detector. The detector is coupled to the antenna transmission line and
extracts and rectifies a small percentage of the transmitter output
energy for monitoring purposes. This detected signal is fed to the
master control unit via a terminated 72-ohm coaxial line. The control
unit provides video monitoring outputs to the cathode -ray oscilloscope
and the video monitors. It also houses a line amplifier for feeding the
output of either camera to the transmitter: Synchronizing signals are
added to the output of this line amplifier.
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
385
By means of push buttons provided on the control unit, any monitor
or the oscilloscope may be independently switched to Camera No. 1
output, Camera No. 2 output, the control unit line amplifier output, or
the output of the radio frequency detector. The front panel of the
control unit is also the location of all electrical controls for the two
cameras.
While the control circuits used are conventional, the video switching arrangement is thought to present novel features and will be
described in some detail. This can best be done by reference to a
simplified block diagram of the video and switching portion of the
8
GANGED INPUT
GAIN CONTROL
CAU
I VIDEO
OUTPUT
DRIVING
CHORINP UT
72 OH
72 OHM LINE
\
TRANSMITTER
INPUT SWITCHING
Iy
CAM .2 VIDEO
OUTPUT
SYNC INPUT
\IVARIAGLEI
2.o
72 OHM LINE
KEYED oSYNC
SIGNAL BLACK LEVEL
CLAMP
LIN AMPLIFIER
LINE
Illy
LINE AMPLIFIER
CUT PUT TO
TRANSMITTER
72 OHM LINE
*Cg
72 OHM LINE TO
MONITOR N0.3
72 OHM
LINE
AMPLIFIER
I
Fig.
7- Master control unit.
I
(Simplified video block diagram.)
master control unit, Figure 7. It will be noted that Camera No. 1 video
output passes through four monitoring networks in cascade, the last
network being terminated in a resistance of 72 ohms in the form of an
output gain control potentiometer. The Camera No. 2 output, line
amplifier output and radio frequency detector output also pass through
sets of monitoring networks in cascade. By means of groups of mechanically- interlocked push button switches indicated in vertical rows on
Figure 7, the No. 1 monitor, No. 2 monitor, or the oscilloscope can be
switched to any of these four outputs independently, the interlocks
being arranged to prevent any monitor or the oscilloscope from being
bridged across more than one network at a time. Only three input
www.americanradiohistory.com
TELEVISION, Volume IV
386
points are provided for No. 3 (utility) monitor as there is no occasion
to use this monitor to look at the radio frequency output during operations. A line amplifier with bridging input and 72-ohm output is provided so that No. 3 monitor can be placed at any desired location in the
aircraft and fed from a 72 -ohm cable.
The operation of the monitoring filters is shown in more detail in
Figure 8. It will be noted that during normal operation, with the
switch in position A, the shunt capacitance consists of C2 and C3 in
parallel. When the switch is thrown to position B to feed the video signal to a monitor, an equivalent capacitance to C3 consisting of C4 and C5
in parallel replaces C3. The capacitance C5 represents the value of the
stray capacitance in the connecting lead and input to the monitor, while
C4 is a trimmer capacitor adjusted to take care of different lengths of
monitoring cables in various installations. This monitoring network
CI
VIDEO INPUT
WITCH
.MONITORING
OUTPUT
C
Fig.
R
TERMINATION
gigC5
8- Monitoring network.
is a constant resistance, bridged -tee type utilizing negative mutual
inductance. Values which give constant resistance, approximately zero
attenuation and linear phase are :
Ll
0.158R
=
M=
f
- 0.050R
f
Cl=_
C2
+
0.029
Rf
0.315
C3
=
Rf
Where R = surge impedance of the cable
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
387
f = top video frequency component considered
M, C1, C2 and C3 as in Figure 8.
further requirement is that the shunt resistance of the monitoring
tap be high compared to the surge impedance of the line. This requirement is easily met in the usual monitoring input which has a direct current resistance of 250,000 ohms or more and an equivalent shunt
input capacitance, including several feet of coaxial input cable, of 100
micro -microfarads or less. This type of monitoring filter has been
successfully used in the commercial television operations of NBC for
the past several years. It has economical advantages over the use of a
low- impedance video bus for monitoring purposes. It can also be designed to fit in a small space, the overall dimensions of the shield
housing for the filters of Figure 7 being 11/2 x 6 x 12 inches approximately.
As indicated on Figure 7, the line amplifier feeding the transmitter
also incorporates a keyed "black level clamp" circuit.' This circuit
performs the function of maintaining the tips of the video blanking
signals at a constant direct -current level at the grids of the cathode
follower tubes feeding the transmission line. Low frequency surges
caused by switching the input of the line amplifier from one camera
to another are thereby ironed out and prevented from affecting the
transmitter. While this is not a new idea, it will be described here in
some detail since it is also used in the television cameras to discriminate
against microphonic disturbances.
Figure 9 shows the basic circuit of the keyed black level clamp. In
operation, the diodes D1 and D2 of V3 are periodically rendered conducting by the application of keying pulses supplied by V2. The keying
pulses are of line frequency and are timed to occur during the "back
porch" portion of the video planking pedestals. Application, simultaneously, of positive pulses to the plate of D1 and negative pulses to
the cathode of D2 provides a low resistance discharge path for the
coupling capacitor C1. The potential on the grid side of C1 is therefore
brought back to a fixed value ( -C in this case) at the end of each
scanning line. By adjustment of R3 and R4 to provide keying pulses of
equal amplitude and some adjustment of the relative values of R1 and
R2, the keying pulses can be balanced out so that only a small residual
pulse signal appears at the grid of V4. Provided that the peak -to -peak
video signal at the point where the clamp is applied amounts to several
volts, the residual unbalanced pulse signal can be made so small as to
have negligible effect on the output of V4.
A
7
K. R. Wendt, U. S.
Patent No. 2,299,945, October,
www.americanradiohistory.com
1942.
TELEVISION, Volume IV
388
CAMERA DESIGN
As already indicated, two types of television cameras were designed
for Project Ring. The PH- 536 /AXS -1 camera utilizes the Image
Orthicon,' a super- sensitive pickup tube producing a satisfactory picture with 1 /100th the light required by the iconoscope. The PH537 /AXS -1 camera utilizes a developmental type orthicon similar to
the type 1840, but incorporating an electron multiplier signal amplifier.
The sensitivity of this tube is intermediate between the iconoscope and
the Image Orthicon. The use of an electron signal multiplier, with the
resulting increase in effective sensitivity, makes this tube useable over
a greater dynamic light range than is possible with the type 1840.
OUTPUT
CATHODE
FOLLOWER
CI
DI
___102
o
V.
+B
72011M LINE TO
TRANSMITTER
D
HORIZ. DRIVING
PULSE INPUT
PEAK WHITE
BACK
BACK PORCH
LEVEL
SYNC.
SIGNALS
il
RELATIVE
(TIMING
HORI2
DRIVING
PULSES
T
Fig. 9 -Keyed black -level clamp circuit.
Photographs of these two camera types are shown in Figures 10 and 11.
Functionally, these cameras differ from the usual commercial designs in that they incorporate within a single housing all the tubes
and components needed to produce a complete video signal (except the
synchronizing signal generator) . Their video outputs have a peak -topeak voltage level of 1.0 volt into a 72-ohm load and are complete with
video blanking pedestals. The chief advantage of this arrangement is
that there is a saving in overall weight of the equipment as compared
8
Albert Rose, Paul K. Weimer and Harold B. Law, "The Image Orthi-
con-A Sensitive Television Pickup Tube ", Proc. I.R.E.,
pp. 424-432, July, 1946.
www.americanradiohistory.com
Vol. 84, No. 7,
TELEVISION RECONNAISSANCE
Fig.
10 -Image
Fig.
389
orthicon camera with viewfinder. (Hinged cover opened to
show video amplifier -top access door open.)
11- Multiplier
Orthicon camera with 7% inch f/2.5 lens.
www.americanradiohistory.com
390
TELEVISION, Volume IV
to the commercial practice of locating camera deflection amplifiers and
the main video amplifier remote from the camera proper. The image
orthicon camera is 10% x 93/4 x 211/4 inches in size and weighs 46
pounds, less lens and lens mount. The camera utilizes 21 tubes, of
which nine are dual types. The Multiplier Orthicon camera is 101/4 x
151/2 x 203/4 inches in size and weighs 74 pounds. This camera utilizes
21 tubes, of which eight are dual types.
Since the subjects to be televised are always at a distance, in normal
operation, the camera lenses are locked at infinity focus for this application. A simple viewfinder such as the combination ball sight and
plano- convex lens with cross-hairs, as shown in Figure 10, may be used.
A cathode- ray -tube viewfinder using a green zinc -orthosilicate fluorescent screen with a green optical filter has also been found to be
convenient for airborne television use. The No. 3 (utility) video monitor has been used for this purpose in connection with the nose position
camera where operating space is restricted.
In addition to special selection of non-microphonic tubes for the
first video amplifier stages and careful design of the vibration isolators
on which the cameras are mounted, keyed black level clamp circuits are
used to eliminate microphonic signals from the camera video outputs.
The early video amplifier stages of both types of cameras are purposely
designed to have a drooping gain characteristic for signals below 500
cycles. This is followed in each case by a stage incorporating a black
level clamp at a point where the video signal is at a sufficiently high
level so that microphonic disturbances produced by subsequent stages
are negligible. Video stages following the black level clamp, which
effectively "restores" the low frequency components of the signal, are
designed for uniform response from 20 cycles up to the top video frequency.
Another circuit artifice which counteracts microphonics is the use
of a value of signal load resistor which is high compared to the effective shunt capacitance existing at the output of the pickup tube electron multiplier. The resulting loss in video signal output and the phase
shift at high frequencies is compensated for in one of the higher level
video stages by use of a "high -peaker" circuit. This results in a greater
video signal input to the thermionic amplifier at low and medium frequencies than if the load resistor were chosen for more uniform
response over the entire video band. A form of "high -peaker" compensation used in the Ring equipment is shown diagrammatically in
Figure 12. Since both the electron multiplier and the amplifier tube V1
can be considered as constant -current generators, and since the gain
versus frequency characteristic of the intermediate video stages is uni-
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
391
form, the overall response of the system is proportional to the product
of the impedances of the reactance arms consisting of R1 and C1 in
parallel and R2 and L2 in series. These reactance arms became inverse
networks (Z1 Z2 = A2 = Constant) when the values of L2 and R2 are
chosen so that:
L._,
R1
C1=
R2
When this is done, the phase and amplitude distortion produced by the
R1C combination at the input to the amplifier can be exactly compensated over a wide band of frequencies within the limitations imposed
by resonance of L2 with the interstage shunt capacitance C2. As a
practical matter, satisfactory compensation can be obtained by choosing
L2 small enough so that the parallel resonant frequency of L2 and C2 is
well above the top video signal frequency of the system. A ratio of
resonant frequency to top video frequency of 1.5:1 was found to be
satisfactory.
HIGH
PEAKER
I
PICKUP TUBE
NTERHEDIATE
VIDEO STAGE
CL_D
VIDEO°
AMP.
MULTIPLIER
LOAD
RESISTOR
Fig. 12-Block diagram illustrating "high- peaker" correction network.
By use of such a compensating system, the orthicon low- frequencysignal output fed to the grid of the first thermionic amplifier tube can
be increased many times over the value obtained with a flat amplifying
system. This results in a worthwhile discrimination against micro phonic disturbances originating in the thermionic amplifier portion of
the video system.
AIRBORNE VIDEO TRANSMITTER
The video transmitter was based upon a preliminary design worked
out for this project by RCA Laboratories Division. It was designed
and constructed to provide for either of two types of installation. The
low -power installation -100 watts average carrier or 400 watts peak
carrier output
comprised of the transmitter unit, transmitter power
supply unit (units on left side of Figure 13), remote control box, modulation indicator unit, antenna coupling unit, antenna, shock mounts,
-is
www.americanradiohistory.com
392
TELEVISION, Volume IV
wavemeter, test meter and cables. The installed weight is 200 pounds
and the power required is 14 amperes at 28 volts direct current and 14
amperes at 115 volts 400/2400 cycle alternating current. Cooling air
at the rate of 200 cubic feet per minute is required. The high power
installation-350 watts average carrier or 1400 watts peak carrier output-consists of the video transmitter, power amplifier and two power
supplies shown in Figure 13 plus the associated equipment, cables and
Fig. 13 -The video transmitter and linear power amplifier. (These are
shown at the top, mounted together as a unit on a special shock mounting.
The remote control box and modulation indicator unit are shown mounted
below the transmitter. The lower two units are the transmitter and power
amplifier power supplies.)
antenna. The installed weight is 400 pounds (30 -pound cathode -ray
oscilloscope not included). The power required is 40 amperes at 28
volts direct current and 45 amperes at 115 volts 400/2400 cycle alter-
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
393
nating current. Cooling air at the rate of 800 cubic feet per minute is
required. The air for cooling the transmitter and power amplifier is
supplied in flight by an air scoop mounted on the fuselage of the aircraft. An auxiliary cooling blower is used for ground testing. The
high -power transmitter is shown in Figure 13.
The radio frequency section of the transmitter contains the following tube and circuit arrangements A type 826 oscillator tube is operated in an ultra -audion (colpitts) type oscillator circuit, at a plate
potential of 750 volts, at either 90 or 102 megacycles. The oscillator
is very loosely coupled to two type 3E29 pentode tubes, connected in
parallel push -pull, operating as a buffer stage. The buffer drives a grid modulated stage through a four -terminal inductively -coupled band -pass
circuit. The secondary of this network is heavily damped to provide a
band -pass of 85 to 107 megacycles and also to improve the voltage regulation of the buffer stage. Provision of this wide bandpass eliminates
the need for retuning for operation on either of the two carrier frequencies, 90 or 102 megacycles. Four type 4E27 pentode tubes, connected in parallel push -pull, are operated at a plate potential of 1100
volts as a grid -bias amplitude -modulated Class C stage. The output of
this modulated Class C amplifier unit may be fed to an antenna system
through an antenna coupling unit or may be used to drive the linear
power amplifier unit used in the higher power installation.
The first stage of the video modulator section of the transmitter
consists of two type 6AG7 pentodes with the plates connected in
parallel. Input connections to each of the grids are provided so that
synchronizing signals may be fed to one grid, and video signals to the
other as required in some applications. When being fed mixed video
and synchronizing signals from the pickup equipment described here,
the two modulator inputs are operated in parallel. A video pre- emphasis
network in the first video stage provides for the selection of several
degrees of high -frequency peaking as may be required to compensate
for sideband attenuation in the modulated radio -frequency section of
the video transmitter. The second video stage consists of a type 3E29
pentode tube operating in a conventional video amplifier circuit. The
video modulator stage contains two type 3E29 tubes operated in
parallel. The video output signal of the modulator, approximately 120
volts peak -to-peak, is applied in series with the grid bias supply of the
amplitude -modulated radio -frequency amplifier. High- frequency compensation is effected in the video stages by conventional means, i.e.,
combination (series shunt) peaking and LCR (inductance- capacitanceresistance) networks in the cathodes.
The Class B linear power amplifier unit contains two type 827R
:
www.americanradiohistory.com
394
TELEVISION, Volume IV
air- cooled screen -grid tubes connected in 'push-pull and operated at a
plate potential of 2400 volts. The coupling to this stage is effected
through a heavily damped four -terminal band -pass network. An LCR
two -terminal network is inserted in series with the grid of each 827R
tube. The combination of the tube impedance and this equalizing network provides a reasonably constant input impedance to the 827R stage
between 85 and 107 megacycles. An LCR network is also placed in
series with each filament lead of the 827R tubes. The capacitive reactances of these networks are adjusted to equalize the filament lead
inductances over the range of 85 to 107 megacycles, thus effectively
placing the filaments at radio frequency ground potential. The output
of the power amplifier stage is inductively coupled to the 50 ohm transmission line and antenna. A 35Z5 diode tube connected as a detector
and bridged across the transmission line provides a video monitoring
voltage that can be used to check the modulation and overall performance of the transmitter. A direct -current milliammeter connected in
the diode rectifier load circuit indicates the transmission line voltage
or relative power output of the transmitter. Meters are also provided
to read total grid, screen and plate currents of each tube of the transmitter.
The power supply for the low -power section of the transmitter contains five full -wave rectifiers and supplies the following direct -current
voltages: 350 volts to the video modulator section, 350 volts to the
3E29 buffer stage, 1100 volts to the oscillator and 80 volts for bias.
Xenon gas type 3B25 rectifiers are employed in all rectifier circuits
except for the 5Z4 full wave high vacuum rectifier used for the bias
supply. The Xenon gas rectifiers are preferred to the mercury vapor
type for this application, since they will operate in any position and
also over the wide temperature limits encountered in aircraft installations.
The ripple filtering circuits are conventional with the exception
of
the 1100 -volt supply which contains a 50-ohm constant impedance
network. This LCR network has constant impedance to all modulating
frequencies up to 5 megacycles. The power supply unit also contains
a thermal 40-second delay relay tube for timing the application
of high
voltages and the necessary fuses, interlock circuits, and contactors.
The power -amplifier power supply consists of eight 3B25 tubes
connected in a bridge rectifier circuit and supplies 2400 volts for the plates
and 1400 volts for the screens of the 827R tubes. A 50 -ohm constant
impedance network is used for filtering.
The modulation indicator unit, see Figure 13, is used in connection
with a cathode-ray oscilloscope to give a visual indication of the per-
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
395
tentage of modulation of the transmitter. The video signal from the
radio-frequency-detector is fed to the modulation indicator unit and
thence to the cathode-ray oscilloscope. The modulation indicator units
contains a mechanical vibrator which short-circuits the detector load
circuit at a rate of 300 -400 times per second. The time of short- circuit
of the video line (zero voltage) corresponds to zero power output of
i
,Ib
Ilid
.
,11iiii
IIIIIIIIIIII IIIIII111111 11
025 V.
SETUP APPROX.
OUTPUT OF CONTROL UNIT LINE AMPLIFIER
FEEDING TRANSMITTER AS SEEN ON OSCILLOSCOPE
USING 20 CYCLE SWEEP
VIDEO
((<<f
fÍjfr
VJII
iu
IN I1111111111
.
90? Mo
111011
100°
SYNC.'
MODULATION INDICATOR OUTPUT AS SEEN
20 CYCLE SWEEP.
ON THE OSCILLOSCOPE USING
Fig.
14- Signal
level and modulation check using oscilloscope.
the transmitter. The tips of the synchronizing signals, as seen on the
oscilloscope fed by the detector, correspond to maximum carrier amplitude, as the transmitter is operated with an input level such that the
Y
T. J. Buzalski, "A Method of Measuring the Degree of Modulation of
a Television Signal ", RCA REVIEW, Vol. VII, No. 2, pp. 265 -271, June,
1946.
www.americanradiohistory.com
TELEVISION, Volume IV
396
tips of the pulses just begin to be compressed in the output of the
modulated stage. The percentage of modulation of the carrier may
therefore be quickly estimated visually on the oscilloscope screen as
shown in Figure 14B. The dots at the top of the trace in Figure 14B
occur when the contactor is closed and correspond to zero carrier while
CONNECTOR FOR
COAXIAL LINE
ZOO
50
OHM
OHMS
-
,
-
\\
I.
,
y
:s
METAL
SKIN OF AIRCRAFT
ØLIØ
Om
` `
`
=
.
8
`_
fW "f
`_S
6:
W. _'S
s
W
r' .ÿ
fll INME
M. WI
y
`W.
lea 0
\_`
:
162
"STYRAMIC"
DIELECTRIC
O.i I
s
NaMag
MOM
NIMI
I
'WW1
a
IblIMIT
01
NOV
Mr
4
Fig.
DIA. BRASS SLEEVE
DIA. STEEL ROD
15- Transmitting
antenna.
the tips of the synchronizing pulses at the bottom of the trace correspond to maximum carrier.
The transmitting antenna follows a design by Mr. P. S. Carter.
This antenna, shown in Figure 15, provides satisfactory operation on
carrier frequencies of 90 or 102 megacycles without readjustment.
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
397
Fig. 16- Measured. radiation pattern of relative field intensities of a vertically polarized quarter -wave antenna mounted below forward bomb bay of
JM -1 aircraft model, scaled 30 to 1 and measurements made at 3,000
megacycles.
www.americanradiohistory.com
TELEVISION, Volume IV
398
The location of the antenna on the aircraft was determined on the
basis of scale model tests performed by the engineers at Rocky Point,
Long Island. Figure 16 shows a plot of data obtained in this way for
the location finally chosen. As indicated, this provides a reasonably
uniform horizontal radiation pattern.
AIRCRAFT INSTALLATION
The installation of the Ring equipment in a Navy JM -1 "Martin
Marauder" aircraft, as used in developmental flight tests, is shown in
Figures 17 and 18. The total installed weight including the two television cameras, control and monitoring equipment, the transmitter,
video and transmitter power supplies, power and video cabling, cable
Trannitter Control
Picture
Picture U. 1
Picture
Box
Ibdulation Innutar Ont
Sontag on bullhead mart to
nitre M.
Tt+ualtter
Operators
Sut
2
Video
Tncsaltter
Poorer Supp
Powr
Srqpl,
Cathode
Oq
Oscilloscope
portion Vannte areas
.here equipment la looted
Shadod
Gttoda
Pour Amplifier
Powe
Panel
Shatterproof
Plate Glass
Valet klatches
Orthicon
Coen
Carr.
Yonlwe
Trwaltte
Tnewitte
Power
Master
1ap11f1e
Gntro1
Snit
Synchronizing
Generator
prge
lloecopm
hu-puni.
Pour
Gntnl
.7
Ose
Syqi,
1. C.
Junction
Box
Opmnt rs
ennui
Junction
Box
Pullinn
Tin
of
Pew
Fig.
17- Phantom
view of the location of the various units in the JM -1
aircraft. The transmitter is located in the after bomb bay, and the antenna
is located just forward of the transmitter beneath the fuselage.
junction boxes, equipment shock mountings, an assortment of camera
lenses, spare tubes, meters, tools, etc. is 1400 pounds. These figures
do not include the engine- driven generators and their voltage regulators.
Power for operating the equipment in flight is obtained from two
Navy type NEA -8, 7.25 kilovolt-ampere 115 -volt, 400- to 2400 -cycle
alternators, one mounted on each aircraft engine. Direct-current
power at 28 volts for operating the video and transmitter filaments is
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
399
obtained from two 200 -ampere generators, one on each engine, operated in parallel. The power required for the video pickup and control
equipment is 2.2 kilovolt-amperes at 115 volts 400 to 2400 cycles and
19 amperes at 28 volts direct current. The transmitter requires approximately 5 kilovolt -amperes at 115 volts 400 to 2400 cycles and 40
amperes at 28 volts direct current. The transmitter and video equipments are supplied from separate alternators. These alternators, operated on separate engines, are not synchronous but owing to adequate
power supply filtering no trouble has been experienced from beats
between the frequencies of the two power sources. At normal cruising
speed, the alternator frequencies are in the vicinity of 1000 cycles.
Fig. 18-Photograph of the master control position in the aircraft. (The
picture monitors and cathode -ray oscilloscope (with calibration unit attached) are shown mounted above the synchronizing generator, master
control unit and video switch panel. This position provides complete adjustment and control of the transmitted picture-including "on the air" monitoring of the picture, check of per cent of synchronizing signal transmitted
and measurement of the modulation percentage of the transmitter.)
Power changeover switches and a power connection plug are provided for operation from an auxiliary power sources for pre -flight testing and maintenance when the aircraft is on the ground.
www.americanradiohistory.com
TELEVISION, Volume IV
400
GROUND STATION EQUIPMENT
The R -90 /UXR -2 television receiver shown in Figure 19 is continuously tunable over a range of 76 to 116 megacycles and accepts a
video band extending 5 megacycles above and 5 megacycles below the
carrier. The receiver sensitivity is such that an input signal of 35
microvolts with 40 per cent modulation produces an output signal with
unity signal -to-noise ratio. The signal -to -noise ratio is expressed in
terms of the ratio of peak -to-peak video signal output to peak -to-peak
output noise. A half -wave vertical dipole antenna matched to a 72 -ohm
coaxial line is normally used to feed this receiver. While the design
of this receiver follows conventional superheterodyne principles, it
incorporates a very important feature in the form of a fast -acting
automatic volume control operated from the peak value of the detected
synchronizing signals. This automatic volume control serves to iron
Fig. 19
-Type
R -90 /UXR -2 television receiver.
out the fluctuations in received signal which occur due to addition or
cancellation of the direct wave from the airborne transmitter by waves
received over an indirect path.
The receiver provides an output of 1 volt peak -to-peak composite
video signal at an impedance of 72 ohms for feeding one or more of
the 12 -inch viewing monitors shown in Figure 20.
TEST RESULTS
Numerous flight tests of the Ring equipment installed in a JM -1
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
401
aircraft were conducted at the Naval Air Station, Willow Grove, Pennsylvania and at the Naval Air Test Center, Patuxent River, Maryland,
between November 1944 and July 1945 when Navy altitude and acceptance tests were completed.
In one test to determine the range of the transmitter, an acceptable
television picture was received over a path length of 205 miles with
the plane carrying the transmitter flying at 23,000 feet over Putnam,
Connecticut and the receiver located at the Naval Air Station at Willow
Grove, Pennsylvania. A vertical half -wave dipole at a height of 50
feet above the ground was used at the receiving end. A monoscope was
used as a video signal source in the aircraft for this test so that the
maximum transmission range could be determined independently of
conditions of weatherand visibility.
On another occasion with the aircraft flying over Philadelphia at
an altitude of 10,000 feet, observers at the Naval Air Test Center at
Fig. 20 -Type ID- 86 /UXR -2 12 -inch viewing monitor.
Patuxent River, Md., 120 miles away, were provided with views of the
city buildings, railroad yards, ships in the Delaware River and other
objects of interest. Using a 20 inch f /10 lens on the Image Orthicon
camera, which provides a viewing angle of only 3.4 degrees, telephoto
views of the Philadelphia- Camden Bridge showing moving traffic in
sufficient detail to distinguish between automobiles, trucks and buses
were received at Patuxent. Alternate overall views provided by the
www.americanradiohistory.com
402
TELEVISION, Volume IV
Multiplier Orthicon camera equipped with a 7.5 inch f/2.5 lens (17.5
degree viewing angle) were also transmitted. After circling the Philadelphia- Camden area, the aircraft proceeded to Washington, D. C., via
Baltimore, Maryland, providing television views of ships, oil refineries,
etc., arriving over the Capitol at an altitude of 17,500 feet. Using the
Image Orthicon tube with the 20 inch f /10 lens, at this altitude,
moving trucks, buses and automobiles were easily distinguishable.
Circling down to 5,000 feet over Washington views of the National
Airport and the Naval Air Station at Anacostia were transmitted to
Patuxent, a distance of 55 miles, in clear enough detail so that expert
observers were able to tell the number and types of planes parked on
the airfields.
The altitude from which objects on the ground can be picked up by
the television cameras is naturally controlled by conditions of illumination and visibility. The extent to which weather conditions govern the
operating altitude is similar to that existing in the case of aerial
photography. As in aerial photography, various optical filters, such
as the Wratten No. 23, have been found helpful in improving picture
contrast in the presence of ground haze. The type LM -15 Image
Orthicon tubes used in these tests had considerable sensitivity in the
near infra -red light region. The use of a Wratten No. 89A infra -red
filter produced improved results with this tube in dealing with ground
haze under some conditions.
ACKNOWLEDGMENT
The design, development and testing of the Ring equipment was
the cooperative effort of a number of NBC television engineers under
the direction of Mr. O. B. Hanson, Vice -President and Chief Engineer,
Mr. R. E. Shelby, Director of Technical Development and Mr. G. M.
Nixon, Assistant Director of Technical Development. Among the
engineers taking an active part in the work were Messrs. R. M. Fraser,
C. L. Townsend, Edward Wade, E. Stolzenberger, R. A. Monfort, W. L.
States, E. C. Wilbur, A. W. Protzman, C. W. Turner and W. C. Resides.
Preliminary transmitter design drawings were supplied by Mr. T. L.
Gottier of RCA Laboratories Division ; final transmitter development
and construction were handled by Messrs. T. J. Buzalski, W. L. States
and A. L. Hammerschmidt of NBC. Mr. F. J. Somers was the project
engineer on design, development and engineering of the Ring equipment. Mr. H. P. See was in charge of installation, field testing and
technical liaison with the Navy Department. In addition to the participation of the engineers already mentioned, substantial assistance
was obtained from a number of groups and individuals in various
www.americanradiohistory.com
TELEVISION RECONNAISSANCE
403
divisions of RCA. Administration of the several contracts under which
the project was carried out was by the Government Development Section of RCA Victor Division; these contracts were NXss- 20596,
NXsr -47375 and NXsr -66811 between Radio Corporation of America
and the U. S. Navy. Lt. Comdr. L. R. Moffett, USNR, was in active
charge of the project for the Bureau of Ships of the Navy Department.
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION EQUIPMENT *t
BY
R. D. KELL AND G. C. SZIKLAI
Research Department, RCA Laboratories Division,
Princeton, N. J.
Summary -A developmental television camera, designed especially for
airborne applications and using the image orthicon, is described. This
camera is part of a complete airborne television transmitter system weighing 50 pounds. The transmitter has a power output of eight watts in the
260 -to- 380 -megacycle range. Experimental results in guiding a mediumangle bomb with the aid of the miniature equipment are given.
INTRODUCTION
it became apparent that
extended application for television might be found if the transmitting equipment could be made smaller and lighter than the
then-existing field equipment. One particular application for a miniaI[)URING the course of the World War II,
Fig.
1
-The
"Roc bird" (Courtesy of Douglas Aircraft Company, Inc.)
ture television equipment was a new medium -angle guided -bomb type
of missile developed by the Douglas Aircraft Company, Inc., and known
as the "Roc ". The "Roc bird" is shown in Figure 1. After a preliminary study of miniature tubes and other components, it was decided to
develop miniature television equipment for this project. The system
was to consist of a small cylindrical camera unit placed in the nose of
*
Decimal Classification: R583 X R560.
Reprinted from RCA REVIEW, September, 1946.
404
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION
405
the missile, a small transmitter and power supply placed in the after
part, and a dipole antenna placed on the rear of the missile.
Some of the preliminary work was based upon tube operating conditions so severe that the life of the tubes would be materially shortened. The object was to get maximum performance for a short time
with minimum apparatus, since, in use, the entire unit was expended
after a few minutes of service. This design consideration was applied
in the case of numerous expendable electronic apparatus, but due to
the complexity of the equipment, it was decided that the advantage of
the slight weight reduction, obtainable by severely overworking the
tubes, was offset by the disadvantages of frequent tube replacements
Fig.
2
-The
Mimo tube.
during development and the difficulty of adjusting and testing the
equipments with the identical tubes to be used in service.
THE MINIATURE IMAGE ORTHICON
Since the space for the television camera unit was limited, a special
pickup tube was developed for this project.' The tube is shown in
Figure 2. It is called the "Mimo" tube (miniature image orthicon) .
The tube is 11/2" in diameter and 9" in length.
P. K. Weimer, H. B. Law, and S. V. Forgue, "Mimo- Miniature Image
Orthicon ", RCA REVIEW, Vol. VII, No. 3, pp. 358-366, Sept., 1946.
1
www.americanradiohistory.com
TELEVISION, Volume IV
406
THE CAMERA UNIT
The camera unit is shown at the right in Figure 3. An airborne
iconoscope camera of earlier design is shown at the left for comparison
of their sizes. The requirement for a cylindrical camera unit was met
by mounting the tubes and components on three disc -shaped chassis
surrounding the focusing coil and camera tube, as shown in Figure 4.
The chassis, focusing coil, deflecting coils and alignment coils are
assembled on and in a steel tube which also supports the lens mounting. The cylindrical case of the camera unit is so constructed as to be
airtight, providing normal atmospheric pressure for the circuits re-
t
Fig.
3
-The
BC -1212 iconoscope camera and the Mimo camera.
gardless of altitude. The required controls pass through special
vacuum -tight bushings in front of the camera unit case. The lens end
of the case is covered with a flat disc of optical glass treated on both
sides with non -reflecting film.
The schematic diagram of the camera unit is shown in Figure 5.
The video amplifier uses type 6AK5 tubes. It has a frequency characteristic which is approximately flat to 4 megacycles. The third videoamplifier stage has a conventional "high peaking" grid input circuit
which compensates for the high-frequency attenuation of the orthicon
output circuit. The grid of the fourth video -amplifier stage is
"clamped" by a 6AL5 duo-diode to black level; thus the low- frequency
components, lost in the small coupling capacitors preceding this stage,
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION
407
are reinserted.' Also, the clamping removes amplifier microphonics with
the exception of those components near or above the line frequency of
14,000 cycles. Pulses for the clamp circuit are obtained from the balanced horizontal pulses appearing across the horizontal deflecting coils.
Video blanking is added in the plate circuit of the fourth video stage.
The cathode output stage acts as a clipper, setting a level of 0.3 volts,
thus tending to limit the video output to 0.6 volts peak -to -peak.
The vertical deflection system, which operates at 40 cycles, consists
of a 3A5 blocking oscillator and discharge tube (tube No. 8) and another 3A5 tube (No. 7) with both sections operating in parallel as the
final amplifier. Blocking oscillators are used for driving both vertical
and horizontal scanning circuits. The oscillator transformers are constructed with sma:1 mu -metal cores. The vertical speed of 40 cycles
Fig. 4 -Mimo camera with case removed.
remains constant to approximately 1 /10 cycle for supply voltage variations from 21 to 24 volts.
The 14- kilocycle horizontal deflection circuit consists of a 3A5 dual triode (tube No. 1) operating as a blocking oscillator and as a discharge
tube, a 25L6 power pentode (tube No. 2) as the final amplifier and
one-half of a 3A5 dual triode (tube No. 14) as a horizontal deflection
regulator. This regulator, which uses the 50 volts across an NE -2 tube
as a reference value, increases the plate voltage of the horizontal dis2 R. D. Kell and G. C. Sziklai, "Image Orthicon Camera ", RCA REVIEW,
Vol. VII, No. 1, pp. 67 -76, March, 1946. (See page 70).
www.americanradiohistory.com
408
TELEVISION, Volume IV
charge tube and thus tends to increase the deflection if the +335
voltage decreases. This cancels the reduction of deflection which would
occur otherwise and provides substantially constant amplitude of horizontal deflection. This regulator also serves to keep the high voltages
derived from the deflection -flyback voltage constant.
The horizontal oscillator has a regulated power supply,; but its
heater supply varies directly with the primary power sources.. Under
this condition the deflection frequency tends to increase with decreasing battery supply voltage. To counteract this effect a special compensating bias arrangement is used in the grid circuit. The grid leak
is supplied with a positive voltage obtained partly from the 26 -volt
battery and partly from the regulated +150 volts. A drop in the +26
volts tends to cause the speed to decrease. The bleeder resistor for
combining the two voltages is proportioned to make this effect cancel
the effect of the heater temperature change.
The core of the horizontal output transformer is two small loops
of hypersil core material. The secondary winding is shielded and
carefully balanced to reduce pickup from the beam -deflecting yoke to
the video input. Resistance -capacitance damping is used. The positive
high voltages for the image- orthic4 electron multipliers are supplied
by a 1654 rectifier. The photo -cathode and ring voltages are supplied
by a 6AK5 (tube (No. 4) in a constant -voltage rectifier circuit which
is also self-regulating, using the regulated +150 volts from tile power
supply unit as a reference. This circuit operates on the same principle
as that described for the large orthicon camerae, and provides a very
well regulated source of -350 volts for the photocathode and ring of
the camera tube.
The first half of a 3A5 (tube No. 6) is used to produce a wide vertical blanking pulse. This is mixed with horizontal pulses in the first
half of a 3A5 (No. 14) to provide kinescope blanking. The two pulses
are also mixed in the second half of a 3A5 tube (No. 6) to provide
orthicon blanking. Three volts of orthicon blanking are supplied to
the target at a direct -current level variable from about -11/2 to +11/2
.
volts.
The video input circuit picks up an appreciable amount of undesired
horizontal pulse voltage, mostly from the target blanking. This is
partially neutralized by mounting near the input capacitor a terminal
with a few volts of the opposite-polarity horizontal pulse on it, obtained
from the clamp circuit. The remainder of the pickup is removed by
blanking. The focusing coil and the alignment coil, connected effectively in parallel, are supplied with current from an .Amperite regulator.
Horizontal and vertical sync (synchronizing) pulses are supplied
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION
.4
www.americanradiohistory.com
409
TELEVISION, Volume IV
410
through the two halves of 3A5 tube No. 5. Both pulses are delayed in
their respective grid circuits, providing a "front porch" on each blanking pulse. Since the plates, which are in parallel, operate at a low
voltage, the tube tends to limit the horizontal pulses during the vertical
sync pulse. Four volts peak-to -peak of sync signal appears on the output when it is coupled into a 150 -ohm load.
The cathode of the image orthicon is at ground potential. Voltage
for its bias adjustment is obtained from the negative supply of the
dynamotor. The "wall" voltage for the tube is provided by the regulated +150 -volt supply.
The camera unit has ten potentiometer adjustments. Four of these
may be adjusted with the sealed cover in place. They are the image orthicon grid bias, "wall" voltage or beam focus, photocathode or image
Fig.
6
-The
BC- 1212 -T3 and the Mimo transmitter.
focus, and video gain. The other adjustments are the two centering
controls, the photocathode ring voltage, vertical linearity, alignment coil current, and target voltage. No adjustments are provided for the
deflection amplitudes and speeds, these being permanently adjusted
during test.
THE TRANSMITTER
The Mimo transmitter unit is shown at the right in Figure 6. An
earlier airborne television transmitter is also shown to indicate the
relative size. The schematic diagram of the transmitter is shown in
Figure 7. The transmitter is tunable between 260 and 380 megacycles
and has a power output of 8 watts. Figure 8 is a side view of the
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION
411
transmitter, showing the relative location of the oscillator and the
power amplifier. Figure 9 is a view of the bottom of the chassis, showing the master oscillator. Figure 10 shows the top of the chassis, with
the power amplifier and modulator tubes.
The master oscillator is a 2C43 lighthouse tube, with a resonant line tuning circuit and Colpitts feedback. The feedback capacitor is
adjusted for the proper plate current, which may be measured across
a 10 -ohm resistor in the cathode. A tuned link circuit feeds the grid
----
--011r---
HH
9004
4v4
TulC1
(
VIDEO
?
1
1
w
Fig. 7 -Mimo transmitter diagram.
circuit of the push -pull power amplifier. The link circuit is fed through
the chassis and it has a short -circuiting bar which is ganged with the
short -circuiting capacitor of the oscillator. The power amplifier consists of two 2C43 lighthouse tubes in push -pull neutralized with a dual
ceramic trimmer. The amplifier is grid modulated by two 6V6 beam
tetrodes in parallel. The video and the sync signals coming from the
camera unit are mixed at the grid of the first video amplifier (6AK5).
www.americanradiohistory.com
TELEVISION, Volume IV
412
Each input terminal is terminated by a 150 -ohm resistor and the 4-volt
peak -to -peak sync signal is attenuated fifteen times before mixing.
This circuit also attenuates the undesired flow of the video signal
toward the camera, thus preventing cross -modulation which would
cause parts of the picture signal to act as spurious sync pulses.
The second video amplifier has an "unbypassed" 1000 -ohm variable
cathode resistor acting as a gain control. With the gain control fully
closed, the 0.7-volt peak -to -peak video input from the camera provides
approximately 50 volts peak -to -peak across the plate lode of the 6V6
Fig.
8 -Mimo
transmitter chassis,
side view.
Fig. 9 -Mimo transmitter chassis,
bottom view.
tubes, corresponding to a modulation in excess of 90 per cent. The
video amplifier is flat out to 4 megacycles.
A 9006 diode coupled to the antenna provides a monitor signal or
a direct antenna tuning indication to a plug -in meter. The plate and
grid currents of the power amplifier can also be measured by means of
a plug -in meter.
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION
Fig.
10 -Mimo
413
transmitter chassis,
top view.
POWER SUPPLY UNIT
The power supply is shown on the right -hand side of Figure 11.
This may be compared in size with the dynamotor and junction box
of the earlier airborne equipment in the same photograph. Figure 12
Fig.
11
-The
BC- 1212 -T3 dynamotor and junction box and the Mimo
junction box including the dynamotor.
www.americanradiohistory.com
TELEVISION, Volume IV
414
is a view of the Mimo power supply showing the Amperite regulator
for the focusing coil in the camera, the 150 -volt regulating tube, the
electrolytic capacitors and the dynamotor. The schematic diagram of
the power supply is shown in Figure 13.
ANTENNAS
The project involves the simultaneous operation of two radio links,
namely, the Mimo link by which the picture signal is transmitted from
the "Roc bird" or missile to the control plane, and the radio link by
Fig.
12- Internal view of the
Mimo junction box.
which the control signal is transmitted to the missile. This requires
four antenna installations which will be referred to as Mimo-Roc,
Mimo-plane, control -plane, and control -Roc, respectively.
The desired characteristics of the antennas are prescribed by certain operational and tactical conditions which will be discussed. In
the first place, adequate coupling between the sending and corresponding receiving antennas must be maintained for every likely position of
the "Roc bird" with respect to the launching plane from the time of
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION
415
dropping until impact. Since the plane is always roughly at the rear
of the bird, the Roc antennas should have maximum radiation in this
general direction to provide a favorable signal -to -noise ratio. Also,
these antennas should have negligible radiation in a forward or downward direction in order to avoid strong signals being reflected from
the ground. This is particularly true of the Mimo antennas because a
television picture is inherently very susceptible to multipath reception.
Also, in this case, the effect of multipath reception would be made worse
by the Doppler effect due to the high velocity of the bird.
Obviously the antennas on the plane should have their maximum
r
DYNAMOTOR
- ---
-1
owi .QFU
I
D
l
L--
RLO
-
--
W..T
-vR-bo-30
Y
a
CAMERA
pu[ mor
Yu. w+T
GRw-
Fig.
13 -Mimo
T
power supply diagram.
radiation in a downward direction in order to bracket the Roc missile
during its fall to earth. Also, adequate coupling between sending and
receiving antennas should be retained even when the plane turns. It
is not expected that the directivity of the plane antennas can be of
any value in excluding multipath reception.
Secondly, it was required that it be possible to adjust the Mimo
antennas to match the impedance of the lines at any spot frequency in
the band (260 to 380 megacycles) and have a several -megacycle- bandwidth in order to transmit a picture of adequate detail.
Both the directivity requirements and the fidelity requirements are
www.americanradiohistory.com
TELEVISION, Volume IV
416
somewhat less severe for the control link than for the Mimo because
the signal transmitted is of a simpler character. However, the control
frequency (84 megacycles) is several times lower so that more space
would normally be required to obtain a prescribed antenna performance.
At the start, the required size of the control antenna presented a serious problem in the "Roc bird" because of the danger of interfering
with the aerodynamic performance of the missile. After considerable
study, a satisfactory solution of the problem was found by insulating
Fig.
14
-The
brake ring and dipole antennas on the "Roc bird."
the brake ring, which is primarily part of the aerodynamic equipment,
and using it as the control -Roc antenna. This brake ring, however,
affects the radiation pattern of the Mimo -Roc antenna. On this account
it was found necessary to devise a method of effectively grounding the
brake ring at the Mimo frequencies, and at the same time insulating
it at the radio- control frequencies. Figure 14 shows the Mimo dipole
transmitting antenna and the brake ring used for the receiving control
antenna on the rear of the "Roc bird."
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION
417
The supporting rods of the brake ring are made up as coaxial lines
which are a quarter wavelength long at the Mimo frequency, and thus
ground the brake ring to the body of the "Roc." The impedance of these
lines is high at the control frequency, and consequently insulates the
brake ring from the body of the "Roc" at this lower frequency. Figure
15 shows the radiation pattern of this assembly in the elevation and
azimuth planes. The reduction of secondary lobes from the Mimo antenna was of great importance in reducing the interference that reception of the radiated energy could cause in the control receiver. The
brake ring is connected at a point 90 degrees with respect to the Mimo
antenna, as shown in Figure 14, thus reducing the coupling between
the two antennas to very low value.
The Mimo -plane antenna consists of two dipoles, identical in conI.0
./
.9
8
E LE
VAT ION
AZI MUT H
NOSE
o
180160 I4Ó 120° 100°
Fig.
80°
60° 40°
20°
15-Radiation patterns of
0
20° 40°
60° 80°
10O° 120°
140° 160`180°
the "Roc" dipole at 310 megacycles.
struction with those used on the "Roc." They are mounted on the underside of the plane with centers spaced 0.417 of a wavelength at mid band, and oriented so that the extended axes of the dipoles intersect at
90 degrees at a point equidistant from the two dipoles. The physical
layout is shown in Figure 16. The dipoles were fed equally and in
phase.
COMPARISON OF WEIGHTS OF EARLIER AIRBORNE TELEVISION AND
MIMO EQUIPMENTS
While photographs give a fair indication of the size reduction
accomplished in the Mimo design, it may be interesting to note that
www.americanradiohistory.com
TELEVISION, Volume IV
418
the weight of the total equipment was cut in half. The weights of the
various components of the two systems are as follows:
Camera Unit
Transmitter
Power Supply Unit
Shock Mounts, Cables, etc., approx.
Earlier Airborne
Television Equipment
331/4 lbs.
26% lbs.
211/2 lbs.
181/2 lbs.
Total
100 lbs.
Mimo 3
20 lbs.
7 lbs.
15 lbs.
8 lbs.
50 lbs.
Fig. 16-Two dipoles as Mimo -plane antenna.
PERFORMANCE
In the early flight tests, made in the East, the camera and transmitting equipment was mounted in an AT -11 type aircraft. The signal
was received on the ground. The camera unit was mounted in the
plastic nose of the plane on a tiltable platform. The transmitter and
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION
419
power supply were in the rear of the plane and the antenna was
mounted on the underside of the plane. A single horizontal dipole
mounted on a wire screen was used as the receiving antenna at the
ground station. On several occasions very satisfactory signals were
obtained from a distance of 20 miles with the plane at 10,000 feet altitude. A useful signal was obtained at the same altitude even 40 miles
away, when the nose of the plane was tipped down, thus taking greater
advantage of the antenna directivity.
In actual drops of the "Roc" missile, of course there were no distances of such magnitudes involved. There were, however, several other
difficulties encountered in obtaining good signals from the missile. The
most important among these difficulties was a predominant Doppler
effect, due to the multipath reflections from the ground (in spite of the
directivity of the antenna) and the high speed of the "Roc."
When the source of a radiation is moving, an apparent wavelength
will be observed according to the relation:
c
_
±vt
(1)
f
where A' is the apparent wavelength, c is the velocity of propagation,
vt is the velocity of the radiating source, and f is the frequency. The
sign is positive when the path increases, and it is negative with a
reduced path. Since
C
f=-
(2)
A
the frequency f' observed at the receiver is:
C±
f' =
v,.
(3)
À'
where v,. is the velocity of the receiver. Substituting the value of
from (1), we have:
f' =
A.'
c-_*vr
c
±vi
f
(4)
Assuming two straight paths, one increasing as the bird travels
from the plane, and one reflected from the ground decreasing as the
bird approaches the ground, two signals will be received producing a
beat frequency according to the relation:
of=fi'-f:'=f
(C+Vr
C+v,
C+v,
C-vt
www.americanradiohistory.com
TELEVISION, Volume IV
420
Of=f
Since
c2
»
2c vt
(5)
vt2,
of-f
2v
,
(6)
c
assuming
a missile velocity of approximately 500 miles per hour, or
224 meters per second and a carrier frequency of 300 megacycles,
A
f = 448 cycles per second.
Fig. 17 -Two "Roc birds" on a B -17 plane (Courtesy of Douglas Aircraft
Company, Inc.)
Another difficulty encountered was the microphonics caused by the
control surfaces. The microphonic problem, while considerably less
serious than the Doppler effect, aggravated the situation. Since these
two difficulties produced similar effects in the form of horizontal bars,
the causes were not easily separated. In the course of actual drops
these difficulties were reduced to the extent that very satisfactory pictures were obtained for guiding purposes.
Figure 17 shows two "Roc birds" attached to a B -17 plane. Figure
18 shows a photograph of a test target in the western United States.
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION
Figures
19 -21
421
are enlargements from 16- millimeter moving picture
films taken from the television picture received at the control plane.
In this particular drop the plane flew at 19,200 feet, which was 15,000
feet above the target area when the missile was dropped. The vertical
bars caused by the Doppler effect are noticeable in all three pictures,
particularly in Figure 21, which is taken less than a second before
impact, but they do not destroy the value of the information. The
reproductions through the 16- millimeter motion picture film enlarging
and finally the printing process destroy much of the detail and clarity
of the picture appearing on the monitor, but even from these repro-
Fig.
18- Photograph
of a test target.
ductions, the value of the information for guiding the missile may
easily be seen.
ACKNOWLEDGMENT
Acknowledgment is made to the engineers of the Douglas Aircraft
Company, Inc., and RCA Laboratories Division who cooperated on the
www.americanradiohistory.com
422
TELEVISION, Volume IV
Figs. 19- 21- Progressive photographs of the monitored picture.
www.americanradiohistory.com
MINIATURE AIRBORNE TELEVISION
423
project. In particular, special credit is due Messrs. R. R. Thalner and
K. R. Wendt, who helped in the design as well as the tests, to M. A.
Jackson of NBC, who assisted in the drop tests, and to Dr. G. H. Brown
and Mr. J. Epstein, who designed the antenna systems. The work
described in this article was done in whole or in part for the Office of
Scientific Research and Development under Contract OEMsr -441 with
Radio Corporation of America.
www.americanradiohistory.com
MIMO- MINIATURE IMAGE ORTHICON': t
BY
PAUL K. WEIMER, HAROLD B. LAW, AND STANLEY V. FORGUE
Research Department, RCA Laboratories Division,
Princeton,
Summary-A miniature image orthicon, known as the "Mimo" tube, has
for use in airborne television equipment. Its reduced size, and
power requirements permit a substantial reduction in the dimensions and
weight of the pickup -tube camera. The Mimo incorporates an improved
mounting technique and employs additional fine mesh screens in front of
the photocathode and target for the purpose of shaping the electric fields
and simplifying operation. The resolution and signal -to -noise ratio of the
Mimo approach that of the larger image orthicon at high light levels under
carefully controlled conditions. Performance considerations as a function
of the tube size are discussed.
been developed
INTRODUCTION
MILITARY applications for a miniature television camera have
prompted the development of a pickup tube consilIërably
smaller than any of the tubes in commercial use. At the
same time, the aim was to approximate the performance of the larger
tubes. Of all of the well -known types of pickup tubes, the image
orthicon because of its high sensitivity and high signal level output
was most suited for scaling down. Accordingly, a miniature image
orthicon called the "Mimo" tube has been designed for use in airborne
television equipment.' Figure 1 shows a comparative photograph of the
Mimo tube and an image orthicon.#
The mechanism of operation of the Mimo tube is essentially the
same as the image orthicon whose cross- sectional diagram is shown in
Figure 2. The optical image is projected on the semi- transparent photocathode laid on the inside surface of the glass envelope. The resulting
photoelectrons are focussed by the uniform magnetic field, and they land
at high velocity on the thin, semi- conducting glass target. Since the
secondary emission ratio of the glass is greater than unity, a positive
charge pattern is built up on the glass corresponding to the light and
Decimal Classification: R583
Reprinted from RCA REVIEW, September, 1946.
1 A.
Rose, P. K. Weimer, and H. B. Law, "The Image Orthicon
Sensitive Television Pick -up Tube ", Proc. I. R. E., Vol. 34, pp. 424 -432, July,
*
-A
1946.
2 R. D. Kell and G. C. Sziklai, "Miniature Airborne Television Equipment", RCA REVIEW, Vol. VII, No. 3, pp. 338 -357, Sept., 1946.
# Throughout this paper the term "image orthicon" will refer only to
the tube described in Reference 1.
424
www.americanradiohistory.com
MIMO
Fig.
425
1- Comparison of the Mimo tube with
an image'orthicon.
shade in the optical image. The target screen collects the secondary
electrons from the glass and serves to limit the maximum potential to
which the glass may rise. A low- velocity beam scans the other side of
the glass target and deposits sufficient electrons in the positive areas to
drive the glass down to the potential of the thermionic cathode of the
gun. The conductivity of the glass is so chosen that the charge is conDECELERATING RING
(ZERO)
SECONDARY
ELECTRONS
CATHODE (ZERO)
SECONDARY
ELECTRONS¡-
ELECTRON IMAGE
DEFLECTION YOKE
PHOTO- CATHODE
SIGNAL OUTPUT
ELECTRODE
( -300V)
(.1500V)
ALIGNMENT COIL
Fig.
TARGET SCREEN
(ZERO)
TWO-SIDED TARGET
2-Cross -sectional diagram of the image orthicon.
www.americanradiohistory.com
TELEVISION, Volume IV
426
ducted through the glass in a frame time while lateral leakage is
negligible. Excess beam electrons not deposited on the glass form a
modulated return beam which is directed into a five stage multiplier.
The video signal from the output of the multiplier is fed into the
camera video amplifier.
Aside from reduced size, the principal ways in which the Mimo tube
differs structurally from the image orthicon are:
(1) All electrodes including the image section are mounted on a
single assembly with all the electrical connections brought out through
a single 18 -lead stem of the miniature button type.
(2) Additional fine mesh screens are used in front of the target and
photocathode for the purpose of controlling the shape of the electric
fields.
The results of these changes are described in the following sections
along with a discussion of performance as a function of target area.
SIGNAL OUTPUT
ELECTRODE
I
MIL APERTURE
"PINWHEEL "PERSUADER°
MULTIPLIER ELECTRODE
WALL SCREEN
DYNODES
18
LEAD
MINIATURE
TYPE STEM
Fig.
THIN GLASS
TARGET
PHOTOCATHODE
TARGET
SCREEN
METAL WALL
CYLINDER
3- Cross -sectional
PHOTOCATHODE
SCREEN
diagram of the Mimo tube.
STRUCTURAL DETAILS OF THE MIMO
A cross sectional drawing of the Mimo tube is shown in Figure 3.
The overall length is 9" (as compared to 151/4" for the image orthicon)
and the maximum diameter has been reduced from 3" to 11/2 ". These
dimensions allow a substantial reduction in weight of copper, and power
required for the focusing and deflection coils, as well as the use of a
smaller lens.
In the type of assembly used in the Mimo tube the metal wall
cylinder is of thin nichrome and replaces the platinum coating used
in the image orthicon. Ceramic tubing supports all electrodes (except
within the gun) and the target connections are made to the stem by
means of wire leads pushed through the ceramic tubing. Mandrels are
www.americanradiohistory.com
MIMO
427
used to align the cylinders during assembly, making possible a more
accurate alignment than if the target structure were mounted independently as was done in previous tubes.
A gap of 180 degrees between two ceramics is left between the
target structure and the wall for convenient insertion of the glass target and wall screen just prior to sealing. A spring contact to a metal
button on the inside of the glass envelope connects the photocathode to
the proper lead in the stem.
A glass envelope of uniform diameter is used for the Mimo tube in
order to take advantage of the single unit construction without requiring the additional sealing operation over the target. This fact, combined with the elimination of the leads at the shoulder, greatly simplifies the glass blowing operations. The molded miniature type stem
which requires no basing is extremely convenient.
It was found that the performance of the multiplier was unaffected
by scaling it down from 11/2" to 1" diameter. The persuader electrode
is tied electrically to the first stage instead of being brought out on a
separate connection as before. The gun is made slightly smaller in
diameter and the defining aperture reduced to about 1 mil.
Vibration tests showed the Mimo to be structurally quite strong.
One tube was found to be operable after having been subjected to an
acceleration of 25 g's.
USE OF THE WALL AND PHOTOCATHODE SCREENS
The availability of fine mesh screens of high transmission and uniformity have made practical the use of screens for controlling the electric fields in front of the target and photocathode. These screens,
labelled S2 and S3 in Figure 3, are mounted directly on the wall and
target structures, replacing the separate "decelerating ring" and
"photocathode ring" of the image orthicon. Unlike the target screen,
S1, they are positioned far enough from a nodal plane of the electron
stream that their meshes are not superimposed on the transmitted
picture.
One advantage of using screens in this manner is that good focus at
the edges of the picture and freedom from distortion are automatically
assured without requiring separate adjustment of ring voltages.
Furthermore, the screens permit high fields in front of the target and
photocathode without requiring high voltages. Uniform landing of the
low-velocity beam at the edges of the target is easily obtained, and the
position of the deflection coil for best landing is less critical.
Another important advantage gained by the use of the wall screen
is the elimination of the multiplier shading control found in the image
www.americanradiohistory.com
TELEVISION, Volume IV
428
orthicon. The screen compels the electric field in front of the target to
become more nearly parallel to the magnetic field. This reduces the
translational effect on the beam which is the major cause of the scanning of the first stage of the multiplier by the return beam. The consequence of this reduction in scan is that the requirement of uniform gain
of the first stage is somewhat less stringent. As a result there is no
need to adjust the persuader voltage for controlling uniformity of gain.
The persuader electrode of the Mimo is connected internally to the first
stage lead.
It should be pointed out that the screens result in some loss in
signal -to -noise ratio (in some cases as much as 30 per cent). Also, the
two extra screens are potential sources of spurious signal. The wall
screen is the more critical of the two because the beam passes through
it twice. In spite of the fact that both the scanning beam and the
return beam are out of focus when passing through this screen a
spurious interference pattern simulating a mesh appears under certain
conditions. This pattern can be minimized by proper spacing of the
wall screen and the target.
In the airborne application for which the tube was designed, the
advantage of the screens in simplifying operation considerably outweighed the accompanying disadvantages.
PERFORMANCE
As
A
FUNCTION OF SIZE
The active target area of the Mimo tube is slightly more than one
quarter of that used by the image orthicon. This reduction affects performance from the standpoint of resolution, signal output and optics
of the camera lens.
1.
Resolution
Assuming that the resolution is limited electron -optically only by
chromatic aberration and the stiffness of the beam at the target, it
follows that a reduction in size of the tube, while keeping the voltage
constant, should have no effect on the number of television lines which
may be transmitted. The higher fields in the smaller tube should reduce
the spot size in proportion to the change in dimensions. Actually, other
less fundamental factors enter in to determine resolution
factors
whose contributions are not as readily scaled down with tube size.
Loss of resolution by target leakage, for example, may arise from
the volume conductivity of the glass or from the surface conductivity
caused by a conducting coating of caesium on the glass. The first cause
depends mainly on the resistivity and thickness of the glass and may be
practically eliminated by using thinner glass in the small tube.
-
www.americanradiohistory.com
MIMO
429
0.1 mil were used in the Mimo tube). Howwhen
present, will deteriorate resolution to a
caesium
leakage,
ever,
greater extent with a target of small dimensions.
The superposition of the target screen upon the transmitted picture
requires a finer mesh screen for the Mimo tube. An improvement in
maximum resolution resulted when a screen of 500 meshes per linear
inch was replaced by a screen of 1000 meshes per linear inch.
Cross talk in which the stray deflection fields disturb the paths of
the photoelectrons in the image section might be expected to scale down
proportionally with tube size. However, in the Mimo tube the deflection coil, for compactness, has been placed relatively closer to the target
than in the image orthicon. This makes the cross talk a more critical
problem. An effective solution is the use of iron wire wound over the
deflection coil in combination with a cylindrical copper shield over the
image section, but the position of the copper shield is quite critical.
Alternative methods of reducing cross talk are an iron ring on the end
of the deflection coil or a "bucking coil" over the image section fed by
a small fraction of the horizontal deflection voltage. The cross talk
from the horizontal deflection coil is more persistent than from the
vertical coil. The shortened storage time of the image orthicon type of
target, when the light is raised, rapidly erases the effect of "vertical"
cross talk but has no effect on the "horizontal" cross talk until extremely high lights are reached.
The resolution required of the Mimo tube for the airborne television
project was 250 lines at high lights, and this was easily met. (See
Figure 4 and Figure 5). A number of tubes when carefully set up
under experimental conditions with high light illumination showed
more than 500 lines. The high limiting resolution of the scanning
beam is evidenced by the fact that by under -scanning the target (to
remove the video amplifier frequency band limitation) the individual
wires of the 1000 mesh target screen can be resolved. This is equivalent
to 2000 television lines per inch. Separate tests have shown that under
ideal conditions the image section is also capable of equal resolution.
The contrast ratio near the limiting resolution is, of course, very low.
The limiting resolution of the small tube is enhanced by the use of
the wall and photocathode screens as well as by the use of a smaller
defining aperture in the gun. However, it should be pointed out that
high light resolutions approaching that of the image orthicon can be
attained only if great care is taken in selection and adjustment of the
tube.
(Targets as thin as 0.05 to
2.
Signal Output
At very low light levels, where full storage occurs, the signal output
www.americanradiohistory.com
TELEVISION, Volume IV
430
at the target is independent of tube size. This assumes that the camera
lens aperture is adjusted to give the same depth of focus.
At high lights the area of the target is important in determining
signal output. For a "close spaced" target (i.e. glass- screen spacing
less than one picture element) the signal output varies as the target
area and signal -to -noise ratio varies as the diameter. For "wide
spaced" targets (i.e. glass- screen spacing wide compared to a picture
element) the signal output varies more nearly as the diameter of the
target and signal -to -noise ratio as the (diameter). The target spacing
of the Mimo is of the order of two mils which is about the same as in
Fig.
4- Photograph of a test pattern
transmitted by the Mimo tube.
the image orthicon. Because this is somewhat intermediate between
"close" spacing and "wide" spacing, the drop in signal -to -noise ratio is
bracketed by the above limiting cases.
Another factor influencing the variation of signal with target area
at high lights is the degree of overlapping of the beam spot in adjacent
lines. Some overlapping does occur in the Mimo tube (in spite of the
high limiting resolution quoted above), and this would contribute to enhanced signal at high lights owing to the recharging of the target
between successive scans.
www.americanradiohistory.com
MIMO
3.
431
Choice of Lens
The first consequence of the smaller photocathode of the Mimo tube
is that a shorter focal length lens may be used for the same angle of
view. This results in a saving in space although a faster lens is required.
If, in addition, the lens diameter is also scaled proportionally, so that
the numerical aperture remains unchanged, increased depth of focus
is obtained at the expense of operating sensitivity.
5-
Enlarged section of a test pattern transmitted by the Mimo tube.
Fig.
(The optical pattern was projected on the photocathode at normal size
while the scanning amplitude was reduced to cover only the center portion
of the target. This procedure tests the resolving power of the tube by
reducing the limitations of the amplifier frequency response as well as cross
talk in the image section. The numbers on the pattern should be multiplied
by ten to give the resolution in television lines.)
At high light levels, in which range the signal output is substantially independent of scene brightness and in which range the Mimo
was mostly used, lens speed is of no importance. Here the shorter focal
length lens is an unalloyed gain in conserving space.
The size of the image projected on the photocathode of the Mimo is
approximately the same as that of one frame of a 35- millimeter motion
picture film. Thus a wide choice of lenses for the Mimo is at hand. An
f/2.0 lens was used in the camera but light conditions in the airborne
application were such that the lens was often stopped down to as small
as f/22.
www.americanradiohistory.com
432
TELEVISION, Volume IV
CONCLUSIONS
A useful television pick -up tube of reduced size has been developed
for airborne television purposes. This tube retains the high sensitivity
and stability under adverse lighting conditions which have characterized the image orthicon. At the same time changes in design have been
incorporated which make for simplified operation of the camera. It is
believed that the Mimo tube represents a first step toward the development of a television camera which approaches the miniature photographic camera in convenience and portability.
ACKNOWLEDGMENTS
The writers wish to acknowledge the interest and valuable suggestions of Drs. V. K. Zworykin and Albert Rose. The production of the
tube was greatly aided by the contributions of R. R. Goodrich, P. G.
Herkart, and C. S. Busanovich.
www.americanradiohistory.com
TELEVISION EQUIPMENT FOR AIRCRAFT'*
BY
M. A. TRAINER AND W. J. POCH
Engineering Products Department, RCA Victor Division,
Camden, N. J.
Summary-The design considerations involved in the development of
lightweight television equipment for airborne use are discussed in Part I
of this paper. Following this is a description of Block I television equipment
developed in accordance with these considerations. Flight testing of Block I
equipment brought to light several difficulties peculiar to the transmission
of television signals from aircraft. In Part II a number of these difficulties
are discussed as well as methods developed for minimizing them.
PART I
DEVELOPMENT AND DESIGN OF LIGHTWEIGHT TELEVISION EQUIPMENT
THE first development work on television transmitting apparatus
for use in aircraft was undertaken in 1936 under the direction
of R. D. Kell. This was a direct outgrowth of Dr. V. K.
Zworykin's memorandum to David Sarnoff of April, 1934.' Equipment
was built using the 1850 type iconoscope and was installed and tested
in a Ford trimotor airplane. Results obtained with this equipment
clearly demonstrated the potential usefulness of television for the
armed services. However, it was apparent that this equipment, although
much smaller than other commercial equipment of the same type, was
still appreciably larger and heavier than was considered desirable.
The advent of the 1848 iconoscope, a smaller tube than its predecessor, the 1850, made possible the design of television cameras of
greatly reduced size. Commercial equipment using this tube and designed specifically for field pickup use was introduced in 1939. All the
equipment associated with the camera was built into suitcase -type
units, making it convenient to transport.
At this time both the Army and the Navy began to take a very
serious interest in the military possibilities of television and equipment
quite similar to the suitcase -type commercial design was constructed
for airborne use in 1940. This equipment has been described in previous literature.' While this apparatus was useful in supplying a need
for experimental and training equipment it was not particularly suitDecimal Classification R583 X R520.
from RCA REVIEW, December, 1946.
1
V. K. Zworykin, "Flying Torpedo with an Electric Eye," RCA
REVIEW, Vol. VII, No. 3, pp. 293 -302, Sept., 1946.
2 C. J. Marshall and L. Katz, "Television Equipment for Guided Missiles," Proc. I.R.E., Vol. 34, No. 6, pp. 375 -401, June, 1946.
*
t Reprinted
:
433
www.americanradiohistory.com
434
TELEVISION, Volume IV
able for military use since it had been designed primarily to meet
commercial standards.
In 1940 a program was undertaken to develop television equipment
specifically for use in military aircraft and having the following design
objectives:
(1) Light weight
(2) Compactness
(3) Reasonably low power drain from a 12 -volt direct -current
source
(4) Reliable line -of -sight range up to 10 miles
(5) Resolution capability close to commercial standards
(6) Unattended operation possible at the camera and transmitter
(7) Overall ease of installation, operation and maintenance.
Naturally, a certain amount of compromise was necessary in order
to produce results in reasonable agreement with the above requirements. Some of the fundamental decisions made in order to achieve
these results are listed below:
(1) It was decided to place the iconoscope with its associated circuits and the transmitter in a single unit instead of two separate units. Although some difficulty was anticipated because
of feedback from the transmitter into the camera circuits it
was felt that advantages were to be gained with regard to
overall size and weight considerations as well as in simplification of the interconnection problem.
(2) The dynamotor power supply was made a separate unit so that
a dry battery supply might be substituted for the dynamotor
under certain operating conditions.
(3) The frame frequency was made approximately 40 cycles and
the line frequency approximately 14,000 cycles without definite relationship between the two as is necessary for commercial
interlaced scanning. This decision greatly simplified the problem of developing suitable synchronizing and other line and
frame frequency pulses. The choice of 40 cycles for frame frequency was a compromise. A lower frequency would have re
sulted in objectionable flicker and a higher frequency woudc
have resulted in lower overall resolution for the available bandwidth. The line frequency of 14,000 cycles was chosen to give
350 scanning lines. This resulted in approximately 275 lines
resolution in the vertical direction and 350 lines in the horizontal direction. Equal values of resolution could have been
obtained by using a higher frequency for horizontal scanning
and a consequent increase in the number of lines. However, it
.
www.americanradiohistory.com
o4
omliv
"
.., . .
,
...
..
..
Fig.
www.americanradiohistory.com
.
....
C.,.,.C,,.,.
10- Receiver-schematic
diagram.
n
www.americanradiohistory.com
le
usr
.14111
ea
|Ì
w
Celt
-}
||
Imo
¢
w§
//
]
|
(-W
c
§
!
3;
it
1
fit
....».
,I
...1.1.,..-
1,1
1
-112
4-1
r-,
;
t
1
e
:!-.',1'....
,
.
..
'111-...,.....r.r-
t
:: I )I
t
_4±I
,
ti9
f-
TELEVISION FOR AIRCRAFT
was considered desirable to keep this frequency somewhat low
that the horizontal scanning circuits would require less
power and would have lower dissipation.
The vertical synchronizing pulses were made approximately
one and a half lines long so that only one or at the most two
horizontal synchronizing pulses would be lost during the vertical synchronizing period. It was considered unnecessary to
add sufficient tubes and circuit elements to place slots in the
vertical pulse as is done with the standard Radio Manufacturers Association synchronizing signal. The timing of the
leading edge of the vertical synchronizing pulse was also made
coincident with the leading edge of the vertical blanking pulse.
This simplification of the synchronizing signals was made possible by the choice of sequential rather than interlaced scanning
and resulted in a considerable saving in tubes and circuit
elements.
The overall video band width of the system was made 4.5
megacycles and both the transmitter and receiver were designed
for double sideband operation. Obviously, single -sideband operation would have resulted in a more complex transmitter.
The decision to make the receiver accept both side bands was
based on the desirability of eliminating tuning controls in
order to simplify receiver operation. The band width chosen
was considered to be a fair compromise between the resolution
capability of the 1848 iconoscope, normally somewhat better
than 350 lines, and the number of video and intermediate frequency amplifier stages required in the camera, transmitter
and receiver in order to obtain adequate amplification. Field
tests with earlier types of television equipment indicated that
the resolution capability of the proposed system was entirely
adequate for most of the military applications proposed.
A transmitter frequency was chosen in the neighborhood of
100 megacycles. This made possible efficient tuned circuits
with lumped constants of relatively small size and also an
antenna structure not considered to be excessively large.
Picture monitoring facilities were not included with the camera transmitter unit since unattended operation was expected.
Instead a separate monitor unit employing a 7 -inch kinescope
was designed so that it could be connected to the camera transmitter with a single cable connection. When this was done, the
B+ supply to the transmitter output tube was automatically
transferred to the monitor circuit. When this connection was
so
(4)
(5)
(6)
(7)
435
www.americanradiohistory.com
TELEVISION, Volume IV
436
Fig.
1
-Block
Fig.
I equipment at
transmitting location.
2- Camera transmitter -tube
side.
Fig. 3- Camera transmitter -circuit side.
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
437
made, the power drain on the dynamotor remained approximately the same as with the transmitter load.
Other design c, nsiderations involved in the development of this
equipment will be discussed in the detailed descriptions of each unit
which follow. The three units necessary at the transmitting location
are shown in Figure 1. Figure 2 presents a side view of the camera transmitter showing most of the tubes. As níay be seen in the photograph the type of construction used resembles an "I" beam with closed
ends and results in a light- weight but very rigid unit. The center
chassis section on which most of the tubes are mounted is welded to
the "wrap around." Both side covers of the unit, which are removable,
are fastened in place by six "Dzus" type fasteners. Special spring contacts are placed on all the outside flanges of the case to insure a good
ground between the covers and the case. The front compartment houses
the lens mounting and iconoscope; the central portion of the case contains the video amplifier and deflecting circuits ; and the transmitter is
located in a narrow section at the rear of the unit.
Ten control knobs are recessed in the top of the case. These,
together with the power switch on the back of the unit may be considered as "operating" cent: ols. Their functions are as follows:
(1) Horizontal Sawtooth Shading
(2) Vertical Sawtooth Shading
(3) Horizontal Parabola Shading
(4) Vertical Parabola Shading
(5) Iconoscope Horizontal Centering
(6) Iconoscope Vertical Centering
(7) Iconoscope Size
(8) Iconoscope Focus
(9) Video Gain
(10) Iconoscope Bias
Five additional controls are accessible after removal of the right hand cover. These controls require adjustment only when associated
tubes are chr.nged. The five controls are arranged along the partition
separating the transmitter section from the rest of the unit. Reading
from top to bottom they are:
(1) Synchronizing Amplitude
(2) Iconoscope Vertical Size
(3) Iconoscope Blanking
(4) Vertical Speed
(5) Horizontal Speed
The power from the dynamotor is brought in through a plug in the
www.americanradiohistory.com
438
TELEVISION, Volume IV
rear of the unit.
A coaxial connection for a 75 -ohm radio -frequency
output line from the transmitter is also located at the rear of the unit.
Figure 3 shows the other side of the case where most of the circuit
elements are placed. In spite of the large number of components most
of them are readily accessible. Electrically there are four main groups
of circuits making up this unit. These are:
(1) The master scanning oscillators and deflecting circuits
(2) The iconoscope with its associated video amplifier, shading,
blanking and synchronization mixing circuits
(3) The transmitter circuits
(4) Power supply circuits
Considering first (1) above, this comprises that part of the schematic shown in Figure 4 (facing page 434) associated with tubes V6.
V7, V8, V9, V10, V13, V14 and V15. Cathode -coupled multivibrators
were chosen as master oscillators for both line and frame frequencies
because they produce pulse wave shapes which can easily be transformed into synchronizing and blanking pulses. Across the common
cathode resistor of this type of oscillator is found a positive pulse with
steep vertical sides, a decreasing amplitude during the pulse itself and
a constant voltage level between pulses. The wave shape of the voltage
on the output plate of the oscillator is similar to that on the cathode but
it has the opposite polarity. It may be noted that a relatively high value
of cathode resistor is used in the multivibrator circuits in order to
obtain sufficient pulse voltage on the cathodes for satisfactory operation
of the mixer circuits. This requires a positive bias on the multivibrator
grids.
An adjustment of this bias voltage on the output triode of the
multivbrator provides a convenient speed control. The circuit elements
cf the multivibrators were so chosen that the resulting widths of the
pulses appearing on the cathodes have a duration such that these pulses
are suitable for use as kinescope blanking signals. Narrower pulses obtained by differentiating the cathode pulses are used in developing the
synchronizing signal. The same cathode pulses are also fed to the usual
type of discharge -tube sawtooth generating circuits in order to create
the necessary scanning waveforms. The pulse appearing in the plate
circuit of the vertical multivibrator is used as a source of blanking
signal for the iconoscope grid.
Resistance mixing circuits are used to combine the horizontal and
vertical pulses from the multivibrator cathodes in the grid circuit of
the blanking amplifier (Section A of tube V6), and the resultant is
added to the video signal in the common plate circuit of the video
amplifier tube V5, Section A, and tube V6, Section B. The amplitude of
these pulses on the grid of the tube V5, Section B, is sufficient to drive
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
439
the grid beyond cut -off. Thus no video signal is transmitted during
the pulse time with the result that the kinescope screen of the receiver
can be made black during this period and the retrace of the cathode
ray beam will not be visible. The synchronizing pulses also obtained
from the pulses appearing on the multivibrator cathodes, are much
narrower than the corresponding blanking pulses because of the insertion of small coupling capacitors in the mixing circuit at the grid of
tube V6, Section B, which pass only the steep wave front of the pulses.
These short pulses appear in the plate circuit of the tube and are added
to the combined video and blanking signal appearing at the cathode of
the video amplifier tube, V5, Section B.
The function of the 18- micromicrofarad, additional shunt capacitance from grid to ground of the synchronizing pulse amplifier tube
V6, is to delay slightly the leading edge of the horizontal synchronizing
pulses with respect to the leading edge of the horizontal blanking
pulses thereby producing a narrow but adequate "front porch" or delay
period. The function of this delay is the same as that in the standard
Radio Manufacturers Association signal
prevent video signals near
the leading edge of the blanking signal from changing the timing of
the synchronizing signals. The same capacitor also causes a slight
delay of the leading edge of the vertical synchronizing signal with
respect to the vertical blanking signal with the same result.
As noted previously, the necessary saw -tooth wave forms for deflecting the iconoscope cathode -ray beam are derived from the pulses appearing on the cathodes of the two multivibrators by impressing them
on the grids of the two discharge tube sections of tube V10. The vertical saw-tooth is amplified in tube V9 and fed through a step -down
transformer to the vertical deflecting coils of the iconoscope yoke. A
horizontal saw -tooth appears in the plate circuit of Section B of tube
V10. The plate supply for this circuit comes from the vertical output
tube V9. Thus the horizontal saw -tooth is modulated by a vertical saw tooth thereby producing the keystone correction made necessary by
the construction of the iconoscope. This requires a greater amplitude
of horizontal scanning current when the beam strikes that part of the
mosaic nearest the electron gun. A control is not required on the
amplitude of keystone correction since the circuit constants of the
horizontal saw -tooth generating circuit were chosen to give the exact
amount of correction necessary. This keystone-corrected horizontal
saw-tooth is amplified in the horizontal output tube (V14), and the
horizontal damping tube V13 supplies the necessary damping to suppress horizontal frequency transients. A step -down transformer in the
plate circuit of tube V14 feeds the deflecting signal to the horizontal
deflection coils of the iconoscope yoke.
-to
www.americanradiohistory.com
TELEVISION, Volume IV
440
In order to obtain a direct -current second anode voltage of approximately
1000 volts for the iconoscope, the plate transformer of the
tube V14 is also provided with a step -up winding which feeds enough
horizontal signal to the high voltage rectifier tube V15 to obtain the
necessary voltage.
Since a horizontal size control would also cause a variation in the
second anode voltage which would, in turn, require a readjustment of
vertical size, it was considered more desirable to keep the horizontal
scanning current in the yoke at a fixed value and provide an adjustment of the second anode voltage supply.
This adjustment changes the deflection amplitude of horizontal and
vertical scanning circuits simultaneously so that a quick check of scanning amplitudes can be made by decreasing the second anode voltage
slightly which should normally bring the frame of the iconoscope
mosaic uniformly into view at all four edges of the picture. The second
anode voltage is made adjustable by inserting a variable resistor in
the cathode lead of the rectifier tube V15.
The second major division of the camera transmitter mentioned in
(2) above is the iconoscope and its associated video amplifier consisting of tubes V1, V2, V3, V4, and V5.
The iconoscope output resistor has a relatively high value in order
to improve the signal -to -noise ratio at low frequencies, thereby reducing microphonics and ripple. This causes a drop in the high- frequency
content of the video signal which is compensated for by the "high
peaker" coupling circuit between tubes V2 and V3.
All the video stages except the one associated with V5 were made
using both a series and a shunt inductance in the amplifier plate circuits in order to equalize the high frequency response. The single
inductance in the plate circuit of the first triode section of V5 was
adjusted to produce a broad resonance in the middle frequency range
in the neighborhood of 2.5 to 3.0 megacycles in order to equalize the dip
in response resulting from the other four circuits in cascade. In this
manner an overall flat response was obtained to approximately five
megacycles. The low- frequency phase and amplitude response, affected
mainly by the choice of coupling capacitors, grid leaks, and plate -filter
by -pass capacitors, is such that the amplifier is capable of passing a
40 -cycle square wave with very little distortion. In order to eliminate
some of the disturbing effects in the picture caused by microphonics
especially at the lower frequencies, a small series coupling capacitor
(270 micromicrofarads) is placed in series with the grid of the video
amplifier V5. This is normally short -circuited with a short length of
wire which can then be cut if noise and vibration conditions are espe-
-
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
441
cially severe. The consequent reduction in low- frequency response
attenuates most of the usual microphonic frequencies but also results
in some loss of picture intelligence. The most objectionable effect is
the introduction of rather long horizontal light streaks after dark
objects in the picture.
The potentiometers and associated circuits in the low- impedance
side of the two deflection transformers provide shading signals which
are mixed with the video signal in the iconoscope output circuit. These
signals compensate to a certain extent for the spurious signal generated
by the iconoscope itself (black spot signal). Four compensating signals are generated; horizontal and vertical saw -tooth wave forms, and
horizontal and vertical parabola wave forms. Since the two deflecting
output circuits are balanced to ground (by having the connection
between pairs of deflecting coils grounded through the low- impedance
centering circuits) the mid -position of the four potentiometers corresponds to zero shading signal. Moving the potentiometers to one side
of this point provides one polarity of shading signal, moving them to
the other side provides the opposite polarity.
The waveform of the signal appearing at the vertical transformer
secondary is a reasonably linear saw -tooth so that a single differentiation circuit consisting of R65 (220,000 ohms) and C28 (0.1 microfarad) supplies a satisfactory vertical parabola. The 330- micromicrofarad capacitor (C68) is for the purpose of by- passing horizontal
frequency pulses introduced into the vertical circuits through coupling.
in the deflecting yoke. Since the voltage waveforms appearing on the
secondary side of the horizontal output transformer are pulses, it is
necessary to use a single differentiating circuit to obtain saw -tooth
waveforms and two of these circuits in series to obtain parabolic waveforms. The series resistors in the mixing circuit used to combine all
four different waveforms are so chosen that the maximum voltages
derived from all shading potentiometers are approximately the same.
The radio -frequency oscillator and buffer circuits are also shown in
the schematic (Figure 4). V17 is the radio -frequency oscillator, the
tank circuit of which consists of L4, C108, and C109. C109 may be
removed for change of frequency and C108 is variable to permit tuning
to the exact frequency desired within the range. L4 is tapped in two
places, the center tap going to the plate supply voltage through L2 and
R40, the other tap being the oscillator output and feeding through C82
to the buffer grid. The voltage supply is connected through the circuit
of L4 to both the plate and the screen grid. This oscillator has the control grid and screen grid closely coupled together so that the tube oscillates as a low "mu" triode. The output is taken from L4 through C82
www.americanradiohistory.com
442
TELEVISION, Volume IV
to the control grid of the buffer stage V18. This buffer stage is a
radio- frequency amplifier with a tuned plate circuit consisting of L5
and C110. Connected to L5 is a link circuit which has a coupling coil
L6 feeding to the radio -frequency output stage, V19.
The radio -frequency output and modulation circuits are also shown
4. The coupling coil LO couples to L7 which is the grid tuning
circuit for the radio- frequency output tube V19. This is a double section tube having essentially two beam -power structures within one
envelope. L7 has a center -tap connection for grid -modulation and bias
for V19. The bias supply, from a 45 -volt dry battery, passes to L7
through the output circuit of the last video amplifier which contains the
video signal, synchronization and blanking. Consequently, the grid
voltages on V19 will be modulated with these various voltages. The
plates of this tube are connected to their respective ends of L8. L8 is
part of the output tank circuit which consists of L8, C58 and C57. C58
and C57 are tuning condensers ganged together and controllable by one
adjustment. The heater circuit of V19 is supplied directly from the
12.5 -volt power supply through a switch, shown as S4, which opens the
heater circuit when the monitor cable plug is inserted. L9 is a radio frequency output coil coupled to L8 and is connected to the antenna
in Figure
output terminal.
To insure that synchronizing pulses will always be transmitted at
full power, further modulation is accomplished in the plate circuit of
V19 for the synchronizing pulses only. The output from V20 consists
of synchronizing pulses applied to the grid of V21. The plate current
for V19 passes through R117, V22 and R150 to the mid -tap of L8.
V22 is a half wave rectifier tube through which the plate current of
V19 flows. Condenser C103 is connected from the plate circuit between
V22 and R150 to ground through a 4,000 -ohm resistor. This 4,000 -ohm
resistor is also in series with the cathode of V21. Between synchronizing pulses, V21 is not drawing plate current and consequently C103
will be charged to full plate voltage. When a synchronizing pulse
arrives at the grid of V21 it will draw considerable plate current,
causing current to flow through R68, making the lower end of C103
approximately 180 volts positive. This 180 volts added to the approximately 400 volts already on C103 will raise the plate voltage of V19 to
about 580 volts. The condenser C103 supplies V19 with power during
the short interval that the synchronizing pulse continues, as V22 will
not allow a discharge of this condenser back through the power supply.
Consequently, during the period of a synchronizing pulse, V19 is
driven completely to the limits of operation and definite modulation of
synchronizing pulses is secured. In order to make the rise of plate
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
443
voltage completely effective, C102 is placed in this circuit. This raises
the screen grid voltage on V19 in much the same manner as the plate
voltage was raised. Therefore, by having both grid and plate modulation for the synchronizing pulses, it is possible to hold the receiver in
synchronism under severe conditions of noise and interference.
The power supply unit consists of a dynamotor, a power -control
relay and two filter systems and is shown schematically in Figure 5.
All voltages required for operation of the camera transmitter and
monitor are obtained through connections to this unit. The dynamotor,
which comprises the means for converting the 12 -volt primary power
into the high -voltage direct -current potential required for operation of
the camera -transmitter and monitor, is mounted on the top of the unit,
thus providing adequate ventilation and ease of access for routine
service of the rotating equipment.
All cable connections required for normal operation of the equipment are completed by means of an 8- connector receptacle (cameratuir
Dr-
Fig.
5
-Power supply- schematic diagram.
transmitter connections) and a 2- connector plug (power connection).
A third receptacle is provided for connection of a remote power' control
switch if such is required.
The dynamotor is designed for operation from a 12.5 volts direct current supply. Two secondary windings and commutators are provided; one insulated for 1000 volts delivering 6.3 volts direct -current
to the filament of the iconoscope and thé other delivering 400 volts
direct -current to the camera- transmitter and to the input of a filter
system. Approximately 280 volts is supplied to the camera -transmitter
from the output of this filter. The iconoscope centering voltage is
derived from the heater supply but is filtered to keep ripple on this
supply from affecting the deflecting circuits.
In the camera transmitter unit itself are located two regulator tubes
www.americanradiohistory.com
TELEVISION, Volume IV
444
V11 and V12 in series. These tubes provide a constant voltage supply
to those circuits which are sensitive to slight changes in anode voltage.
Such circuits include the scanning oscillator circuits, the screen grids
of the video amplifier tubes, the blanking and synchronizing amplifier
tubes and the clipper tube V5. The bias supply is derived from a 45volt "C" battery. This was chosen primarily to supply the transmitter
output tube with a low impedance bias source so that grid current
would not change the bias voltage. A bleeder is used to provide the
necessary bias voltages required for the video amplifier. This bleeder
is supplied through one diode section of tube V15 so that the drain on
the battery is automatically removed whenever the heater supply is
removed. The second -anode supply for the iconoscope is derived from
Fig.
6-Monitor -with cover removed.
the horizontal output transformer and was discussed in detail in the
description of the scanning circuits.
The monitor designed for use with the camera transmitter is shown
with the transmitting equipment in Figure 1. In Figure 6 it may be
seen again with the cover removed. The monitor unit consists of a
type 1811 -P1, 7 -inch kinescope, together with its associated deflection
circuits and video amplifier. Connection of the monitor to the camera
transmitter automatically opens the heater circuit of the Type 829
output tube in the transmitter (V19), thus Compensating for the power
required by the monitor unit, and maintaining normal B+ voltage.
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
445
Four controls required for normal operation of the monitor unit
are located on the front panel, below the screen of the kinescope. These
controls, progressing from left to right are, respectively, the "horizontal hold" control, the "focus" control, the "brightness" control and
the "vertical hold" control. The "width ", "height ", "vertical linearity ",
"vertical centering" and "horizontal centering" controls, being used
infrequently, are screwdriver adjustment. They are arranged from
front to back along the right -hand side of the unit in the order named.
The monitor is provided with a removable light shield.
To obtain access to tubes and circuit components, the chassis can
be removed from the case after loosening the three "Dzus" fasteners
on the front of the unit and also removing the power cable.
The monitor was designed specifically for operation from a com-
Fig.
7- Monitor -schematic diagram.
posite picture signal of the type used to feed the transmitter modulator
and having approximately one volt peak -to -peak amplitude. It contains
a video amplifier, separating circuits, line and frame frequency oscillators and their associated deflecting circuits. A schematic diagram is
shown in Figure 7.
A 12SN7GT (V1) forms a two stage video amplifier which receives
a video signal from the monitor output of the transmitter unit and
feeds the signal to the kinescope grid. From the plate of the first
amplifier stage the signal is fed to the synchronizing separator tube, a
12SL7GT (V3). There is sufficient synchronizing signal amplitude at
this point so that the video signal drives the grids of this tube beyond
www.americanradiohistory.com
446
TELEVISION, Volume IV
cut-off and only synchronizing pulses appear on the two cathodes from
which the vertical and horizontal oscillator (V4 and V6) are fed. Both
horizontal and vertical oscillators are conventional blocking oscillators
except that speed controls are provided by adjustments of the amount
of positive bias on the oscillator grid leak ground returns.
A resistance- capacitance damping circuit is used across the horizontal deflection transformer in place of a damping tube in order to
save space, although the circuit is somewhat less efficient.
This same transformer also has a step -up winding and a filament
winding which supplies heater and plate voltage to a Type 8016 rectifier, V8. The rectified direct -current output (approximately 4500 volts)
is applied to the second anode of the kinescope. A potentiometer in a
bleeder circuit on this supply furnishes an adjustable first anode
voltage. A VR-150 -30 regulator tube (V2) supplies a substantially
constant 150 -volt potential to the video amplifier and the vertical and
Fig. 8 -Block I receiver and voltage control.
horizontal oscillators, so that gain and scanning oscillator frequency
will be reasonably constant regardless of change in supply voltage.
Figure 8 shows the receiver and a voltage control box designed
specifically for adjusting the input voltage to the receiver. In Figure
9 the top cover of the receiver has been removed to show the general
construction and arrangement of parts. It may be seen that in design
features it is somewhat similar to the camera -transmitter unit.
The receiver is completely self- contained and includes its own power
supply, designed for operation from a 12- to 14 -volt storage battery or
the equivalent. The entire unit mounts, by means of thumb -screws, on
a shock- mounting base supplied with the equipment. Plate power is
supplied by an internal dynamotor developing 330 volts (direct -current), when connected to a 12.5 -volt direct -current source. All cable
connectors and controls necessary for set -up and operation of the unit
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
447
are brought out to the front panel. Battery connections are terminated
at a polarized plug, mounted at the lower right -hand side of the panel.
The antenna connection is completed through a connector mounted at
the upper left -hand side of the receiver. The antenna circuit is designed to work out of a 75 -ohm coaxial line.
Controls requiring adjustment under normal operating conditions
Fig. 9 -Block I receiver-top cover removed.
are mounted on the front panel. Those controls requiring only infrequent adjustment are recessed below the level of the panel.
The controls mounted on the panel are the "vertical hold" control
located on the left end of the upper row of controls, the "brightness"
control at the right of the vertical hold control, the "horizontal hold"
control at the left of the lower row of controls, the "contrast" control
at the right of the horizontal hold control, and the "focus" control at
the right and below the kinescope. The power switch is mounted
directly above the focus control.
www.americanradiohistory.com
448
TELEVISION, Volume IV
The controls which are recessed are (starting at the left end of the
top row) ; the "vertical size" control, the "vertical linearity" control
and the "vertical centering" control, the "horizontal linearity" control
and the "horizontal centering" control.
The connector plug located at the right of the antenna plug is used
when it is necessary to send a signal to a remote monitor. When it is
desired to utilize the monitor signal, the small switch mounted next
to the monitor output plug must be switched to the monitor ouput
position; otherwise, the switch should always remain in the "internal"
position.
Electrically, the receiver may be separated into four main divisions:
(1) The radio-frequency amplifier, intermediate -frequency amplifier and video amplifier chassis is located at the left of the
kinescope when the receiver is viewed from the wiring side
with the face of the kinescope tube at the top.
(2) The seven -inch kinescope, Type 1811 -P1 is mounted in the
middle of the receiver. The kinescope is housed in a mu -metal
shield, within which is mounted the magnetic deflecting yoke.
(3) The deflection chassis, mounted to the right of the kinescope,
contains the circuits which separate the synchronizing signal
from the video signal, the blocking oscillators and discharge
tubes, the saw -tooth voltage amplifiers and output tubes. These
are utilized for generating the saw -tooth current wave in the
magnetic deflecting yoke which deflects the electron beam in
the kinescope. The horizontal output circuit which provides
the high voltage power supply for the kinescope is also located
in this section.
(4) The dynamotor, located at the right rear corner of the deflection chassis supplies all the low- voltage direct -current power
required for operation of the receiver.
A schematic diagram of the receiver is shown in Figure 10 (facing
page 435). The radio frequency amplifier, Type 9003 (V1), the first
detector, Type 9003 (V2), and the fixed frequency radio -frequency
oscillator, Type 9002 (V3), are mounted on a small chassis. This chassis
is so designed that it can be removed from the radio -frequency and
intermediate -frequency chassis of the receiver unit and replaced with a
unit designed to receive another carrier frequency within reasonable
limits if that becomes desirable. The radio frequency unit is followed by
a six -stage intermediate frequency amplifier (tubes V4, V5, V6, V7, V8
and V9). The first five stages of the intermediate frequency amplifier
utilize Type 6A C7 tubes and the sixth stage a Type 6AG7 tube.
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
449
The coupling circuits in the intermediate frequency amplifier circuit are "constant K" type filter sections with resistance loading on
the input or plate side, and a full shunt arm on the grid or output side.
The sixth intermediate frequency amplifier stage drives the second
detector, one section of a Type 6H6 tube (V10). The other section of
V10 is used as a limiter tube to limit the noise peaks coming through
input voltages.
The picture signal output from the second detector, after going
through the limiter, is amplified in the video amplifier tube, Typ6AC7 (V12), and the two sections of the Type 12SN7GT output tube
(V13). The first half of V13 furnishes at its cathode an output signal
for the external monitor and the video signal for the synchronizing
pulse separating system. Its plate circuit furnishes video signal to
the contrast control. The second half of V13 is used as an output tube,
the signal on its grid being obtained from the contrast control, the
plate circuit driving the grid of the Type 1811 -P1 (C7466) kinescope
(V27)
.
The automatic volume control amplifier and rectifier tube, Type
12SL7GT (V11), is also mounted on this chassis. Its purpose is to
hold the output of the set at a constant level over a wide range of signal
input voltages.
Mounted on the deflection chassis is the video signal amplifier and
direct -current setting tube, Type 12SN7GT (V14), the vertical synchronizing separator and clipper tube, Type 12SN7GT (V15), the
vertical oscillator and discharge tube Type 12SN7GT (V16), and the
vertical output tube, Type 12SN7GT (V17). Tube (V17) furnishes
saw-tooth deflection current to the vertical coils in the deflection yoke
through the vertical output transformer.
Contained also in this chassis are the horizontal synchronization
separator and clipper tube, Type 12SN7GT (V18), the horizontal oscillator and discharge tube, Type 12SN7GT (V19), the horizontal
saw -tooth voltage amplifier tube, Type 12SN7GT (V20), the horizontal
output tube, Type 807 (V21) which furnishes a saw-tooth current wave
to the horizontal deflection coils through the horizontal -output transformer, the controlled damper tube, Type 71A (V31), and the highvoltage rectifier tube, Type 8016 (V22). This tube rectifies the "kickback" voltage generated in the horizontal output transformer during
the return line time to provide high -voltage (direct- current) required
by the second anode of the kinescope.
The bias amplifier and rectifier tube, Type 12SN7GT (V24) which
amplifies and rectifies the horizontal saw -tooth voltage to provide a
negative bias voltage, and the two voltage regulator tubes, V25 and
www.americanradiohistory.com
TELEVISION, Volume IV
450
V26 (Type VR -105), which provide regulated voltage to the plates of
the several oscillators (where stability of frequency is important) and
to the synchronizing separator and clipper tubes, are also located on
this chassis.
Antennas for use with this equipment are not discussed herein, but
will be covered in a separate paper at a future date. However, one of
the frequently -used antenna types was a quarter -wave vertical rod
antenna working against two quarter-wave ground rods extending on
opposite sides at the base of the vertical radiator. A matching unit
was used to transform the impedance of the antenna to 75 ohms in
order to match the transmission line.
Tests of the overall equipment have already been described in some
detail.2 These tests involved extensive work over a long period by both
government and company engineers. Although a number of operating
difficulties were experienced (see Part II) which were corrected in
later equipment, it was found that the results obtained were in substantial agreement with the original design objectives.
More than 500 equipments of this type were built, most of them
by the production engineering group in charge of A. Wright and K. A.
Chittick. Later this group redesigned the Block I equipment in order
to make it more suitable for quantity production. The new design was
called Block III equipment and was used by the armed services under
actual combat conditions.
PART
II
TRANSMISSION PROBLEMS IN AIRBORNE TELEVISION SYSTEMS
Introduction
Early flight tests of Block 1 equipment indicated that the demand
for satisfactory performance under conditions of actual operation in
military aircraft imposed extremely severe requirements on the operating characteristics of the equipment. Normally after take -off of the
transmitting plane it was impossible to readjust any controls to compensate for changes in power -supply voltage, light conditions, signal
strength, interference, temperature, and vibration. This frequently
resulted in inferior performance because of poor picture quality, poor
synchronization or both.
Investigation showed that some of the principal defects were caused
by the following conditions:
(1) Microphonics.
(2) Power supply voltage variations.
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
451
(3) Interference caused by ignition -type noise or by other carrier
frequencies, particularly those having radar modulation.
(4) Unstable synchronization.
(5) Variation in light conditions on pickup tube.
(6) Insufficient signal at receiver.
(7) Multi -path transmission and frequency modulation of the
transmitter master oscillator.
Considerable time and effort has been devoted to these problems
and some progress has been made toward satisfactory solutions. There
follows an account of the work done to date.
Micro phonics
There are two basic methods for approaching this problem:
(a) by mechanically isolating the unit involved from noise and
vibration; and
(b) by providing electrical means for reducing the resultant effects
in the units themselves.
Where especially severe conditions are encountered, a combination
of the two methods is probably necessary in order to obtain satisfactory performance. It is hoped that the improvements developed in
electrical circuits will make it possible to operate under normal conditions without elaborate mechanical isolation.
The most troublesome microphonics originate in those tubes in the
video amplifier which are followed by maximum low frequency gain.
There has been some improvement in performance resulting from
better tube construction and it is hoped that even mq'e rugged tube
types will be available for future equipment. A considerable improvement in performance was also obtained by applying the clamp circuit
principle used in pre -war orthicon equipment to the video amplifier.
This circuit effectively allows the gain at low frequencies in the video
amplifier, where microphonics are normally most troublesome, to be
reduced almost to zero but does not interfere with the reproduction of
low- frequency signal components. Circuits of various types employing
this fundamental principle were tried and flight- tested and a relatively
simple circuit easily adaptable to Block equipment wuus finally evolved
In Figure 11 are shown modifications of some of the circuits in a
Block I camera -transmitter unit. One of these modifications is the
addition of a clamp circuit in the video amplifier. The double diode
V23 receives pulses from the horizontal output transformer in such
a
manner that positive pulses are impressed on the plate of one diode
and negative pulses on the cathode of the other diode. The negative
pulses are also fed to the iconoscope grid in order to provide a con1.
www.americanradiohistory.com
TELEVISION, Volume IV
452
stant reference voltage from the iconoscope during the horizontal
blanking period. A complete explanation of the operation of the clamp
circuit may be found in an article by C. L. Townsend .3
More effective shock mounts than the original design, were developed by the Robinson Company and contributed to a marked reduction
in noise signal originating in microphonic tubes. However, it was
apparent that under some conditions microphonics originated from
noise reaching the tubes through the walls of the unit itself. It was
possible to reduce this effect considerably by enclosing the unit in a
separate container lined with sound absorbing material. In such cases
it was also necessary to provide forced ventilation in the unit to prevent overheating.
`,
Fig.
11 -Block I
camera transmitter modifications.
Power Supply Voltage Variations
Occasionally the primary direct-current voltage source supplying
the equipment was reasonably constant but frequently, especially in
aircraft, the supply voltage was subject to considerable variation.
This made the proper adjustment of equipment quite difficult since the
voltage with the plane engine stopped or idling was usually very much
lower than that obtained in flight. In order to overcome this difficulty
some type of voltage regulator appeared to be essential. Automatic
voltage regulator circuits, commonly used in other types of equipment
2.
3 C. L. Townsend, "The Clamp Circuit," Broad. Eng. Jour., Vol. 12,
No. 2, pp. 5 -8, Feb., 1945; No. 3, pp. 5 -7, March, 1945.
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
453
for the B+ supply, were adapted for use in Block equipment. In
addition to providing a constant output voltage over a considerable
range of primary supply voltage, this circuit has the advantage of
providing a low impedance source for the various electrical circuits in
the equipment, thereby eliminating to a large extent the necessity for
large electrolytic by -pass capacitors normally used for preventing
coupling at low frequencies between various circuits.
Some undesirable effects caused by variation in the primary voltage
supplying heater power to the equipment still remained. It was possible to reduce these effects considerably by the use of devices such
as ballast lamps, which provide a relatively constant current for a
wide range of input voltage. This combination of a regulated B+ supply and ballast lamps on the low voltage supply resulted in satisfactory
performance but the efficiency of these circuits was rather poor because
of the power loss in the regulator tubes.
Recently it has been possible to obtain dynamotor power supplies
which will supply a constant output voltage for the B+ supply over a
considerable range of primary voltage change. These have the advantage of being much more efficient than the other devices described but
it may still be desirable to use a regulated B+ supply in order to obtain
a low impedance source for preventing crosstalk. Since this B+ regulator circuit does not need to be designed to accommodate a large
variation of input voltage, the tubes can be designed to operate much
more efficiently than in customary regulator service. Another possibility is the use of a circuit quite similar to the normal regulating
circuit which will take out low frequency disturbances and provide a
l'w impedance source but which will not regulate the direct -current
voltage itself.
3.
Interference
On some of the first flight tests it was discovered that radar equipment operating near the Block equipment carrier frequency would
completely eliminate the picture signal appearing on the receiver. This
would occur when sufficient interfering signal reached the automatic
gain control circuit at the receiver to decrease the gain of the intermediate- frequency amplifier enough so that none of the desired signal
was visible on the kinescope. A long series of experiments and flight
tests were made in an effort to reduce this effect as much as possible.
For radar -type interference it was noticed that while the amplitude
of the signal was large, the energy content was quite small. This characteristic was used to improve the operation of the automatic volume
control. Under normal conditions the automatic volume control is con-
www.americanradiohistory.com
454
TELEVISION, Volume IV
trolled by the amplitude of the synchronizing pulses appearing at the
output of the receiver second detector. Since the energy content of the
synchronizing pulses is normally much greater than that of the radar
pulses it was possible to alter the constants of the automatic volume
control circuit in such a way that it operated to a large extent on
signal energy rather than peak amplitude. This type of circuit is called
the low impedance type automatic volume control circuit and is illustrated in Figure 12. It is capable of permitting satisfactory operation
of the receiver under interference conditions from radar at least 100
times as severe as was formerly possible. Another type of "low impedance" automatic volume control circuit is illustrated in Figure 13.
The input circuit to the automatic volume control diode is tuned to
the line frequency in order to discriminate between the synchronizing
pulses intended for rectification and the interfering radar pulses.
This circuit, however, showed little improvement over the preceding
one since shock excitation of the tuned circuit by radar or noise pulses
also contributed to the automatic volume control bias developed.
A still further improvement was made possible by introducing
pulses from the horizontal output circuit into the automatic volume
control circuit in such a way that signals can reach the automatic
volume control detector only during a short interval corresponding to
slightly more than the period of the horizontal synchronizing pulse.
This circuit is called a "keyed automatic volume control circuit" and
prevents all the interference occurring during the interval in which
picture signal is transmitted from affecting the operation of the automatic volume control. An improvement of approximately 10 to 1 was
noted over the normal low impedance automatic volume control circuit.
This type of circuit is shown in Figure 14.
Suitable limiter circuits on the signal output to the automatic
volume control circuit are also helpful in improving performance.
However, these circuits require rather careful adjustment so that the
limiting level is always somewhat higher than the amplitude of the
synchronizing signal itself. This also means that the automatic volume
control characteristic must be quite flat so that with any reasonable
value of signal input the synchronizing level will not exceed the limiter
adjustment.
4.
Instability of Synchronization
Another effect of noise and interference was to disturb seriously
the synchronization of the picture at the receiver. In many cases this
resulted in a loss of picture intelligence much greater than that caused
by the presence of the interference in the picture signal itself. During
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
/-
455
AVC RECT.
AYC AMPL
2
TO
I.
F.
AMPL. GRIDS
Fig. 12-Low impedance automatic volume control circuit.
flight tests it was frequently necessary for the operator at the receiver
to adjust the hold controls in order to keep a stationary picture on the
kinescope screen.
A fundamental difficulty with the first Block equipment lay in scanning oscillator instability. Both the line and frame frequency scanning
oscillators were cathode -coupled multivibrators whose frequency was a
function of a large number of variables including the supply voltage,
¡-AVC AMP
AVC
RECT.-
IYSL7
L
P. TO S.F.
AMPL. GRIDS
Fig. 13 -Low impedance automatic volume control circuit with
tuned circuit discrimination.
ambient temperature, and all the resistors and capacitors used in the
feedback circuit. This type of oscillator has the advantage that it produces pulse wave shapes which can be used directly for blanking and
synchronizing signals. Various other types of relaxation oscillators
were tried in an effort to find some whose frequency stability was
superior to those in use. However, none of them exhibited a frequency
stability comparable to that obtained from a sine wave oscillator.
-
-A VC AMPL.
V
AVCRECT.
25L7
Ií000
.25
POSITIVE
^ROM HOR
PUI SE
C
12
.
TO I.F.
AMPL. GRIDS
25
.
Fig. 14 -Keyed low impedance automatic volume control circuit.
www.americanradiohistory.com
456
TELEVISION, Volume IV
It was finally decided to use sine -wave type oscillators for both
line and frame frequency scanning circuits and relatively simple shaping circuits were developed to obtain the necessary blanking and
synchronizing pulses from the wave shape obtained from the sine wave
oscillator. The horizontal oscillator was of the inductance-capacitance
type with an adjustable powdered iron core for tuning the circuit to
the proper frequency. The vertical oscillator was of the resistance capacitance network type. This was used in place of a conventional
inductance -capacitance oscillator because of the large size of the inductance required for a 40 -cycle oscillator of the latter type. The
stabilization of these scanning frequencies at the transmitter resulted
in a considerable improvement in performance since the receiver
synchronizing circuits did not have to compensate for any changes of
frequency at the transmitter. It was also possible to adjust two different camera transmitter units to have exactly the same scanning oscillator frequencies so that a receiver could be tuned from one transmitter
to another without requiring readjustment of the synchronizing hold
controls. Figure I1 shows modifications made in the scanning oscillator circuits of a Block camera transmitter unit in order to permit the
lise of stable oscillators.
One -half of the 12SL7 (V8) is the horizontal frequency oscillator.
The other half is an amplifier and clipper of the pulse -wave form
appearing across the 2200 -ohm resistor in the plate circuit of the
oscillator. The plate circuit of the pulse amplifier is coupled to both
triodes of the 12SN7 (V10). One triode is the horizontal discharge
tube which develops a saw -tooth in its plate circuit for the horizontal
output tube. The other triode is the horizontal blanking amplifier
supplying blanking to the 12SN7 clipper tube (V5) in the video amplifier. One-half of the 12SL7 (V7) is the vertical scanning oscillator;
the other half is an amplifier stage which amplifies a vertical pulse
obtained by differentiating the plate voltage wave of the oscillator.
One -half of the 12SN7 (V9) is the vertical discharge tube which converts the pulse to a saw -tooth wave form. The other half of (V9) is
the vertical output tube which feeds the output transformer. Vertical
shading voltages are also obtained from the plate and cathode voltages
of this tube. Vertical blanking signal is obtained from one -half of
12SN7 (V6). In the plate circuit the horizontal and vertical blanking
signals and the video signal are mixed together to be impressed on the
connected to the horizontal and vertical output circuits.
clipper grid. The other half of (V6) is the synchronizing signal amplifier which obtains its grid voltage from differentiating circuits
However, even with this improvement severe noise and interference
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
457
conditions caused the picture on the receiver to be unstable. Radar
interference of insufficient intensity to interfere with the picture signal
was enough to cause the picture edges to be quite irregular so that
a considerable amount of picture intelligence was lost from that cause.
To improve this condition the automatic frequency control principle
was applied to the scanning oscillators in the receiver. This circuit
uses a phase detector which operates in such a way that it maintains
the pulse from the synchronizing signal and the pulse from the horizontal output circuit within a very short time interval of each other.
The output of the phase detector operates on a reactance tube which
changes the frequency of the scanning oscillator in such a way that
the proper phase relationship is maintained. This circuit has a sufficiently long time constant so that noise pulses of short duration have
practically no effect on the bias supplied to the reactance tube.
In this way the average frequency of the scanning oscillator at the
receiver is made the same as the average frequency of the scanning
oscillator at the transmitter. If, for instance, both horizontal oscillators maintain this constant frequency over a period greater than one
vertical frame then the picture edges will automatically be exactly
vertical. The longer the time constant of this circuit the better noise
immunity it will have for low. frequency interference. However, the
long time constant circuit requires an appreciably longer time to come
into synchronism if the signal is temporarily lost. Another difficulty
noticed with the long time constant circuit on the horizontal oscillator
is the lateral movement of the picture caused by rapid changes in the
path length of the signal reaching the receiver. This happens only
when the receiver is obtaining its incoming carrier signal over two
separate paths, one of them usually a direct path and the other a
reflection from the ground or some large object. The best value of the
time constant of the horizontal automatic frequency control is a matter
of compromise. Experiments indicate that an optimum value will
permit the horizontal receiver scanning oscillator to stay in synchronism with the transmitting scanning oscillator even though a synchronizing signal may be lost for one complete frame and still allow
a reasonably fast pull -in after temporary loss of synchronizing signal.
The time constant for the automatic frequency control circuit controlling vertical oscillator speed is more difficult to adjust. In order
to obtain superior noise immunity over the standard lock -in circuit the
time constant must be made long compared to one vertical frame. This
makes the recovery time after a temporary loss of signal too slow to be
satisfactory. What is necessary is a circuit which has a long time constant while the vertical oscillator is locked in but permits a short
www.americanradiohistory.com
458
TELEVISION, Volume IV
recovery time when the oscillator is out of step. One outstanding
advantage of this circuit arrangement for both horizontal and vertical
oscillators is that it practically relieves the operator at the receiving
location from the necessity of adjusting the receiver hold controls once
they have been properly adjusted.
Figure 15 shows the horizontal and vertical lock -in circuits of a
Block 1 receiver modified to include automatic frequency control of
both scanning oscillators. Referring to Figure 15, one -half of the
12SL7 (V24) is the horizontal sine -wave oscillator. The 6AC7 (V20)
is a reactance tube operating in a conventional circuit. A change in
the direct -current bias of this tube causes it to change its gain. Since
the plate current of the tube is almost 90 degrees out of phase with
the plate voltage due to the resistance capacity network in its grid
circuit, the change in gain causes a change in frequency because of the
change in reactive current through this tube which is in shunt with the
oscillator tank circuit. The bias which operates the reactance tube and
thereby changes its frequency comes from a 6H6 discriminator tube
(V19) which receives signal from a synchronizing amplifier and horizontal synchronizing separator (V14). The plate circuit of the separator has a transformer which supplies a synchronizing signal of
opposite polarity to the two discriminator diodes. A horizontal pulse
from the horizontal output is introduced into the discriminator circuit
in such a way that the same polarity pulse appears on both diodes.
When the phase relationship between the incoming synchronizing
pulses and the pulses from the horizontal deflecting circuit changes,
the direct -current output from the discriminator changes. If the phase
shift is in one direction, the direct -current output is positive; if in the
opposite direction, the direct -current output is negative. This bias
applied to the reactance tube tends to shift the oscillator frequency just
enough to keep the synchronizing pulse and the horizontal output pulse
in proper phase relationship. The time -constant of the resistance capacitance network in the output of the discriminator determines the
speed of response of the horizontal automatic frequency control system.
If this time constant is slow, the oscillator will hold its frequency over
longer periods of absence of synchronizing signal than is possible
with
a short time constant. However, if the oscillator is out of synchronism
either at the time .the receiver is turned on, or because of excessive
noise or lack of signal, the time of recovery is much slower with
the
long -time constant circuit than with the other.
It may be noted that the horizontal discriminator circuit is not
balanced with respect to ground. Test indicated that it
was highly
desirable to have the discriminator output nearly zero when
no input
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
www.americanradiohistory.com
459
460
TELEVISION, Volume IV
signal was present. Since a pulse is used from the horizontal output
instead of a saw -tooth, the discriminator output will not be zero unless
the circuit is unbalanced, with respect to ground. If this condition is
not fulfilled, the recovery or pull -in time is quite long because of the
necessity for charging the one -microfarad capacitor through the
10- megohm resistor to a definite direct -current potential.
The speed control for the horizontal hold adjustment is a rheostat
in the cathode circuit of the reactance tube (V20). A powdered iron
plug in the oscillator coil acts as an auxiliary speed control. The
cathode voltage adjustment also makes it possible to change the sensitivity of the automatic frequency control system. When the control is
near the lowest value of cathode voltage possible, the sensitivity is
greatest. With this adjustment the horizontal oscillator will lock in
about 400 cycles above and below the 14,000 cycle normal frequency.
At the highest value of cathode voltage, this range drops to about 100
cycles each way.
The automatic frequency control circuit for the vertical oscillator is
quite similar to the circuit used for the horizontal oscillator. The oscillator is one -half of a 12SL7 (V16) with the resistance-capacitance
feedback circuit mentioned before. Here the reactance tube changes
the impedance to ground of one leg of the resistance- capacitance network. By changing the bias on the grid of one -half of the 12SL7 (V18)
the plate impedance of the tube varies and therefore the impedance in
series with the .01- microfarad capacitor to ground also changes,
thereby changing the frequency. The vertical separator (V18) and
discriminator (V15) have the same circuit arrangement as the equivalent horizontal circuits. Vertical synchronizing pulses appear on the
two discriminator diodes of the 6H6 with opposite polarities and a
vertical sawtooth from the vertical output has the same polarity on
both diodes. The resultant direct-current voltage which is generated
when the diodes are unbalanced tends to shift the oscillator frequency
so that the correct phase relationship between the vertical synchronizing pulses and the vertical saw -tooth output is maintained. The time
constant used here is again a compromise between the time required
for the absence of synchronizing signal to cause the oscillator to fall
out of synchronism and the time for recovery of synchronism once the
oscillator is out of step. A choice is more difficult here than with the
horizontal circuit because of the low frequency of vertical scanning.
The time constant should normally be quite long compared to the time
required for one picture so that the vertical automatic frequency control circuit, in effect, obtains its synchronizing information from a
large number of vertical pulses. However, if this time constant is very
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
461
long, and the vertical oscillator does lose synchronism so that the
vertical blanking pulse appears near the center of the picture, a number
of seconds may be required before the automatic frequency control
circuit can reestablish normal synchronization.
The vertical hold control is a rheostat in a resistive element of the
resistance -capacitance oscillator network. With a reasonably strong
signal, the automatic frequency control circuit will keep the vertical
oscillator in step over a range of plus or minus two cycles variation
from the normal 40 -cycle frame frequency.
Variation in Light Conditions
The difference in light conditions prevailing at even moderate altitudes compared to the light conditions on the ground is usually extremely great. The normal characteristics of a scene observed from
an airplane during daylight include an extremely high light level and
very low contrast, caused by either haze or smoke in the atmosphere.
This complicates the problem of adjusting the equipment on the ground
for satisfactory performance in the air. For the early tests of Block 1
equipment, a test bench was developed for setting up the camera transmitter unit on the ground and was arranged so that a slide with very
little contrast could be projected on the iconoscope mosaic at a light
level comparable to that obtained during flight. This, in general, improved performance considerably over that obtained when preliminary
adjustments were made for a scene on the ground.
A number of experiments were made in an effort to improve the
signal output from the iconoscope under high light level and low contrast conditions. It was discovered that many of the earlier iconoscopes
would saturate at very -high light levels and would produce considerable
noise output and very little signal. In many cases improved performance was obtained by a reduction in the aperture of the lens used in
the camera. The effect of various types of filters was studied but the
results were somewhat disappointing since the insertion of a filter
which cut down the effect of haze also reduced the signal output from
the iconoscope to the point where hiss noise from the first video amplifier tube became quite noticeable. In most cases it was found that a
Wratten #25 orange filter would improve picture contrast without
seriously increasing noise.
An investigation of the saturation phenomena in the iconoscope by
the tube engineers resulted in a slight modification in the tube which
permitted better operation at high light levels. A high -light test was
also included in the iconoscope test specifications which would insure
5.
satisfactory operation.
www.americanradiohistory.com
462
TELEVISION, Volume IV
Two methods for obtaining a more nearly constant signal output
from the iconoscope were also investigated. One method was the
development of an automatic volume control circuit which would change
the gain of the video amplifier as a function of its signal output. This
was somewhat difficult because of the spurious signal delivered by the
iconoscope during the blanking intervals and the spurious signals
commonly referred to as "dark spot" and "edge flare ".
Two types of automatic volume control circuits were developed. In
one circuit only the high frequency components of the video signal were
used to operate the automatic volume control system. This operated
satisfactorily as long as there was a sufficient amount of fine detail in
the picture from which high frequency signals could be obtained. A
circuit somewhat more satisfactory was made by keying the automatic
volume control circuit in such a way that only the picture information
in the central part of the picture was used to operate the automatic
volume control. In actual flight tests these circuits showed very little,
if any, advantage over the normal circuit because the usual light conditions normally required all of the video gain available consistent with
a reasonably satisfactory signal to noise ratio. An automatic iris control was also suggested for this purpose and was actually developed,
although primarily for use with the orthicon tube.
There were two other possibilities suggested for improving the
performance of the iconoscope under flight conditions: (1) the possibility of introducing automatic shading circuits; and (2) the possibility
of controlling automatically the beam current of the iconoscope itself.
These investigations were discontinued because it was felt that the
development of the small orthicon and the image orthicon would eliminate the necessity for circuits of this type.
Low Signal Strength
With a carrier power output of approximately 15 watts from the
transmitter of Block I equipment the reliable operating range from
aircraft to ground was approximately 10 miles. Operation beyond this
range was possible but the signal would sometimes be lost because of
the change in position of the plane. An increase in this range was
considered very desirable but it was felt that considerable increase in
transmitter power would be necessary for any appreciable improvement
in performance. This would, of course, mean a much larger unit at
the transmitter with a corresponding increase in weight and power
drain. It was decided first to improve the signal -to -noise ratio of the
receiver as much as possible before attempting to provide more transmitter power. It was also considered desirable to provide the receiver
6.
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
463
operator with a tuning adjustment so that interference effects from
carrier frequencies close to the desired frequency could be minimized
as much as possible. A new type of head -end tuner for the receiver
was developed which had the following advantages:
(a) an improvement in signal-to -noise ratio of at least 2 to 1 over
previous circuits;
(b) considerable improvement in the radio- frequency selectivity
and prevention of excessive oscillator radiation by the inclusion
of a radio -frequency stage; and
(e) better performance, as evidenced by flight tests of a receiver
with the improved tuner, not only because of the improved
sensitivity but because it was possible by slight changes of
tuning to eliminate or greatly reduce interference near the
edge of the channel. (This was impossible with the earlier
fixed -tuned type receivers.)
Fig.
16
-Block
I receiver tuning unit -top cover removed.
It is estimated that the addition of this tuner increased the effective
operating range of the receiver to at least 25 miles. A photograph of
one of these tuners is shown in Figure 16 and a schematic diagram is
shown in Figure 17.
The general principle of this tuner, which has wide application, is
that of connecting appreciable fixed inductance between the low potential side of the variable capacity and ground or virtual ground. Thus,
www.americanradiohistory.com
TELEVISION, Volume IV
464
inductance is added to the circuit as the variable capacitance is increased. This action occurs most effectively when the minimum capacitance is small compared to the distributed capacitance immediately
external to the capacitor. Under this condition, the circuit current
follows a shorter path when the variable capacitor is at minimum than
when at maximum.
With ideal capacitor construction and with the largest useful auxiliary inductance, the maximum possible tuning ratio should increase
from two to three by the addition of the inductance. In practice, of
course, the maximum tuning range of 2 cannot be realized because
of limited capacity range. Hence, the addition of inductance in the
capacitor may increase the tuning range by more than the theoretical
maximum factor of 3 to 2. In the actual application, the frequency
Fig.
17 -Block I
receiver tuning
unit- schematic
diagram.
coverage was doubled by the inductance. However, the tuning ratio
has been limited to less than 2 in order to reduce the crowding at
the low-frequency end of the dial when semi-circular plates are used.
"Midline" or straight -.ine- frequency plates may also be used to improve
dial linearity. Inasmuch as the tube capacitance at one end of the
circuit is in series with the tuning capacitor at the other end, there is
little to be gained by using a tuning capacitor with a greater maximum
than two or three times the tube capacitance.
7.
Multipath Transmission
One of the troublesome problems encountered in the transmission
of television signals from one plane to another is the effect of two
out -of -phase signals of nearly the same field strength reaching the
www.americanradiohistory.com
TELEVISION FOR AIRCRAFT
465
receiver simultaneously. This frequently causes two pictures to appear
simultaneously on the face of the picture tube. The one having lower
intensity is sometimes called a ghost. Synchronization difficulty is also
experienced because of the presence of two signals. Considerable improvement can be obtained by using directional antennas and by
locating them in such a way that reflections from the ground are
minimized.
The multipath problem becomes considerably more serious if frequency modulation is present on the transmitter carrier especially if
the frequency modulation is caused by video signal. Peculiar interference patterns are caused by the combination of multipath transmission
and transmitter frequency modulation which may be so strong as to
destroy entirely the usefulness of the picture. When transmitting from
plane to ground the difficulty usually arises because a second signal
arrives at the receiving location, reflected from a large building or
other similar object in the area near the receiving location. When the
transmission path is from plane to plane the second signal at the
receiver is the one arriving by virtue of a reflection from the ground.
A similar effect can be obtained even though the receiver is at a ground
location if the antenna is located high enough so that a ground- reflected
signal will have à phase difference, compared to the direct signal, sufficient to cause trouble. How this may come about is explained briefly
in the following paragraphs.
Consider the elementary propagation system consisting of a transmitter (T), a receiver (R), the propagation medium, and the earth's
surface (E). In the simplest case, the signal is transmitted to the
receiver, R, from the transmitter, T, over just two paths, one being
the direct path TR and the other the indirect path, or path of reflection
TER. The resultant signal at R is the vector sum of the two.
Obviously, the direct signal from T will arrive at R ahead of the
indirect signal, the time difference, t, depending upon the heights of
the transmitter and receiver as well as the distance between them.
Thus, if a simple rectangular pulse of very short duration as compared
to t is sent from the transmitter, two pulses will be received, spaced t
seconds apart. Now let us assume that the transmitter is sending out
a continuous signal of frequency f, and after a period of time the
frequency is suddenly changed to f2 for a time much less than t after
which the frequency again returns to f1. No amplitude changes occur
at the transmitter during the transition from fl to f2 to fl. In t seconds
after the pulse of frequency f2 has left the transmitter two signals
will be received, fl and f2. Both will be demodulated at the receiver
producing a beat note having a frequency f1 ± f2 but only during the
www.americanradiohistory.com
466
TELEVISION, Volume IV
times when different frequencies arrive by the two paths. The frequency f i -}- f2 is too high to be accepted by the receiver, but f 1 f2
may be a very low frequency falling within the video acceptance band.
When this is the case, the beat note will appear superimposed on the
-
picture signal.
In actual television practice, large repetitive changes in modulation
occur mainly at the beginning and ending of the horizontal blanking
pulses. This means that, when the reaction from power amplifier to
oscillator is appreciable, there is a beat note of large amplitude during
the horizontal return time, or fly -back, with consequent vertical bar
patterns appearing in the picture. Often, the disturbance is further
complicated by the presence of not just one but several paths of reflection so that a corresponding number of beat notes can appear simultaneously and for varying lengths of time.
Because of this difficulty it is extremely important to keep frequency modulation of the transmitter to an absolute minimum. A
buffer stage between the transmitter oscillator and modulator is almost
an absolute necessity. It is also important that mechanical vibration
of the oscillator circuit elements be kept below a level which will cause
excessive frequency modulation. In general, it seems that more than
1 per cent frequency modulation can cause noticeable effects in the
picture.
#
r
The development of Block I television equipment and the subsequent
investigations carried out in an effort to improve the performance were
the collective work of a large group of engineers working under the
direction of R. D. Kell, in charge of television research at RCA Laboratories Divisicn, and G. L. Beers, Assistant Director of Engineering in
charge of Advanced Development at RCA Victor Division. The development and field test of this equipment would have been at best
extremely difficult without the whole- hearted interest and cooperation
of NDRC, Military, anci Naval personnel associated with this project.
This paper covers work done in whole or in part under the following
contracts: W535sc238, NOs- 86775, PDRC -29, NXs3405 -A, and OEMsr441, all with Radio Corporation of America.
www.americanradiohistory.com
TELEVISION -A REVIEW,
1946'`
BY
E. W. ENGSTROM
Vice President in Charge of Research, RCA Laboratories Division,
Princeton, N. J.
Summary-After reviewing the history of television very briefly the
four problems attending its commercialization are discussed. These problems- transmission, networking, reception and pickup-are treated in detail.
Developments made since 1939 are enumerated and the results which have
produced today's commercial television equipment are considered.
TELEVISION had its start as a public service during the year
and a half before the United States entered the war. That
called a stop to further progress. Television has marked time
during the war period, but now sufficient time has elapsed since the
close of hostilities to allow commercial television to again get under
way. Television is now growing and expanding and is taking its place
among the other regular household facilities which we have come to
regard as essential in the modern home.
Television has had a long and interesting history, and while it is
not the purpose of this paper to discuss the background or principles
of operation, since these have been adequately covered, a few items of
background material will be reviewed briefly.
Shortly after some of the first principles of electricity and particularly photo-electricity were known, people conceived methods for
achieving the results which we now call television. While it was
possible, even before the break of this century, to make crude setups
which illustrated the principle, these were not good enough to reproduce moving pictures at a distance. That had to await the coming
of the electron tube.
Television made rapid progress in the latter part of the 1920's,
but there was virtually an insurmountable obstacle ahead. The methods
chosen at that time involved mechanical devices- revolving scanning
discs, mirror wheels, etc. -all of which definitely limited how good the
picture detail might be made. There were also severe mechanical limitations in the devices themselves. What was needed to carry the art
beyond this impass was the conception of some new tools which would
give us an electronic method of operation to replace the mechanical
*
Decimal Classification: R583.17.
467
www.americanradiohistory.com
468
TELEVISION, Volume IV
processes used in television up to that time. Specifically, we needed
an electron tube in the receiver, which was capable of producing directly
on its optical surface a light image corresponding to the original
scene. Zworykin's kinescope or picture tube was the first electron tube
to fulfill this specification. Television was quick to take advantage of
this new development, but it was further handicapped by the lack of
a similar device at the transmitter. This handicap was soon overcome,
however, by Zworykin's iconoscope and the Farnsworth dissector tube.
With the development of the iconoscope, Zworykin and his associates
created the cornerstone of the structure which has become electronic
television.
The conception of this electronic method was followed by a period
of active reasearch in which the tools were made better and sharper,
so as to serve more effectively the purpose for which they were created.
There followed, then, consideration of a number of factors related to
over-all system planning, such as the practical aspects of building a
system using the new tools, the evaluation of the work that had been
done by a number of people both in this country and abroad, and the
assembly of all of these ideas and tools so as to make the best use
of the technical facilities available. It was then necessary to take the
resulting system into the field and put it to a practical test. This was
done. After the completion of these tests, the technical personnel
again reached agreement among themselves as to the fundamental
standards, such as the number of scanning lines, method of synchronizing, method of transmitting sound, etc. This standardization was
necessary so that the transmissions from all stations would be receivable on any television receiver. These operating standards were modified from time to time as additional operating experience was obtained.
All of these changes in the operating standards were in the direction
of improving the over -all television system.
At the completion of this work, the standards had to be approved
and set up by a Federal agency, the Federal Communiciations Commission. In this work the Commission had the cooperation of all of the
technical personnel of the television industry.
Television then moved into its commercial phase, which was
abruptly halted by the war. All television developments were then
turned toward objectives of a military nature.
During the short period of commercialization, study of the television systems problem indicated that a satisfactory commercial
television service would be dependent upon the successful solution
of four important problems. These four problems are represented by
the four sections of Figure 1.
www.americanradiohistory.com
TELEVISION REVIEW
469
The first of these problems concerns the television broadcasting
transmitter in a given city. In this country, broadcasting is done on
a competitive basis. It is desirable, under such a system, to have a
number of transmitters in each of the service areas. This means that
in a city or metropolitan area such as New York, there should be
several television transmitters. The radiation from each television
transmitter, however, occupies a space in the radio spectrum, several
times as wide as the whole of the region now assigned for standard
sound broadcasting, so, if the there are to be several transmitters, a
wide expanse of radio frequency channel space is required. This means
0'1
Fig. 1.- Illustrations depicting the four principal problems of television:
transmission at very high frequencies, television networking, production
of larger and brighter pictures in the home and increased camera -tube
sensitivity.
that it is necessary to use very high radio frequencies. In the years
just preceding the war, it was possible to build television transmitters
for channels 1, 2, 3, 4, or 5, but in the case of channels 6, 7, 8, 9, 10,
etc., the required transmitter tools were not available. This resolved
itself into a research problem which would be basic to a transmitter
tube and a transmitter which would permit operation with useful power
on all of the frequencies which would be assigned for television broadcasting. A power of 5 kilowatts was chosen as a goal for this work.
Figure 2 shows the tube resulting from this reasearch program. This
tube has been engineered and is available to use in television trans-
www.americanradiohistory.com
470
TELEVISION, Volume IV
mitters which may be designed for any one of the channels which
have been assigned by the FCC for commercial broadcasting. In the
development of this transmitting tube, there were two conflicting
factors: First, due to the high frequency operation, the tube had to
be small, and, second, the requirement of high transmitter power
meant that the tube must be capable of dissipating large quantities of
heat. In order to meet this double requirement, it was necessary to
develop new principles of the thermo- dynamics, as well as new principles of electronics.
These tubes made possible the construction of a transmitter at the
frequency chosen -288 megacycles- having an output power of 5 kilowatts. This frequency of 288 megacycles is higher than the highest
Fig.
2- Experimental model of a 5-kilowatt tube designed specifically
for television transmission at frequencies up to 300 megacycles.
channel assigned for commercial operation (210 - 216 mc.) and therefore gives performance information with a safety factor. This transmitter was installed together with an appropriate omnidirectional antenna atop the Empire State Building in New York City for propagation
surveys and reception tests. Figure 3 gives a view of this experimental transmitter, and Figure 4 shows the antenna as installed on
top of the Empire State Building. The 288 -megacycle antenna is the
two -layer turnstile system at the very top. From field tests one may
draw the following indications of performance in comparison with
a lower frequency channel where experience is, naturally, much more
comprehensive:
(a) In a comparison of 67 and 288 megacycles, it was found that
"shadow" effects on the higher frequency are much more
severe. This means that signals are reduced more on the
www.americanradiohistory.com
TELEVISION REVIEW
471
higher frequency by buildings, hills, and other obstructions.)
(b) In a comparison of 67 and 288 megacycles where multipath
conditions exist, it was found that multipath phenomena were
about equally serious.
(c) In a comparison of relative field strengths, it was found that
over actual urban regions the field strength on the higher
frequency was, in general, less and subject to greater variations.
The second problem pertains to televison networks -the ability to
tie together transmitters in several cities so that they may all par-
Fig.
3- 288 -megacycle television
transmitter.
ticipate in the same program. A network for television cannot make
use of ordinary telephcn? lines, nor can it make use of the same kind
of radio transmitters as are used for world -wide communication.
Instead, radio frequencies which are many times higher than even
those frequencies assigned to commercial television broadcasting must
be used. These very high frequencies have the property that they do
not bend around the earth's surface as do the longer waves, but travel
in straight paths somewhat analogous to light.
'The radio relaying problerrr is also represented in Figure 1. At the
first city, a broadcast type of service is indicated, and, in addition, a
www.americanradiohistory.com
TELEVISION, Volume IV
472
directive type of radio transmitter is shown, which directs the signal
only in the direction of the second city. Due to the curvature of the
earth, this directed signal will pass over the second city. In order to
redirect and strengthen the signal so that it may travel on, there is
placed, at an intermediate point, a radio receiver for the signal originating in the first city, and a repeating transmitter which will direct
its signal to the second city. Thus the signal is carried around the
curvature of the earth.
Fig.
4- Empire
State Building antenna structure- 288-megacycle
antenna at the top.
A single repeater is shown. The number of repeaters may be increased almost indefinitely. For example, if the two cities are separated
by as much as 300 miles, approximately 10 repeaters might be used to
cover the distance.
Figure 5 is a map showing the arrangement of an experimental
television relay circuit being installed by the American Telephone and
Telegraph Company. The two terminals of the circuit are Boston and
New York.
A second method of providing network facilities for television is to
www.americanradiohistory.com
TELEVISION REVIEW
lWwnT
wrC.
473
'
a RIT TSnELO
SLAB MILL
(LEV. 355'
aSNEEIMSaR Yt.
EL[V.1995'
3s MIL
S
TERMINaL
rinL[
WOaLlrT[[
(
SOWOOIN SRIYRRL BL06
CO.)
- (NEW (MOT. S T
ELEV.225'
SRRIN6PIELD
YRSS.
MASS.
CONN.
[Ow.
11.0
tt0 NILL
!L(Y.1266'
_
,p'1-_
1
Y[MIOENCE
a
smote NIUMM
LLEV.1020'
ÉWIV.UÌÓ'
'
JACKIE
JOLIES
SON'añÑa'
6LPSTONSUPY
.DNS
ELEV.ST5'
ulnN
o Y(W N[Y(N
[LEV 12,10
D
NEW YORK
Fig.
5
-Map
BOSTON RADIO RELAY SYSTEM
showing arrangement of experimental radio relay circuits
suitable for television.
use special coaxial cables. Figure 6 is a map indicating the extent of
the coaxial cable installations in the United States. The filled-in lines
are cables now installed; the heavy dotted lines are routes for installation contemplated so that possibly by 1950 those heavy dotted lines
will be filled in. This map shows a primary network that extends up
Fig.
6- Coaxial
cable routes in the United States, suitable for networking
television programs.
www.americanradiohistory.com
474
TELEVISION, Volume IV
and down the east coast, with a branch that makes a loop around the
central states. A major extension from the eastern and southern states
to the West, joining the large metropolitan area of the west coast
is also shown. When these circuits are completed, there will be
facilities for a nation -wide network of television. This work is considered to be one of the important steps in the development of a national
service. Sound broadcasting depends, to a large extent, on the national
program circuits for its effective operation. The same will be even
more true for television, because it is a means of distributing over a
large audience a very large expense that will have to be borne by the
broadcaster and advertiser who produce the television programs.
The third problem was the operating sensitivity of the television
camera. A camera using an iconoscope or an orthicon was reasonably
satisfactory in the studio, or other locations where the illumination
could be increased to the desired level. In moving the television camera
to points of interest on the street, in the theatre, or wherever one
might wish to go, it was frequently necessary to bring along lighting
facilities similar to those used in the making of motion pictures. The
problems involved in overcoming this serious handicap to the flexibility of television programming formed the basis for the third probtelevision camera so sensitive
lem. The goal for the project was set
that it might be taken wherever people, themselves, might go, and see
with satisfaction. This requirement can not be met by either the
present day still or motion picture cameras. As the U. S. entered the
war, it was thought that the elements of such a camera tube existed.
The development of such a tube was of great military importance, and
research continued. Because of continued research during the war
period, production of such a tube began soon after the war.
This new camera tube, shown in Figure 7, is called an image
orthicon, reflecting the names of two earlier tubes. It is considerably
smaller in size than earlier tubes. This is important, because it is
necessary to form an optical image with a photographic type lens
on the photo -sensitive surface in the tube. To cover this small area,
it is possible to use the excellent fast photographic lenses that have
been developed for regular photography.
Figure 8 is a line diagram showing the elements of this new
tube. The tube is operated in a magnetic field produced by the focusing
coil. The field of this coil is parallel to the axis of the tube. The tube
has a photosensitive surface indicated as the photo- cathode. In earlier
camera tubes, the photo- sensitive surface usually had to perform other
functions than that of merely emiting photo -electrons. In this tube
only the one fundamental function is performed by the surface, thus
-a
www.americanradiohistory.com
TELEVISION REVIEW
Fig.
7
-The
image orthicon
475
-a sensitive television
pick-up tube.
allowing optimum sensitivity. The photo -electrons are focussed by
means of the magnetic field so as to form an electric charge image
on the surface of the target screen. This target is an extremely
thin plate of glass. The electron beam produced by the gun is caused
to scan the back surface of the target screen. The returning electron
beam is deprived of electrons in proportion to the requirements for
neutralizing the charges produced on the glass target by the bombardment of the target by the photo -electrons. These returning signal el_ctrons are collected and passed through several stages of secondary
electron multiplication, which increases the magnitude of the output
current by approximately 1000 times. With this method of operation,
each part of the tube can be so constructed as to serve only that function for which it is designed, thus allowing operation under optimum
conditions.
CATHODE
(ZERO)
DECELERATING RING
(ZERO)
SECONDARY
ELECTRONS
SECONDARY
ELECTRONS
DEFLECTION
I
ELECTRON IMAGE
YORE
i
RETURN BEA
-WALL
SCANNING
COATING
BEAM
(i
180
V
)
- PHOTO- CATHODE
( -600 V )
SIGNAL OUTPUT
ELECTRODE
( +1500V)
TARGET SCREEN
ALIGNMENT COIL
Fig.
8
-Line diagram
.ZERO)
TWO -SIDED TARGET
of the image orthicon.
www.americanradiohistory.com
TELEVISION, Volume IV
476
In addition, there have been improvements in techniques to achieve
each of these functions. The combination of all of these results in a
tube which, under most operating conditions, is more than 100 times
as sensitive as the iconoscope, and under some conditions, more than
1000 times as sensitive.
When one normally thinks in engineering terms that an improvement of 2, 3, or 4 times as good as what existed before has been
achieved, it is considered that real progress has been made. But in
this case one is speaking of 2 or 3 orders of magnitude of improvement. This tube, then, is the answer to that problem of being able to
Fig.
9
-The
image orthicon camera.
television camera wherever one wishes to go, and obtain a picture under practically any light conditions.
The image orthicon is now in regular use and has extended tremendously the horizons for television programs. Figure 9 is a view of one
of the cameras now available using the image orthicon.
The fourth problem was to produce larger and brighter pictures for
the home receiver. Working in this direction, research has recently
made an outstanding advance. This is illustrated in Figure 10. As the
phosphor crystals on the face of a kinescope are excited by an electron
beam, they emit light in all directions. Some of the light goes out
through the tube face to the person viewing the picture. That is the
useful part, but some of it is emitted toward the rear and is scattered
take
a,
www.americanradiohistory.com
TELEVISION REVIEW
477
around in the tube. That part is not useful, and is, in fact, detrimental
because some of it is reflected onto the phosphor and reduces the
picture contrast.
A method has been developed by which an extremely thin layer of
aluminum is placed on the back surface of the phosphor. The film of
aluminum is so thin that the electrons readily penetrate through the
metal and into the phosphor, but it is aLso so smooth that it has a
LUMINESCENT
MATERIAL
CONVENTIONAL KINESCOPE SHOWING LOSS OF LIGHT
AND CONTRAST BECAUSE OF LIGHT ENTERING TUBE
GLASS
FACE
GLASS
LUMINESCENT
MATERIAL
FACE
IMPROVEMENTS ENSUING FROM A REFLECTING BACKING
INCREASED LIGHT OUTPUT
2 IMPROVED LARGE AREA CONTRAST
3 ELIMINATION OF ION SPOT
4 ELIMINATION OF SECONDARY EMISSION DIFFICULTIES
I
ALUMINUM
REFLECTING
BACKING
Fig. 10-Line diagram showing kinescope with and without metal backed screen.
mirror -like surface. Thus the light that would ordinarily pass back into
the tube now is reflected outwards and is added to the useful light.
Because of this, the brightness of the tube is nearly doubled, and
because there is no scattering within the tube there is also an increase
in contrast.
Also, the earlier kinescope, before the war, had an annoying defect.
www.americanradiohistory.com
TELEVISION, Volume IV
478
They showed a yellowish -brown blemish at the center after some use.
This was because- the heavier slower- moving ions would get into the
beam path, strike the phosphor, and cause deterioration. These slowmoving ions can not penetrate the metal film and an aluminized tube
is free of this difficulty.
PANE" M/ F,PO.Q
V/EW/NG
JC,QEEiY
DEP[ EC TiNG
YOKE
CORRECT/NG
4E/Y!
P.POJECT/ON i
K/NESCOPE
¢
SP/iEP/CAL
.1
Fig.
/y,'P.Fo,e
11- Cross -sectional
sketch of reflective optical system used in television
receivers to produce an enlarged image by projection.
The solution to the larger picture, as indicated in Figure 11,
involves a cathotle-ray tube, and its associated optical system. The
tube has a spherical face which is a part of the optical system, and on
it is produced an extremely bright picture-too bright to be viewed
directly. If we placed in front of this tube a conventional projection
lens, we could produce an enlarged view of the image on the screen,
www.americanradiohistory.com
TELEVISION REVIEW
479
but the optical efficiency of the system would be very low. Only a few
per cent of the light produced by the tube would reach the screen. To
overcome this difficulty, the television engineer has taken recourse to
an optical method somewhat related to that used by the astronomer.
The particular type of optical system used in television image enlargement focuses an image from the image source on a viewing screen
located at a finite distance from the area at which the image is initially
produced. This optical system consists of a large spherical mirror,
which is capable of collecting much of the light developed by the image producing tube. The mirror is shown at the bottom of the cabinet in
Figure 11. The spherical mirror directs the light rays toward the final
viewing screen, but between the mirror and the screen it is necessary
to interpose a lens in order to produce a good optical image. This lens
is called a correcting lens and is relatively weak, optically. Its surface
is of the aspherical type, which makes it extremely difficult to grind
from glass and much too expensive fcr the home type of receiver if
produced by normal methods. In order to make such an optical system
practical, it has been necessary to learn how to make a negative of the
lens surface in metal, and how to use this as a mold to produce the lens
from plastic material. With such an optical system, a very practical
means now exists for producing large television pictures.
Figure 12 is a view of an experimental type of projection receiver,
using the above- described optical system, which may be viewed by large
groups. Figure 13 is a view of one of the table model television receivers now available on the market. It has a kinescope, or cathode -ray
tube, 10 inches in diameter and produces a picture large enough to be
viewed by small groups.
This discussion has indicated how definite solutions have been found
to the four major problems that existed as television service was
started in 1939. Now cameras are available which remove the limitations of programming ; transmitters can be built for all the channels
the Federal Communications Commission has allocated; ahead lies the
facilities which will give television networks; and larger, brighter pictures can now be seen in the home. With reference to monochrome
television, the research worker has passed the problem onward; little
more is needed from him. The task now lies with others -to do the
work of producing in the factories, selling in the stores, and programming in the stations. All of this is now under way; the research
men have, for some time, been concentrating on the problems of color
television; believing that the addition of color in a system that is both
technologically sound and compatible with present television is the next
and orderly step in the growth of television service.
www.americanradiohistory.com
TELEVISION, Volume IV
480
Fig.
12
-A
developmental model of a projection -type television receiver
providing a picture 16 by 22 inches in size.
Television, from a technical point of view, is a reality, and for
those who have been so active during the past ten, twenty or more
years in the research and development of television, a goal has been
Fig.
13 -Table
model television receiver with a 10 -inch screen.
www.americanradiohistory.com
TELEVISION REVIEW
481
attained. Now others must appear on the scene to assist in developing
a public service. It is a service of great significance, and the horizon
should be one of great expanse. There is a steep hill to climb, to get
away from the shadows that limit vision and to seek to comprehend
the complicated panorama that lies in view.
Television is going to affect men's lives in many ways. It will permit
men to transport their eyes to distant points so that they can see things
while they are taking place. It will give them new means of entertainment and new means for- improving their education. It can assist merchandising practices and correlate manufacturing operations.
Everyone can participate in this effort, and by their acts of omission or commission will be determined the excellence and character of
the service entrusted to television.
www.americanradiohistory.com
TELEVISION BROADCASTING -1946*
BY
O. B. HANSON
Vice President and Chief Engineer. National Broadcasting Co., Inc.,
New York, N. Y.
Summary -A general review of television broadcasting activities in
1946 is given together with discussions of factors which influenced this
activity. The paper concludes with a consideration of the future of television broadcasting.
THE year 1946 could well go down in history as the year in which
television broadcasting sped from around that mythical corner
where rumor had placed it for the previous ten years. Monochrome television was ready to go, both technically and commercially,
in 1941, and received the blessing of the Federal Communications
Commission in June of that year. It started to emerge from around
the corner when it was stopped in its tracks on December 7, 1941, when
the United States became a combatant in World War II. For a four
year period, hostilities arrested its commercial development.
Television itself went to war and played an important role in our
military developments. It was the television techniques that gave
birth to the now fabulous radar, many wonders of which materially
aided in winning the war and saving the lives of countless thousands
of American citizens.
During the war, television broadcasting was not entirely somnolent,
as WNBT in New York City carried on with limited program schedule
throughout the war and was used by the New York Police Department
and the Office of Civilian Defense to train many thousands of air raid
wardens.
With the cessation of hostilities, television again sprang to life,
and while new television equipment and receivers were as yet unavailable, existing stations reestablished or increased their program schedules to the limit of available trained man power. During 1946, the
Federal Communications Commission, after reviewing the technical
standards for television, again approved those standards and urged
industry to go full speed ahead. The Federal Communications Commission also made new allocations for television, providing thirteen
channels in the ultra -high frequency portion of the spectrum, and these
thirteen channels make it possible to assign frequencies to approxi*
Decimal Classification R583.17.
:
482
www.americanradiohistory.com
BROADCASTING
483
mately four hundred television stations throughout the United States.
Thus television broadcasting is assured of adequate channel space. to
provide spectacular growth within the next several years. During 1946,
the Federal Communications Commission also granted construction
permits for forty -seven commercial television stations (and one educational television station) scattered through major cities from the East
to the West Coast; nineteen more applications are in process with many
others anticipated.
During the early part of the year, several existing television stations made the necessary changes in their equipment to enable them to
broadcast on the new channels assigned by the FCC. The NBC station
in New York, WNBT, installed a new transmitter and erected on the
pinnacle of the Empire State Building a new high -gain television
transmitting antenna. This equipment has been in operation with an
average of twenty program hours per week since May 9, 1946. Many
outstanding programs have been transmitted among which was the
spectacular Louis -Conn heavyweight championship fight. This particular program was picked up by field equipment, using five cameras at the
Yankee Stadium, and transmitted by microwave to Radio City, whence
it was distributed by coaxial cable to television stations in Washington,
Philadelphia and New York, and by a mountain -top radio relay to the
General Electric station in Schenectady. A large audience of government officials seated before projection receivers in the Statler Hotel
in Washington also viewed the event. It is estimated that over two
hundred thousand persons witnessed this event by television.
February 12, 1946, Lincoln's Birthday Anniversary, marked the
opening of the American Telephone and Telegraph Company's coaxial
cable for television program transmission from Washington to New
York. Television cameras viewed the ceremonies taking place at
the Lincoln Memorial and at other spots in the Nation's Capitol.
Television audiences in Washington, Philadelphia, New York and
Schenectady viewed those ceremonies, while they took place, through
the magic of television networking. Thus the practicality of intercity television networking was again demonstrated, emphasizing the
importance of intercity television links, and holding forth the promise
of nation -wide service in the relatively near future.
The American Telephone 2nd T- legraph Company is now in the
process of constructing a transcontinental coaxial cable which will
link the East Coast to the West Coast. This cable will probably be
ready for television transmission in late 1948, and to the many progressive broadcasters throughout the country, this promise of television network service is a great stimulus. The Bell System is also
www.americanradiohistory.com
484
TELEVISION, Volume IV
in the process of constructing a radio -relay system suitable for carrying
television programs between New York and Boston, which will be
available for such purpose late in the summer of 1947. The local telephone companies have also made available to television a method of
transmitting television over telephone wires for short distances to
facilitate pickups in urban areas. This service is also augmented by
the use of radio relays. It was through the use of such a radio relay
that the television audience was able to watch the Army football
games as they were played at the Military Academy at West Point.
As part of regular weekly broadcasts, boxing is televised from
Madison Square Garden and St. Nicholas Arena in New York City, and
through the use of light-weight portable microwave relay equipment,
practically any point within twenty miles of Radio City becomes available for television camera coverage; through the use of a double relay,
even greater distances can be covered, giving television cameras a
greater program flexibility than that which was available before the
war. The New York, Schenectady, Philadelphia and Washington television audiences have witnessed many outstanding baseball and football
games televised during 1946.
The television frontier has been immensely extended due to the
availability in 1946 of new portable pickup equipment, the cameras of
which incorporated the wartime -developed highly sensitive image
orthicon camera tube which is able to produce pictures of excellent
quality at extremely low light levels, exceeding in sensitivity that
attributable to the readily available motion picture film. This one
great development enables television to present any event, day or
night, requiring no more light than that which one encounters in any
night spot where the public gathers.
The President of the United States, Harry S. Truman, was televised as he addressed the commencement exercises at Fordham University on May 11, 1946, which episode gives a glimpse into the future
of the part which television broadcasting will play in the political
affairs of the future. In October, 1946, the Television Broadcasters
Association held its Second Conference and Exhibition which attracted
over a thousand radio executives and engineers from all parts of the
country. All of the main sessions of this conference were televised over
a network covering New York, Schenectady, Philadelphia and Washington. At this exhibition, the first of the post-war home television
sets of several different manufactures were publicly demonstrated.
These sets included table models, console models incorporating standard band broadcasting, FM, and automatic phonographs, and large
screen projection receivers. Great interest was displayed in these sets
www.americanradiohistory.com
BROADCASTING
485
which were just then starting to roll off the assembly lines. It was
estimated that some fifteen thousand receivers would be available for
public purchase before the end of 1946, and it is further estimated that
between two hundred thousand and three hundred thousand sets will
have reached the public by the end of 1947; within three years, approximately one million sets should be in the hands of the public.
Television, destined to play a world -wide role in the future, brought
to the present four -city television audience many of the sessions of the
United Nations as they gathered at their headquarters in Flushing,
Long Island. Famous personalities representing member nations of
the United Nations appeared daily on the home screens of the tele-
vision audience.
Again forecasting the important role which television will play in
the destinies of national politics, television cameras televised the opening session of the 80th Congress as it convened at the Capitol on
January 2, 1947 and January 6, 1947. As President Truman addressed
the combined houses of Congress, he was seen by the television audience in the four above-mentioned cities through the eyes of the
television cameras.
Nineteen hundred and forty -six has, indeed, marked an important
milestone in the growth of television. These important events will
fade, however, into insignificance compared to the anticipated growth
in 1947. Assembly lines producing home receivers will be in full
swing, and factories will be turning out, in quantity, television transmitters, cameras, studio and field equipment, supplying those broadcasters fortunate enough to have received from the Federal Communications Commission their construction permits for television stations.
As always, the research and engineering laboratories of this corporation have pressed their reasearch for new and ever -improved
television devices, and there has emerged in 1946 a promise for the
future of color television. Several methods of attacking this problem
have been demonstrated, but as yet, there remains much work to be
done before color television can provide a public service as practical
and reliable as the monochrome service which the public now enjoys.
The wonders of radio never cease, and television is the greatest
marvel to emerge from electronic laboratories. It holds an influence
for the good of mankind that no man can entirely comprehend.
www.americanradiohistory.com
TELEVISION TODAY AND ITS PROBLEMS -1946*
BY
ALFRED N. GOLDSMITH
Consulting Engineer, Radio Corporation of America,
New York, N. Y.
Summary -This paper is a general review of present day television.
The outstanding characteristics of television as a public service are discussed. Specific topics covered include studio equipment and operations
remote pickups, special studio techniques, programming, network facilities,
receivers, and other television development factors.
OMMERCIAL television broadcasting is now an active and
growing art in America. Despite its comparative youth, it has
attained much of the picture clarity, program flexibility, and
entertainment capabilities of motion pictures. It surpasses even that
well -developed art in its spontaneity of presentation, its occasional
rendering of the unexpected, its instantaneity and immediacy of production and reproduction, and its economical production technique. In
due course, it may be expected to take its place side by side with its
older sister art.
The following discussion of present -day television and certain of its
problems is necessarily a general review. To have listed all details and
to have given due credit to the capable engineers and program planners
who have contributed to present -day television would have far exceeded
the permissible limits of this brief description.
C
STUDIO EQUIPMENT AND OPERATIONS
As the result of several years of careful planning and testing, studio
equipment may be regarded as semi -standardized. It is able to meet all
normal operating requirements to a reasonable extent. Most of its
future development will be more in the direction of detailed improvement, in all likelihood, than of radical modification (unless, at some
future date, basically different methods of television operation should
be discovered and found superior to those now available
somewhat
unlikely contingency) .
Studio and remote -pickup cameras are now highly sensitive and
readily handled. The image orthicon has proven to be a picture tube
of unparalleled capabilities. Its sensitivity is a hundredfold greater
-a
*
Decimal Classification R583.17.
:
486
www.americanradiohistory.com
TELEVISION PROBLEMS
487
than the original iconoscope and its resolution or detail- recording
power has been brought to a satisfactory level. Television cameras may
in the future be made even more mobile and free from limitations on,
or rapid changes of, their position. The view finders, being electronic,
are non -parallactic and truly indicative of the actually transmitted
pictures. A full complement of lenses is available which can be used
at distances of from tens to hundreds of feet from the subject, and
capable of covering anything from a clear wide -angle view to a nearly
flawless close-up. Practically any desired "effect ", wipe, fade, or the
like can now be optically or electronically secured.
At the present time television pictures lack depth of field in some
cases. The problem of increasing this depth can be attacked by the
use of cooler light sources and more of them, more sensitive camera
tubes, or specialized optico -electrical systems of studio illumination inherently capable of delivering an increased range of field.
Studio lighting today is usually adequate in amount, but there is
room for future improvement in its dramatic values and in the more
widespread and skillful use of modelling lighting of types showing
actors and sets to best advantage.
When color television is ultimately commercially achieved, additional lighting problems will arise. The color quality of illumination
must be more carefully controlled. Adequate lighting for color pickup
and reproduction ranges from eight to fifteen times the amount necessary for black-and -white operation, the exact figure depending in part
upon design skill and in part upon the desired fidelity of color reproduction.
Studio control equipment during the last years has become more
dependable, accurate, and convenient in manipulation. The task of the
director in "editing" the television production in the control room, and
as it progresses, has become more effective and less trying. Further,
the control of the activities of studio personnel from the director's
control desk is now more rapid and complete than it was in the past.
There remains for the future the development of methods for simplifying the tasks of the camera men, the microphone -boom operators, and
the actors, through coordinated assistance from the control desk.
REMOTE PICKUPS
A new era in the scope and quality of remote pickups was ushered
in by the modern highly sensitive television camera. Sports events
under normal lighting are now most acceptable elements of the television program. In the days to come, the portability and ease of manipulation of remote equipment will doubtless be even further enhanced.
www.americanradiohistory.com
488
TELEVISION, Volume IV
It is necessary that connections between the cameras, the remote -pickup
control equipment, and the receiver which controls the main transmitter shall be compact, light, easy to handle, and yet capable (as they are
today) of transmitting the sight and sound signals without noticeable
distortion.
SPECIAL STUDIO TECHNIQUES
The optical or electrical insertion of backgrounds has been under
study. It appears likely to prove a useful factor in the future in readily
enabling more elaborate and yet economic productions. If a length of
film or a slide can adequately replace a physical set for background
purposes, many studio problems and economic limitations will be
worked out in the control room rather than on the studio floor. Picture
presentations are increasingly elegant and smooth. Fades, dissolves,
and wipes can be optically or electrically contrived. Numerous novel
and attractive "effects" have been found feasible -and doubtless many
more of these will be evolved in the future.
At present, live orchestras are unavailable because of restrictions
imposed by their labor organizations. These restrictions have, however,
benefited television in one respect. They have made necessary the
development of convenient and economical ways of using orchestral
and vocal phonograph records in television, employing an evolved
dubbing technique. The home audience may, for example, hear a singer
in actuality, accompanied by an orchestral recording of the selection
which he is singing. Or the entire sound reproduction in the home may
originate on a vocal and instrumental record which is transmitted on
the audio channel and which also "cues" the actor in the studio so that,
so far as the audience is concerned, he seems to be singing the actual
selection in exact time. These methods may be further elaborated in
time, even if live orchestras become available, since they have certain
inherent technical, operating, and economic advantages.
The use of film transmissions in television is considerable at
present. The dependability and quality of such transmissions are now
in the main satisfactory. Certain details of operation may be added
or improved in the future. For example, it might prove convenient to
run certain types of film at any desired speed and yet transmit them
at the standard number of frames per second. Experimentation on the
remote control of film projectors from the television studio control
room will doubtless be studied in the future.
Experimenters have indicated the possibility of recording television
programs on 16- millimeter or 35- millimeter film, with the accompanying sound. Such film serves as a permanent record of the program, and
www.americanradiohistory.com
TELEVISION PROBLEMS
489
provides for its later repetition or for its syndication on a transcription
basis. Another application of the recording of a television program
on film is in the field of theater presentations. If a television program
is received in a theater, photographed on film, developed at high speed,
and then projected on the theater screen, it offers one method of large screen television for theater purposes. The use of television in theaters
is a field under development, and one having attractive possibilities.
PROGRAMMING
The program structure of today is relatively simple and is based
largely on most economical production for a limited number of operating hours per week. The future still presents numerous program
problems. It will become necessary to present a far wider variety of
program types, and to learn how to produce thoroughly attractive
program material on a large -scale and economical basis. Television on
such an expanded scale will require locating and training large groups
of skilled personnel capable of meeting television-program needs.
Closely tied into the program problems of the future are certain
economic questions which must be studied. It remains to determine
the most desirable number of stations in each type and size of community, the sort of programs which should be transmitted, and the measure
of sponsor support which will be necessary to provide adequate program
service in various sorts of communities. To determine these factors
may involve the detailed and continued study of audience reactions to
an extent not yet found necessary in radio broadcasting.
NETWORK FACILITIES
As of the end of 1946, over three thousand miles of coaxial cable
and many hundreds of miles of radio -relay systems have been established. By 1950, it is stated that 12,000 miles of coaxial cable capable
of carrying television programs will form a national network. No rate
structure has as yet been established for the use of such facilities. It
will be a complex task, though a promising one, to provide nation -wide
high -fidelity black- and -white television syndication on an economic
basis. The prospects are hopeful in view of the many conceivable and
profitable by- product uses of the wide -channel circuits required for
television syndication.
To secure maximum flexibility and greatest usefulness of networks
for television program distribution, it is desirable that the engineers
develop standards and simplifications in the use of cables, wave guides,
and radio -relay systems. Interchangeable operation of standard equip-
www.americanradiohistory.com
490
TELEVISION, Volume IV
ment will thus be promoted. It may be added that, when color television
becomes commercial, network operations will present technical and
economic problems of increased difficulty.
One television problem, shared with other higher- frequency services, and as yet almost untouched, is that of providing large -area rural
television coverage in non -mountainous regions, and controlling the
corresponding transmissions from national networks. In view of all
that television can offer the farmer, it is greatly to be hoped that a
solution will be found for this problem. To some extent such a solution
will be "tied in" with detailed studies of wave propagation from stations of various carrier frequencies, powers, and heights above ground.
RECEIVERS
Present -day television receivers operate on signals transmitted on
thirteen black-and -white channels, each six megacycles wide, and
located in groups between 44 and 216 megacycles. Some of the receivers show pictures by direct viewing and others by projection of an
intensely brilliant image through a high -efficiency optical system.
Other optical methods of enlarging pictures efficiently have been considered. The present -day picture is a 525 -line (nominal) image with
slightly greater resolution vertically than horizontally. Its brightness
is of the order of 20 -60 foot -lamberts, which is a marked improvement
on earlier pictures and adequate for home reception under normal
conditions. The gradation range of the picture has been greatly in-
creased.
Television receivers of today are compact, attractive in appearance,
and easily handled by most persons. They may fairly be said to be far
in advance, in their stage of technical and operating development, of
the radio receivers of the mid- 1920's during the period of the vast
expansion of radio broadcasting.
Home antennas are mostly of the dipole type, using one or more of
such elements with or without passive reflectors. The problem of providing television service in multiple-apartment dwellings has been
studied in principle. It involves centralized antenna, amplifier, and
distribution systems leading to the various apartments. Technically,
such systems seem practicable. Their economics -as well as the landlord reaction to their installation- remain to be explored.
Television still offers a number of questions to the analyst. It is not
known what number of available channels would meet all reasonable
requirements in each type and size of community. The relative status
and feasibility of operation on 44 -216 megacycles versus operation on
480 -920 megacycles (or even operation on frequencies well above
3000
www.americanradiohistory.com
TELEVISION PROBLEMS
491
or even 5000 megacycles) will require more technical work and field
experimentation than has yet been devoted to this important question.
Then, too, as picture size increases in the home, it may be found that
525 -line pictures are still entirely adequate -or, Alternatively, it may
prove desirable to go to pictures of greater resolution. This last question, however, is far from urgent since it is generally agreed that the
525 -line picture is capable of giving continuing entertainment value
to the television audience.
TELEVISION- DEVELOPMENT FACTORS
The progress of television will involve some further developments
or expansions whereby the television audience will be satisfied at all
times that television progress continues.
One element of television which cannot be slighted is high- quality
servicing. The installation and maintenance of television receivers can
be handled only by highly qualified service men. Such men are being
supplied by the television- receiver manufacturers or by organizations
sponsored by such manufacturers. This is a wise and helpful procdure.
Satisfaction in television reception depends upon the absence of
visible interference with the picture and audible interference with the
sound portion of the program. It is well known that certain medical
and industrial equipment, such as diathermy apparatus, industrialheating devices, and some automobile ignition systems may seriously
affect picture quality and cause noticeable interference with the sound
portion of the program. A combination of regulatory measures and
public education may lead to the reduction or elimination of such
disturbances.
Those who watched the early development of radio broadcasting
will recall the serious interference which resulted from radiating
receivers. This possibility has already been found to exist in television.
The elimination of such interference will require agreement on engineering standards of good practice for "non- radiating" receivers, and
possibly the promulgation of corresponding regulations as to permissible field strength at specific d'_stsnces from suitably designed television receivers.
The cooperation of all major factors in the television industry is
important if the art is to grow normally. The Radio Manufacturers
Association provides such cooperation between manufacturing groups,
and the Radio Technical Planning Board, through its various Panels,
offers an even wider forum to all interested and significant organizations and individuals.
www.americanradiohistory.com
492
TELEVISION, Volume IV
Perhaps the least studied portion of the basis of television today
is its psycho -physical background. The effect of various types of image
presentation upon the observer merits, and will doubtless receive, careful investigation and-analysis in the future. The improvements which
would result in the entertainment value of television would probably
repay extensive investigations along psycho-physical lines.
As of the end of 1946, color television is under vigorous discussion.
Two major types of color television: simultaneous and sequential, have
been demonstrated. The considerations favoring each of these have
been forcibly urged by their respective proponents. Color television is
a more complicated matter than most enthusiasts in that field appreciate at this time. Our knowledge of color reproduction methods, and
of their desirability in television, is certainly in its early stages in
some respects. Also under active analysis are questions of the economics of color television involving transmitter cost, studio and program operation costs, and receiver costs. It is evident, however, that
a healthy development trend exists in color television and that, in the
years to come, this art may also find its sphere of application in home
entertainment.
Some workers have, from time to time, discussed stereoscopic
television, in black -and -white or color. The subject is of limited interest
at present but may attract increasing attention in time.
It may also be mentioned that important industrial television applications, as well as uses of television in military and other fields, will
increase in number and grow in value.
As of the close of 1946, it has been fundamentally established that
television broadcasting is an outstanding contribution to mass communication and that it rests on basically correct technical, esthetic,
and long-term commercial grounds. Among its remaining problems are
further simplification of equipment and operations, achievement of
large -scale and high- quality production and reproduction, expansion
of the scope and coverage of television programs, development of
coordination between television and other methods of entertainment
and instruction, the maintenance of a constructive and helpful attitude
on the part of the governmental regulatory authorities, the growth of
intra- industry cooperation, and the gradual evolution of television
toward even greater technical and artistic capabilities and achievements. Considering its present capabilities and accomplishments, television presents most attractive and hopeful vistas of future achievements and universal acceptance.
www.americanradiohistory.com
SLJ1VIlVIARIES
The following papers are presented in summary form only. The journal
in which the full paper appears is indicated in each case.
MEASUREMENT OF THE SLOPE AND DURATION
OF TELEVISION SYNCHRONIZING IMPULSES t
BY
R. A. MONFORT# AND F. J. SOMERS$
Summary
receivers in the field requires that
operation
of
television
Satisfactory
the waveform of the transmitted synchronizing signals be held to narrow
tolerances. It is therefore essential that suitable measuring equipment and
techniques be available at the transmitter so that synchronizing waveshapes
can be accurately and rapidly checked. This paper describes several measurement methods which have been found to be satisfactory under practical
operating conditions.
* Decimal Classification: R583.13 X R200.
f RCA REVIEW, January, 1942.
# Formerly with the Engineering Department, National Broadcasting,
Company, Inc., New York, N. Y.
Engineering Department, National Broadcasting Company, Inc.,
New York, N. Y.
THE RELATIVE SENSITIVITIES OF TELEVISION
PICKUP TUBES, PHOTOGRAPHIC FILM, AND
THE HUMAN EYE *t
BY
A. ROSES
Summary
The threshold scene brightness which a picture-reproducing device can
record, a measure of its "operating sensitivity", depends not only upon the
lens speed and the exposure time, but also upon the amount of detail in the
recorded image. A general expression for the "operating sensitivity" of a
picture -reproducing device is obtained which includes these factors together
with the threshold number of quanta per picture element. This parameter
498
www.americanradiohistory.com
TELEVISION, Volume IV
494
characterizes the "true sensitivity" of the given device. The "true" and
"operating" sensitivities of four types of television pickup tubes, photographic film, the human eye, and an ideal picture -reproducing device are
obtained. Eye and film have of the order of one one -hundredth the "true
sensitivity" of an ideal picture- reproducing device. Some recent television
pickup tubes have of the order of one one- hundred-thousandth the "true
sensitivity" of an ideal device.
To compare "operating sensitivities ", the same exposure time and
equivalent lens systems are taken for the three devices. A television pickup
tube which has a photoelectric response of 10 microamperes per lumen and
makes full use of the storage principle can record scenes with no more
illumination than that required by some of the "faster" photographic films.
The relatively low "operating sensitivity" of film results from the large
amount of intrinsic picture detail (a picture element is taken to be a single
grain). The human eye has a range of "operating sensitivities" extending
from that of film to a value several thousand times higher, This range depends upon the ability of the eye to coarsen the detail of its perceived image
as the scene brightness is lowered.
*
t
$
Decimal Classification: R583.1.
PrOC. I. R. E., June, 1942.
Research Department, RCA Laboratories Division, Princeton, N. J.
PORTABLE HIGH -FREQUENCY SQUARE -WAVE
OSCILLOGRAPH FOR TELEVISION -t
A
Bx
R. D. KELL,# A. V. BEDFORD# AND H. N. KOZANOW5KI$
Summary
A portable high -frequency oscillograph for television is described by
which a square -wave (100 -kilocycle) response may be viewed as a dotted
wave and readily recorded as a series of readings. The dots are spaced at
1/30- (or 1/20-) microsecond intervals. No electrical connection is
required between the oscillograph and the square -wave generator other than
that established through the apparatus under test since the synchronous
sweep and timing dots are derived from the square -wave response of the
apparatus. Circuit diagrams of the square -wave generator and square -wave
oscillograph are given.
*
Decimal Classification: R371.5 X R583.
f Proc. I. R. E., October,
# Research Department,
$
N. J.
1942.
RCA Laboratories Division, Princeton, N. J.
Engineering Products Department, RCA Victor Division, Camden,
www.americanradiohistory.com
SUMMARIES
:195
CATHODE -RAY CONTROL OF TELEVISION
LIGHT VALVES "t
BY
J. S. DONAL,
JR.:!:
Summary
When a light valve is employed for the reproduction of television pictures, it is desirable to make use of a cathode -ray beam to control the light
valve in order to preserve the all -electronic character of the television system. A number of procedures of cathode -ray control are described, the
majority of which are applicable particularly to the control of the suspension light valve.
The general method employed is shown to be the production of an
electric field through the light valve by bombarding one side of the valve
with electrons of very high velocity, causing the valve areas to be charged
in a negative direction toward the limiting potential of the bombarded surface. Removal of the electric field is then accomplished by charging these
areas back toward their original potential by the use of electrons of substantially reduced velocity.
The most elementary procedure described is one in which a single beam
of electrons of constant velocity is employed, discharge being accomplished
by secondary electrons generated by the action of the beam of primary
electrons.
The effects of polarization of the light valve, resulting from the comparatively low resistivity of the suspension, are described and explained. It
is shown that a suspension of such low resistivity as to be uncontrollable by
the other procedures may be made operative when the valve is used in combination with a spatially modulated electron spray and when, in addition,
the potential of one wall of the valve is increased and decreased at a
moderate frequency.
Of the procedures described, the most effective from the practical
standpoint is shown to be one in which the light-valve field is developed by
a scanning beam, and in which the field is later removed by rescanning with
the same beam at a reduced electron velocity. A photograph is shown of a
picture reproduced by the light valve when controlled by this method.
Decimal Classification R583.15.
Proc. I. R. E., May, 1943.
I Research Department, RCA Laboratories Division, Princeton, N. J.
*
:
A
REFLECTIVE OPTICAL SYSTEM FOR
TELEVISION *t
BY
E. W.
WILBY:i:
Summary
The most logical reason for the use of a reflective optical system for
projection television is that it is the only known simple means capable of
www.americanradiohistory.com
496
TELEVISION, Volume IV
focusing a large field with high efficiency. The maximum efficiency obtainable from conventional projection lenses lies between 6 and 12 percent,
while it is quite feasible to obtain between 20 and 40 percent from reflective
optical systems. The term efficiency, as used here, refers to the fraction of
the total light emitted by a central point on the tube face that is focused
on the image point.
It has long been known that there are very efficient means for focusing
very small sources, for example, parabolic reflectors for infinite throw and
ellipsoidal mirrors for finite throw. A considerable number of good reflective systems capable of handling large fields have been proposed and many
of them have been used. All of these however contain one or more aspherical
elements, and were designed for infinite focus for astronomical use. The
simplest of these was proposed by B. Schmidt and consists of a spherical
mirror and a very weak aspherical correcting lens.
Ill.
* Decimal Classification: R583.15 X R138.3.
f RCA Licensee Bulletin LB -630, November, 1944.
$ Industry Service Laboratory, RCA Laboratories Division, Chicago,
PROJECTION TELEVISION *f
BY
D. W. EPSTEIN# AND I. G. MALOFF$
Summary
Projection television, which is simply the projection onto a viewing
screen of the picture originating on a cathode -ray tube seems, at present
to be the most practical means of producing large television pictures.
The 2 basic problems of projection television are: (1) the problem of
providing a cathode -ray tube capable of producing very bright pictures
with the necessary resolution and (2) the problem of providing the most
efficient optical system so as to utilize the largest possible percentage of
the light generated. These problems were very vigorously attacked over a
period of years and the progress made toward their solution has been very
satisfactory.
Problem (1) has been solved largely by the development of cathode -ray
tubes capable of operating at high voltages. Problem (2) has been solved
by the development of a reflective optical system about 6 to 7 times more
efficient than a good f/2 refractive lens. The reflective optical system consists of a spherical front face mirror and an aspherical correcting lens.
A handicap of this optical system, for use in a home projection receiver,
was the high cost of the aspherical lens. This has been overcome by the
development of machines for making aspherical molds and by the development of a process for molding aspherical lenses from plastics. RCA reflective optical systems are designed for projection at a fixed throw and require
cathode -ray tubes with face curvatures fixed in relation to the curvature'of
the mirrors in the system. A number of such systems, suitable for project-
www.americanradiohistory.com
SUMMARIES
497
ing television pictures with diagonals ranging from 25 in. to 25 ft., have
been developed.
Decimal Classification: R583.
Jour. Soc. Mot. Pic. Eng., June, 1945.
# Research Department, RCA Laboratories Division, Princeton, N. J.
$ Home Instrument Department, RCA Victor Division, Camden, N. J.
*
j
BAND -PASS BRIDGED -T NETWORK FOR
TELEVISION INTERMEDIATE -FREQUENCY
AMPLIFIERS *t
BY
G. C. SZIKLAI$ AND A. C. SCHROEDER$
Summary
Bridged -T networks offer great economy in television intermediate frequency amplifiers for sharp attenuation of the associated and adjacent
sound channels.
A simple design method was obtained by the use of the equivalent
lattice. By the same method, general formulas were obtained for the phase,
attenuation, and delay characteristics. Two designs are given to illustrate
the convenience of the method.
*
$
Decimal Classification: R583.5.
Proc. I. R. E., October, 1945.
Research Department, RCA Laboratories Division, Princeton, N. J.
INPUT IMPEDANCE OF SEVERAL RECEIVING -TYPE
PENTODES AT FM AND TELEVISION
FREQUENCIES*"
BY
F.
MURAL:
Summary
The input impedance of vacuum tubes is an important circuit design
consideration at high frequencies. This report includes information on the
input impedance of a number of currently available r -f pentodes.
Measurements are given on the variation of input resistance with frequency, and on the variation of input resistance and capacitance with plate
www.americanradiohistory.com
TELEVISION, Volume IV
498
current. Measurements are also given for the compensation of input resistance and capacitance variation with plate current by means of unby -passed
cathode resistance.
The frequency range of measurement was chosen roughly to cover the
frequency modulation and television transmission assignments as well as
the recommended intermediate frequencies of receivers for these services.
*
Decimal Classification: R583.6.
Licensee Bulletin LB -661, March, 1946.
Industry Service Laboratory, RCA Laboratories, New York, N. Y.
j RCA
$
TELEVISION HIGH -VOLTAGE R -F SUPPLIES *t
BY
R. S. MAUTNER# AND O. H. SCHADET
Summary
Because of the contemplated large scale production of television receivers, considerable interest is being shown in various methods for obtaining
a source of high voltage simply and economically.
The r-f oscillator type high voltage supply described in this bulletin
offers advantages in economy and space requirements, and provides satisfactory regulation over the normal operating range of cathode -ray tube
beam currents. Furthermore, its safety factor against dangerous shocks is
considerably greater than that obtained with conventional sixty -cycle supplies. The principles of operation and design of these supplies have been
previously described (Schade, O. H., "Radio- Frequency Operated High Voltage Supplies for Cathode -Ray Tubes ", Proc. I.R.E., pp. 158-163, April
1943). These are here reviewed and considered in greater detail. Constructional features of four typical units are shown and their performance is
illustrated by curves indicating the magnitudes of current and voltage
obtained under typical operating conditions., Sample calculations for the
specific cases of a 75 -watt 90- kilovolt supply and a 10 -watt 30-kilovolt supply
are included to illustrate the progressive steps in designing and calculating
the circuit elements and operating conditions for a specified performance.
During the time that these supplies have been in operation they have given
stable trouble -free performance and have required a minimum of attention.
However, careful shielding and filtering are required to minimize undesired
radiation. Corona problems especially require careful consideration to prevent ionization during periods of high humidity.
*
Decimal Classification: R583.5 X R366.
- RCA Licensee Bulletin LB -675. August, 1946.
# Industry Service Laboratory, RCA Laboratories Division, New York,
N. Y.
j
Tube Department, RCA Victor Division, Harrison, N. J.
www.americanradiohistory.com
APPENDIX
TELEVISION
A Bibliography of Technical Papers
by RCA Authors
-
1929
1946
This listing includes some 275 technical papers on
TELEVISION and closely related subjects, selected from those written by RCA Authors and
published during the period 1929 -1946.
Papers are listed chronologically except in cases
of multiple publication. Papers which have appeared in more than one journal are listed once,
with additional publication data appended.
Abbreviations used in listing the various journals
are given on the following page.
499
www.americanradiohistory.com
500
TELEVISION, Volume IV
ABBREVIATIONS
An. Amer. Acad. Polit. Soc. Sci. ANNALS OF THE AMERICAN ACAD-
EMY OF POLITICAL AND SOCIAL
SCIENCES
Broadcast News
Broad. Eng. Jour.
Communications
Elec. Eng.
BROADCAST NEWS
BROADCAST ENGINEERS JOURNAL
COMMUNICATIONS
ELECTRICAL ENGINEERING (TRANSACTION A.I.E.E.)
Electronics
Electronic Ind.
ELECTRONICS
ELECTRONIC INDUSTRIES
FM AND TELEVISION
FM BUSINESS
INTERNATIONAL PROJECTIONIST
JOURNAL OF APPLIED PHYSICS
JOURNAL OF THE FRANKLIN INSTITUTE
JOURNAL OF THE OPTICAL SOCIETY
OF AMERICA
JOURNAL OF THE SOCIETY OF
MOTION PICTURE ENGINEERS
JOURNAL OF THE TELEVISION
SOCIETY
PHYSICAL REVIEW
PROCEEDINGS OF THE INSTITUTE OF
RADIO ENGINEERS
FM and Tele.
FM Business
Inter. Project
Jour. Appl. Phys.
Jour. Frank. Inst.
Jour. Opt. Soc. Amer.
Jour. Soc. Mot. Pic. Eng.
Jour. Tele. Soc.
Phys. Rev.
Proc. I.R.E.
QST
Radio and Tele.
Radio Craft
Radio Eng.
Radio News
Radio Tech. Digest
RCA Rad. Serv. News
RCA REVIEW
RMA Eng.
Short Wave and Tele.
TBA Annual
Tele. News
Televsser
Television
QST (A.R.R.L.)
RADIO AND TELEVISION
RADIO CRAFT
RADIO ENGINEERING
RADIO NEWS
RADIO TECHNICAL DIGEST
RCA RADIO SERVICE NEWS
RCA REVIEW
RMA ENGINEER
SHORT WAVE AND TELEVISION
ANNUAL OF THE TELEVISION
BROADCASTERS ASSOCIATION
TELEVISION NEWS
TELEVISER
TELEVISION
www.americanradiohistory.com
.-11'PF. \7)LV
501
TELEVISION BIBLIOGRAPHY
Year
"The Selection of Standards for Commercial Radio Television ", J.
Weinberger, T. A. Smith and G. Rodwin, Proc. I.R.E. (September) 1929
"Television with Cathode-Ray Tube for Receiver ", V. K. Zworykin,
Radio Eng. (December)
19.29
"Fidelity Tests for Television Systems ", A. F. Murray, Tele. News
(December)
"Description of Experimental Television Receivers ", G. L. Beers,
Proc. I.R.E. (December)
"An Experimental Television System ", E..W. Engstrom, Proc. I.R.E
(December)
"Description of Experimental Television Transmitting Apparatus ",
R. D. Kell, Proc. I.R.E. (December)
"A Study of Television Image Characteristics ", E. W. Engstrom,
Proc. I.R.E. (December)
"Description of an Experimental Television System and the Kinescope ",
V. K. Zworykin, Proc. I.R.F. (December)
"Problems of Cathode -Ray Television ", I. G. Maloff, Electronics
(January)
"The Iconoscope -A Modern Version of the Electric Eye ", V. K.
Zworykin, Proc. I.R.E. (January)
"Transmission and Reception of Centimeter Waves ", I. Wolff, E. G
Linder and R. A. Braden, Proc. I.R.E. (January)
"Television ", V. K. Zworykin, Jour. Frank. Inst. (January)
"Improved Magnetron Oscillator for the Generation of Microwaves ",
E. G. Linder, Phys. Rev. (May 1)
"A Study of Television Image Characteristics ", E. W. Engstrom,
Jour. Soc. Mot. Pic. Eng. (May)
"An Experimental Television System -The Transmitter ", R. D. Kell,
A. V. Bedford and M. A. Trainer, Proc. I.R.E. (November)
"An Experimental Television System -The Receivers ", R. S. Holmes,
W. L. Carlson and W. A. Tolson, Proc. I.R.E. (November)
"The Radio-Relay Link for Television Signals ", C. J. Young, Proc
I.R.E. (November)
"Theory of Electron Gun ", I. G. Maloff and D. W. Epstein, Proc. I.R.E
(December)
"Cathode -Ray Tubes and Their Applications", J. M. Stinchfield, Elec
Eng. (December)
"Transmission and Reception of Centimeter Waves ", I. Wolff, E. G
Linder and R. A. Braden, Proc. I.R.E. (January)
"The Secondary Emission Phototuhe ". H. Iams and B. Salzberg. Prnr
I.R.E. (January)
"A Study of Television Image Characteristics, (Part II) ", E. W.
Engstrom, Proc. I.R.E. (April)
"Television and the Motion -Picture Theater ", A. N. Goldsmith, Inter
Project. (May)
"Applications Illustrating the Use of Cathode -Ray Tubes ", J. M
Stinchfield, Electronics (May)
"Development of Transmitters for Frequencies Above 300 Megacycles
",
N. E. Lindenblad, Proc. I.R.E. (September)
"Luminescent Materials for Cathode -Ray Tubes ", T. B. Perkins and
H. W. Kaufman, Proc. I.R.E. (November)
"Cathode -Ray Tube Terminology ", T. B. Perkins, Proc. I.R.E. (November)
"Influence of Tube and Circuit Properties in Electron Noise ", S. W
Seeley and W. S. Barden, Electronics (December)
"Television". A. V. Bedford and E. W. Engstrom, Section of BOOK OF
KNOWLEDGE
www.americanradiohistory.com
1931
1933
1933
1933
1933
1933
1934
1934
1934
1934
1934
1934
1934
1934
1934
1934
1934
1935
1935
1935
1935
1935
1935
1935
1935
1935
1986
TELEVISION, Volume IV
502
Year
"A Survey of Television Progress in America ", W. R. G. Baker and L.
Malter, Booklet. Geo. Newnes Ltd., England
"Possibilities of the Iconoscope in Television ", V. K. Zworykin and
G. A. Morton, Booklet. Geo. Newnes, Ltd., England
"Input Resistance of Vacuum Tubes as Ultra- High- Frequency Amplifiers", W. R. Ferris, Proc. I.R.E. (January)
"A Turnstile Antenna for Use at Ultra -High Frequencies ", G. H.
Brown, Electronics (March)
"The Secondary Emission Multiplier -A New Electronic Device ",
V. K. Zworykin, G. A. Morton and L. Malter, Proc. I.R.E. (March)
"Scanning Sequence and Repetition Rate of Television Images ", R. D.
Kell, A. V. Bedford and M. Á. Trainer, Proc. I.R.E. (April)
"Applied Electron Optics ", V. K. Zworykin and G. A. Morton, Jour.
Opt. Soc. Amer. (April)
"The Electron Image Tube ", V. K. Zworykin, Broadcast News (April)
"An Urban Field Strength Survey at 30 and 100 Mc. ", R. S. Holmes
and A. H. Turner, Proc. I.R.E. (May)
"The Future of Radio and Public Interest, Convenience and Necessity ",
David Sarnoff, RCA REVIEW (July)
"Television ", C. B. Jolliffe, Statements Made by RCA Before FCC,
RCA Institutes Technical Press, New York, N. Y. (June 15)
TELEVISION, Vol. I, RCA Institutes Technical Press, New York,
1935
1935
1936
1936
1936
1936
1936
1936
1936
1936
1936
(July)
1936
N. Y. (July)
1936
N. Y.
"The Cathode -Ray Tube in Television Reception ", I. G. Maloff, TELEVISION, Vol. I, RCA Institutes Technical Press, New York,
"Television in Advertising ", David Sarnoff, TELEVISION, Vol. I, RCA
Institutes Technical Press, New York, N. Y. (July)
1936
"Television", David Sarnoff, TELEVISION, Vol. I, RCA Institutes
Technical Press, New York, N. Y. (July)
1936
"RCA Television Field Tests ", L. M. Clement and E. W. Engstrom,
RCA REVIEW (July)
1936
"Iconoscopes and Kinescopes in Television ", V. K. Zworykin, RCA
REVIEW (July)
Jour. Soc. Mot. Pic. Eng. (May)
"Electron Optical System of Two Cylinders as Applied to Cathode Ray Tubes ", D. W. Epstein, Proc. I.R.E. (August)
"Magnetron Oscillators for the Generation of Frequencies Between
300 and 600 Megacycles ", G. R. Kilgore, Proc. I.R.E. (August)
"Ultra- High -Frequency Transmission Between RCA Building and
Empire State Building in New York City", P. S. Carter and G. S.
Wickizer, Proc. I.R.E. (August)
"Electrical Measurements at Wavelengths Less than Two Meters ",
L. S. Nergaard, Proc. I.R.E. (September)
"Television Radio Relay ", B. Trevor and O. E. Dow, RCA REVIEW
(October)
"Amateur Applications of the Magic Eye", L. C. Waller, QST (October
and November)
"Electron Beams and Their Applications in Low-Voltage Devices ",
H. C. Thompson, Proc. I.R.E. (October)
"Electron Optics of an Image Tube ", G. A. Morton and E. G. Ramberg,
Physics (December)
"Fourth Estate Views Television ", David Sarnoff, Broadcast News (December)
1936
1937
1936
1936
1936
1936
1936
1936
1936
1936
1936
"Partial Suppression of One Side Band in Television Reception ",
W. J. Poch and D. W. Epstein, Proc. I.R.E. (January)
1937
"Equipment Used in the Current RCA Television Field Tests ", R. R.
Beal, RCA REVIEW (January)
1937
"Some Notes on Ultra- High- Frequency Propagation ", H. H. Beverage,
1937
RCA REVIEW (January)
www.americanradiohistory.com
APPENDIX
"Frequency Assignments for Television ", E. W. Engstrom and C. M.
Burrill, RCA REVIEW (January)
"Television and the Electron ", V. K. Zworykin, Short Wave and Tele
' (March)
"Experimental Studio Facilities for Television ", O. B. Hanson, RCA
REVIEW (April)
"Television Studio Design ", R. E. Shelby and R. M. Morris, RCA
REVIEW (July)
"Television Transmitters Operating at High Powers and Ultra -High
Frequencies ", J. W. Conklin and H. E. Gihring, RCA REVIEW
(July)
"A Circuit for Studying Kinescope Resolution ", C. E. Burnett, Proc
I.R.E. (August)
"Theoretical Limitation of Cathode -Ray Tubes ", D. B. Langmuir,
Proc. I.R.E. (August)
"An Oscillograph for Television Development ", A. C. Stocker, Proc.
I.R.E. (August)
"The Brightness of Outdoor Scenes and Its Relation to Television
Transmission ", W. H. Hickok, R. B. Janes and H. Iams, Proc.
I.R.E. (August)
"Development of the Projection Kinescope ", V. K. Zworykin and W. H.
Painter, Proc. I:R.E. (August)
"Theory and Performance of the Iconoscope", V. K. Zworykin, G. A.
Morton and L. E. Flory, Proc. I.R.E. (August)
"High- Current Electron Gun for Projection Kinescope ", R. R. Law,
Proc. I.R.E. (August)
"RCA Developments in Television ", R. R. Beal, Jour. Soc. Mot. Pic
Eng. (August)
"Television Pickup Tubes with Cathode -Ray Beam Scanning ", H. Iams
and A. Rose, Proc. I.R.E. (August)
"Television Engineering Study Outline ", S. W. Seeley. RCA Licensee
Bulletin LB-405 (August 13)
TELEVISION, Vol. II, RCA Institutes Technical Press, New York,
N. Y. (October)
"Analysis and Design of Video Amplifiers ", Part I, C. N. Kimball and
S. W. Seeley, RCA REVIEW (October)
"The Magnetron as a High- Frequency Generator ", G. R. Kilgore, Jour.
Appl. Phys. (October)
"Field Strength Observations of Transatlantic Signals, 40 to 45 Megacycles", H. O. Peterson and D. R. Goddard, RCA REVIEW
(October)
Proc. I.R.E.
"Commercial Television-And Its Needs ", A. N. Goldsmith. TELEVISION, Vol. II, RCA Institutes Technical Press, New York, N. Y.
(October)
"Television ", David Sarnoff, RCA REVIEW (October)
"Televisual Use of Ultra -High Frequencies ", A. N. Goldsmith. TELEVISION, Vol. II, RCA Institutes Technical Press, New York, N. Y.
(October)
"What of Television ? ", David Sarnoff. TELEVISION, Vol. II, RCA
Institutes Technical Press, New York, N. Y. (October)
"Problems Concerning the Production of Cathode Ray Tube Screens ",
H. W. Leverenz. TELEVISION, Vol. II, RCA Institutes Technical
Press, New York, N. Y. (October)
"Television- Description for Laymen ", A. F. Van Dyck. TELEVISION, Vol. II, RCA Institutes Technical Press, New York, N. Y.
(October)
"A Transformation for Calculating the Constants of Vacuum Tubes
with Cylindrical Elements ", W. van B. Roberts, Proc. I.R.E.
(October)
"Television Economics ", A. N. Goldsmith, Communications (October)
www.americanradiohistory.com
503
Y ear
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
1937
TELEVISION, Volume IV
504
Year
"Television Among the Visual Arts ", A. N. Goldsmith, TELEVISION,
Vol. II, RCA Institutes Technical Press, New York, N. Y. (October)
"A New Method of Remote Control ", S. W. Seeley, H. B. Deal and
C. N. Kimball, RCA Licensee Bulletin LB -398 (October 29)
"Screens for Television Tubes ", I. G. Maloff and D. W. Epstein,
Electronics (November)
"Figure of Merit for Television Performance ", A. V. Bedford, RMA
Eng. (November)
RCA REVIEW (July)
"Direct Viewing Type Cathode -Ray Tube for Large Television Images ",
I. G. Maloff, Proc. I.R.E. (November)
RCA REVIEW (January)
"Television Cathode -Ray Tubes for the Amateur ", R. S. Burnap, RCA
REVIEW (January)
"Video I. F. System Considerations", S. W. Seeley and W. S. Barden,
RCA Licensee Bulletin LB-417 (January 5)
"A Discussion on Television Receiving Antennas", W. S. Barden and
S. W. Seeley, RCA Licensee Bulletin LB -423 (January 27)
"Problems of American Television ", A. N. Goldsmith, Radio Craft
(February)
"Television Receivers ", E. W. Engstrom and R. S. Holmes, Electronics (April)
"Some Notes on Video -Amplifier Design ", A. Preisman, RCA REVIEW (April)
"Effect of the Receiving Antenna on Television Reception Fidelity ",
S. W. Seeley, RCA REVIEW (April)
"Probable Test Equipment Requirements for Design and Test of
Domestic Television Receivers ", J. M. Brumbaugh, RMA Eng
(May)
"Preliminary Discussion of Television Requirements With the Underwriters' Laboratories ", E. T. Dickey, RMA Eng. (May)
"Television I-F Amplifiers ", E. W. Engstrom and R. S. Holmes,
Electronics (June)
"A Discussion on Video Modulation Detection ", W. S. Barden, RCA
Licensee Bulletin LB -435 (June 27)
"A Discussion of Television Deflecting Systems ", S. W. Seeley and
C. N. Kimball, RCA Licensee Bulletin LB -433 (June 30)
"Television Antenna for Good Reception ", S. W. Seeley, Radio Craft
(August)
"Television V -F Circuits", R. S. Holmes and E. W. Engstrom, Electronics (August)
"High Frequency Correction in Resistance -Coupled Amplifiers ", E. W
Herold, Communications (August)
"Television Development and Test Equipment ", H. B. Deal, RCA
Licensee Bulletin LB -447 (August 30)
"Improvements in High Frequency Amplifiers", D. E. Foster and A. E
Newlon, RCA Licensee Bulletin LB-450 (September 22)
"Building Television Receivers with Standard Cathode -Ray Tubes ",
J. B. Sherman, QST (October)
"Review of Ultra- High- Frequency Vacuum-Tube Problems ", B. J
Thompson, RCA REVIEW (October)
"A Survey of Ultra-High- Frequency Measurements ", L. S. Nergaard,
RCA REVIEW (October)
"Selective Side -Band Versus Double Side -Band Transmission of Telegraph and Facsimile Signals ", B. Trevor, J. E. Smith and P. S
Carter, RCA REVIEW (October)
"A Video Mixing Amplifier ", A. A. Barco, RCA Licensee Bulletin
LB -453 (October 11)
"An Experimental Television Receiver ", D. E. Foster and G. Mount joy, RCA Licensee Bulletin LB -458 (October 13)
www.americanradiohistory.com
1937
1937
1937
1937
1938
1937
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
1938
APPENDIX
505
"Iconoscope Pre -Amplifier ", A. A. Barco, RCA Licensee Bulletin
LB-448 (October 17)
"Transmission Lines as Coupling Elements in Television ", S. W. Seeley
and C. N. Kimball, RCA Licensee Bulletin LB -456 (October 17)
"Analysis and Design of Video Amplifiers ", Part II, C. N. Kimball
and S. W. Seeley, RCA Licensee Bulletin LB -461 (October 20)
Jour. Tele. Soc. (October)
RCA REVIEW (January)
"Television Synchronizing and Blanking Signal Generator ", H. B
Deal, RCA Licensee Bulletin LB -452 (October 27)
"A Practical Television Receiver for the Amateur ", C. C. Shumard,
QST (December)
"Television ", A. F. Van Dyck, World Almanac
CATHODE -RAY TUBE IN TELEVISION. I. G. Maloff and D. W.
Epstein, McGraw-Hill Book Co., New York, N. Y.
"Practical Application of an Ultra-High- Frequency Radio -Relay Circuit", J. E. Smith, F. H. Kroger and R. W. George, Proc. I.R.E
(November)
"Television Synchronization ", E. W. Engstrom and R. S. Holmes,
Electronics (November)
"Deflection Circuits in Television Receivers ", E. W. Engstrom and
R. S. Holmes, Electronics (January)
"New Television Amplifier Receiving Tubes ", A. P. Kauzmann, RCA
REVIEW (January)
"Observations on Sky -Wave Transmission on Frequencies Above 40
Megacycles ", D. R. Goddard, Proc. I.R.E. (January)
RCA REVIEW (January)
"A Fixed -Focus Electron Gun for Cathode -Ray Tubes ", H. Iams,
Proc. I.R.E. (February)
"An Ultra- High- Frequency Power Amplifier of Novel Design ", A. V
Haeff, Electronics '(February)
"Some Television Problems from the Motion Picture Standpoint", G. L
Beers, E. W. Engstrom and I. G. Maloff, Jour.. Soc. Mot. Pic. Eng
(February)
Inter. Project (February)
"Television Economics ", Part I, A. N. Goldsmith, Communications
(February)
'
"Using Electromagnetic -Deflection Cathode -Ray Tubes in the Television Receiver ", J. B. Sherman, QST (February)
"Electrostatic Deflection Kinescope Unit for the Television Receiver",
J. B. Sherman, QST (March)
"Television Economics ", Part II, A. N. Goldsmith, Communications
(March)
"Application of Motion Picture Film to Television ", E. W. Engstrom
and G. L. Beers, Jour. Soc. Mot. Pic. Eng. (April)
"Field Strength Measuring Equipment for Wide -Band U -H-F Transmission", R. W. George, RCA REVIEW (April)
"Gamma and Range in Television ", I. G. Maloff, RCA REVIEW
(April)
"Kinescopes for Television Receivers ", L. C. Waller, Communications
(April)
"Measurement of Phase Shift in Television 'Amplifiers", A. A. Barco,
RCA REVIEW (April)
"Power for Television Receivers ", E. W. Engstrom and R. S. Holmes,
Electronics (April)
"Television Economics ", Part HI, A. N. Goldsmith, Communications
(April)
"Television Transmitting Antenna for Empire State Building", N. E
Lindenblad, RCA REVIEW (April)
"Transient Response of Multistage Video -Frequency Amplifiers ", A. V
Bedford and G. L. Fredendall, Proc. I.R.E. (April)
www.americanradiohistory.com
Year
1938
1938
1938
193q
1939
1938
1938
1938
1938
1938
1938
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
TELEVISION, Volume IV
506
Year
"Radio Frequency Generator for Television Receiver Testing ", A. H.
Turner, RMA Eng. (May)
"Television Economics ", Part IV, A. N. Goldsmith, Communications
(May)
"Television Lighting ", W. C. Eddy, Communications (May)
"How NBC Television Evolved ", O. B. Hanson, Radio and Tele. (June)
"Opportunities in Television ", A. N. Goldsmith, Radio and Tele. (June)
"Television Economics ", Part V, A. N. Goldsmith, Communications
(June)
"Simplified Television I -F Systems ", G. Mountjoy, RCA Licensee
Bulletin LB -478 (June)
"Circuit Diagrams of Television Signal Generating Equipment ", E. I.
Anderson, RCA Licensee Bulletin LB -479 (June)
"An Iconoscope Pre -Amplifier ", A. A. Barco, RCA REVIEW (July)
"Antennas ", H. H. Beverage, Radio and Tele., (July)
RCA REVIEW (July)
"Application of Motion Picture Film to Television ", E. W. Engstrom,
G. L. Beers and A. V. Bedford, Jour. Soc. Mot. Pic. Eng. (July)
RCA REVIEW (July)
"A Television Demonstration System for the New York World's Fair ",
D. H. Castle, RCA REVIEW (July)
"Effect of Electron Transit Time on Efficiency of a Power Amplifier ",
A. V. Haeff, RCA REVIEW (July)
"Luminescent Materials ", H. W. Leverenz and F. Seitz, Jour. Appl.
Phys. (July)
"Planning Programs for Television ", T. H. Hutchinson, Radio and
Tele. (July)
"Probable Influences of Television on Society ", David Sarnoff, Jour.
Appl. Phys. (July)
"Television Economics ", Part VI, A. N. Goldsmith,. Communications
(July)
"Television Lighting -Part I ", W. C. Eddy, Jour. Soc. Mot. Pic. Eng.
(July)
"Television Receiving and Reproducing Systems ", E. W. Engstrom,
Jour. Appl. Phys. (July)
"Television Studio Technic ", A. W. Protzman, Jour. Soc. Mot. Pic. Eng.
(July)
"Contrast in Kinescopes ", R. R. Law, Proc. I.R.E. (August)
"Television Antennas and Their Installation ", W. Hollander Bohlke,
Radio and Tele. (August)
"Television Economics", Part VII, A. N. Goldsmith, Communications
(August)
"Recent Improvements in the Design and Characteristics of Iconoscopes ", R. B. Janes and W. H. Hickok, Proc. I.R.E. (September)
"Space Charge Effects in Electron Beams ", A. V. Haeff, Proc. I.R.E.
(September)
"Television Pickup Tubes Using Low -Velocity Beam Scanning ", A.
Rose and H. Iams, Proc. I.R.E. (September)
"Television Economics ", Part VIII, A. N. Goldsmith, Communications
(September)
"The Image Iconoscope ", H. Iams, G. A. Morton and V. K. Zworykin,
Proc. I.R.E. (September)
"A Theoretical Analysis of Single Side -Band Operation of Television
Transmitters ", L. S. Nergaard, Proc. I.R.E. (October)
"Programming the Television Mobile Unit ", T. H. Hutchinson, RCA
REVIEW (October)
"Simple Television Antennas", P. S. Carter, RCA REVIEW (October)
"Television Economics ", A. N. Goldsmith, Radio Tech. Digest
(October)
www.americanradiohistory.com
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
1939
APPENDIX
507
Year
"Television Economics ", Part IX, A. N. Goldsmith, Communications
(October)
"Television Signal- Frequency Circuit Considerations ", G. Moun'tjoy,
RCA REVIEW (October)
"The Orthicon, a Television Pickup Tube ", A. Rose and H. Iams, RCA
REVIEW (October)
"Ultra -High- Frequency Propagation Formulas ", H. O. Peterson, RCA
REVIEW (October)
"Video Output Systems ", D. E. Foster and J. A. Rankin, RCA Licensee
Bulletin LB-494 (October)
"Monitor Kinescope ", C. N. Kimball, RCA Licensee Bulletin LB -500
(November)
"Television Economics", Part X, A. N. Goldsmith, Communications
(November)
"Transatlantic Reception of London Television Signals ", D. R. Goddard, Proc. I.R.E. (November)
"Television Economics ", Part XI, A. N. Goldsmith, Communications
(December)
"Superheterodyne Converter System Considerations in Television Receivers", E. W. Herold, RCA REVIEW (January)
"A New Method for Determining Sweep Linearity ", S. W. Seeley and
C. N. Kimball, RCA REVIEW (January)
"Simplified Television I -F Systems ", G. Mountjoy, RCA REVIEW,
1939
1939
1939
1939
1939
1939
1939
1939
1939
1940
1940
(January)
1940
(January)
1940
"Television Reception in an Airplane ", R. S. Holmes, RCA REVIEW
"RCA Television Field Pickup Equipment ", T. A. Smith, RCA REVIEW (January)
"The Formation and Maintenance of Electron and Ion Beams ", L. P.
Smith and P. L. Hartman, Jour. Appl. Phys. (March)
"A Wide -Band Inductive -Output Amplifier ", A. V. Haeff and L. S.
Nergaard, Proc. I.R.E. (March)
"Selective Sideband Transmission in Television ", R. D. Kell and G. L.
Fredendall, RCA REVIEW (April)
"Mobile Field Strength Recordings of 49.5, 83.5 and 142 Mc. from
Empire State Building, New York -Horizontal and Vertical Polarization", G. S. Wickizer, RCA REVIEW (April)
"Television Studio Technic", A. W. Protzman, RCA REVIEW (April)
"Television Lighting ", W. C. Eddy, RCA REVIEW (April)
"Low Cost Television Receiver ", G. Mountjoy and D. E. Foster, RCA
Licensee Bulletin LB -520 (May)
"Antennas and Transmission Lines at the Empire State Building ",
N. E. Lindenblad, Communications (May)
(April and May)
"A Receiver for the New Amateur Television System ", J. B. Sherman,
1940
1940
1940
1940
1940
1940
1940
1940
1940
1941
QST (June)
1940
"Picture Signal Analyzer ", H. B. Deal, RCA Licensee Bulletin LB -525
(July)
1940
"Field Strength Measuring Equipment at 500 Megacycles ", R. W
George, RCA REVIEW (July)
1940
"Determination of Optimum Number of Lines in a Television System ",
R. D. Kell, A. V. Bedford and G. L. Fredendall, RCA REVIEW
(July)
1940
"A 500 -Megacycle Radio Relay Distribution System for Television ",
F. H. Kroger, B. Trevor and J. E. Smith, RCA REVIEW (July)
1940
"Optimum Efficiency Conditions for White Luminescent Screens in
Kinescopes ", H. W. Leverenz, Jour. Opt. Soc. Amer. (July)
1940
"An Efficient U.H.F. Unit for the Amateur Television Transmitter ",
L. C. Waller, QST
(July)
www.americanradiohistory.com
1940
TELEVISION, Volume IV
508
Year
"Field- Strength Survey, 52.75 Megacycles from Empire State Building", Q. S. Wickizer, Proc. I.R.E. (July)
"A Precision Television Synchronizing Signal Generator ", A. V. Bedford and J. P. Smith, RCA REVIEW (July)
"Some Factors Affecting the Choice of Lenses for Television Cameras ",
H. B. DeVore and H. Iams, Proc. I.R.E. (August)
"The RCA Portable Television Pickup Equipment ", G. L. Beers, O. H.
Schade and R. E. Shelby, Proc. I.R.E. (October)
"Cathodoluminescence as Applied in Television ", H. W. Leverenz, RCA
REVIEW (October)
"Vertical versus Horizontal Polarization", G. H. Brown, Electronics
(October)
"An Electrically- Focused Multiplier Phototube", J. A. Rajchman and
R. L. Snyder, Electronics (December)
TELEVISION, V. K. Zworykin and G. A. Morton, John Wiley & Sons,
New York, N. Y.
TELEVISION BROADCASTING, L. Lohr, McGraw -Hill Book Co
New York, N. Y.
THE MYSTERIES OF TELEVISION, A. F. Van Dyck, The House
of Little Books, New York, N. Y.
"Recent Developments in Television ", E. W. Engstrom, An. Amer.
Acad. Polit. Soc. Sci. (January)
"RCA -NBC Television Presents a Political Convention as First Long Distance Pick -Up ", O. B. Hanson, RCA REVIEW (January)
"A Vestigial Side -Band Filter for Use with a Television Transmitter",
G. H. Brown, RCA REVIEW (January)
"A New Ultra -High Frequency Tetrode and its Use in a One Kilowatt
Television Sound Transmitter ", A. K. Wing and F. E. Young,
Proc. I.R.E. (January)
"Cascade Amplifiers with Maximal Flatness ", V. D. Landon, RCA
REVIEW (January and April)
"Video Output Systems ", D. E. Foster and J. Rankin, RCA REVIEW
(April)
"A Resume of the Technical Aspects of RCA Theatre Television ",
I. G. Maloff and W. A. Tolson, RCA REVIEW (July)
"A Simplified Television System for the Radio Amateur and Experimenter", L. C. Waller and P. A. Richards, RCA REVIEW (October)
"A Method and Equipment for Checking Television Scanning Linearity", V. J. Duke, RCA REVIEW (October)
"Recent Television Developments ", R. E. Shelby and V. K. Zworykin,
Reports on Progress in Physics (British), (Vol. 8) (December)
"Orthicon Portable Television Equipment ", M. A. Trainer, Proc. I.R.E.
1940
1940
1940
1940
1940
1940
1940
1940
,
(January)
"Measurements of the Slope and Duration of Television Synchronizing
Impulses ", R. A. Monfort and F. J. Somers, RCA REVIEW (Janu-
ary)
"Factors Governing Performance of Electron Guns in Television
Cathode -Ray Tubes ", R. R. Law, Proc. I.R.E. (February)
"Television Reception with Built-in Antennas for Horizontally and
Vertically Polarized Waves", W. L. Carlson, RCA REVIEW
(April)
"The Relative Sensitivities of Television Pick -Up Tubes, Photographic
Film and The Human Eye ", A. Rose, Proc. I.R.E. (June)
"Analysis, Synthesis and Evaluation of the Transient Response of
Television Apparatus ". A. V. Bedford and G. L. Fredendall, Proc.
I.R.E.
(October)
1940
1940
1941
1941
1941
1941
1941
1941.
1941
1941
1941
1941
1942
1942
1942
1942
1942
1942
www.americanradiohistory.com
APPENDIX
509
Year
"A Portable High -Frequency Square -Wave Oscillograph for Television", R. D. Kell, A. V. Bedford and H. N. Kozanowski, Proc.
I.R.E. (October)
1942
"Automatic Frequency and Phase Control of Synchronization in Television Receivers ", K. R. Wendt and G. L. Fredendall, Proc. I.R.E.
(January)
1943
"Contemporary Problems in Television Sound ", C. L. Townsend, Proc.
I.R.E. (January)
"The Focusing View-Finder Problem in Television Cameras ", G. L.
Beers, Proc. I.R.E. (March)
Jour. Soc. Mot. Pic. Eng. (March)
"Television- Far -seeing Eye of the Future ", R. E. Shelby, Electronics
(March)
"R -F Operated High -Voltage Supplies for Cathode -Ray Tubes ", O. H
Schade, Proc. I.R.E. (April)
"Cathode -Ray Control of Television Light Valves ", J. S. Donal, Jr ,
Proc. I.R.E. (May)
"A Type of Light Valve for Television Reproduction ", J. S. Donal, Jr ,
and D. B. Langmuir, Proc. I.R.E. (May)
"Electron Bombardment in Television Tubes ", I. G. Maloff, Electronics
(January)
"Automatic Frequency Control of Synchronization in Television."
RCA Licensee Bulletin LB -624 (August)
"Reflective Optical System for Television ", E. W. Wilby, RCA Licensee
Bulletin LB -630 (November)
"Reflective Optics in Projection Television ", I. G. Maloff and D. W
Epstein, Electronics (December)
"Postwar Television Standards ", A. N. Goldsmith, Televiser (Fall)
"Possible Social Effects of Television ", David Sarnoff, An. Amer. Acad
Polit. Soc. Sci. (January)
"Television ", David Sarnoff, TBA Annual (January)
"Cathode Coupled Wide -Band Amplifiers ", RCA Licensee Bulletin
1943
1943
1943
1943
1943
1943
1943
1944
1944
1944
1944
1944
1945
1945
LB-631 (January)
1945
"Future of Theater Television ", A. N. Goldsmith, Television (February)
1945
"The Clamp Circuit ", C. L. Townsend, Broad. Eng. Jour. (February
and March)
"Improved Electron Gun for Cathode -Ray Tubes ", L. E. Swedlund,
Electronics (March)
"Projection Television ", D. W. Epstein and I. G. Maloff, Jour. Soc
Mot. Pic. Eng. (June)
"Multiple- Dwelling Television", A. N. Goldsmith, Television (June)
"Researcher Views Television ", E. W. Engstrom, Televiser (July August)
"Converting from Radio to Video Broadcasting", A. N. Goldsmith,
Televiser (September)
"Planning the Television Station, Part I ", A. N. Goldsmith, Televiser,
(Sept. -Oct.)
"Band -Pass Bridged T Network for Television Intermediate- Frequency
Amplifiers ", G. C. Sziklai and A. C. Schroeder, Proc. I.R.E. (Octo-
1945
1945
1945
1945
1945
1945
1945
1945
ber)
"Cathode- Coupled Wide -Band Amplifiers ", G. C. Sziklai and A. C
Schroeder, Proc. I.R.E. (October)
1945
"Planning the Television Station, Part II ", A. N. Goldsmith, Televiser,
(Nov. -Dec.)
1945
"Transmission of Television Sound on the Picture Carrier ", G. L
Fredendall, K. Schlesinger and A. C. Schroeder, Proc. I.R.E
(February)
"Practical Aspects of Television ", W. H. Bolke and N.
RCA Rad. Serv. News (February)
www.americanradiohistory.com
1946
M.
Brisbin,
1946
510
TELEVISION, Volume IV
Year
"Input Impedance of Several Receiving -Type Pentodes at F -M and
Television Frequencies ", F. Mural, RCA Licensee Bulletin LB -661
1946
(March)
"Improved Cathode -Ray Tubes with Metal- Backed Luminescent
1946
Screens ", D. W. Epstein and L. Pensak, RCA REVIEW (March)
"Local Oscillator Radiation and Its Effect on Television Picture Con1946
trast", E. W. Herold, RCA REVIEW (March)
"Image Orthicon Camera", R. D. Kell and G. C. Sziklai, RCA REVIEW
(March)
"Field Television ", R. E. Shelby and H. P. See, RCA REVIEW
(March)
"Compensating Amplifier ", C. N. Gillespie, Electronics (March)
"Super Turnstile Antenna ", R. F. Holtz, Communications (April)
"Practical Television ", R. A. Monfort, Radio News (May)
"An Experimental Color Television System ", R. D. Kell, G. L. Fredendall, A. C. Schroeder and R. C. Webb, RCA REVIEW (June)
"A Method of Measuring the Degree of Modulation of a Television
Signal ", T. J. Buzalski, RCA REVIEW (June)
"Development of an Ultra Low Loss Transmission Line for Television ",
E. O. Johnson, RCA REVIEW (June)
"The Maximum Efficiency of Reflex Oscillators ", E. G. Linder and R. D.
Sproull, Phys. Rev. (June 1 & 15)
"The Image Orthicon-A Sensitive Television Pickup Tube ", A. Rose
and P. K. Weimer and H. B. Law, Proc. I.R.E. (July)
"Micro -Wave Television Relays, Operating on 6,800 to 7,050 Mc ", W. J.
Poch .and J. P. Taylor, FM and Tele. (August)
"Television for Today -Part III ", M. S. Kiver, Radio Craft (August)
"The Relationship of FM to Television ", A. N. Goldsmith, FM Business
(August)
"Flying Torpedo with an Electric Eye ", V. K. Zworykin, RCA REVIEW (September)
"Naval Airborne Television Reconnaissance System ", R. E. Shelby,
F. J. Somers and L. R. Moffett, RCA REVIEW (September)
"Miniature Airborne Television Equipment ", R. D. Kell and G. C.
Sziklai, RCA REVIEW (September)
"Mimo- Miniature Image Orthicon ", P. K. Weimer, H. B. Law and
S. V. Forgue, RCA REVIEW (September)
"Current Oscillator Television Sweep ", G. C. Sziklai, Electronics (September)
"A Unified Approach to the Performance of Photographic Film, Television Pickup Tubes, and the Human Eye ", Jour. Soc. Mot. Pic.
Eng. ( October)
"Simultaneous All- Electronic Color Television ", RCA REVIEW (December)
"Television Equipment for Aircraft ", M. A. Trainer and W. J. Poch,
RCA REVIEW (December)
www.americanradiohistory.com
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
1946
www.americanradiohistory.com
www.americanradiohistory.com
www.americanradiohistory.com
(91761-n761)
AI aurnioA
uolslAa.i.
www.americanradiohistory.com
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement