Frequency stability standards for television synchronizing generators and methods of measurement.

Frequency stability standards for television synchronizing generators and methods of measurement.
Calhoun: The NPS Institutional Archive
Theses and Dissertations
Thesis Collection
1950
Frequency stability standards for television
synchronizing generators and methods of measurement.
Vanderburg, Elden Robert
Monterey, California. Naval Postgraduate School
http://hdl.handle.net/10945/24730
P'l
-:
-'
-;
•hill;;!;*:!:-';-:'
fTs^ii:Que^lCY
FOR
rHllJiVISION
stajriuty standards
,
SWCIiRO^WlZING GEMERATORS
AMD Mimm^S €f
W'tlASUREih^EIMT
ELDEN ROIMRT VANDB^BUJRG
ri'i'-!'-,''''
IjfijIllSi'
..;
UbiaxY
postgraduate
School.
FREQUENCY STABILITY STANDARDS
FOR TELEVISION SYNCHRONIZING GENERATORS
AND
IvIETHODS
OF IIEASUREI^aEl^
E, R. Vanderburg
ytb
FREC^XinMCY STABILITY 3TA1TDARDS
FOR TELEVISION SYMCHRONIZIMG GENERATORS
AND I^THODS OF I.IEASUREI^dENT
by
Elden Robert Vanderburg
Lieutenant Commander, United States Navy
Submitted in partial fulfillment
of the requirements
for the degree of
IIASTER OF SCIENCE
United States Naval Postgraduate School
Annapolis, Maryland
1950
This work is accepted as fulfilling
the thesis requirements for the degree of
MASTER OF SCIENCE
from the
United States Naval Postgraduate School*
PREFACE
For a period of eleven weeks, from
3
January to 18
March 1950, the author was associated with the Television
Terminal Equipment Group, Engineering Products Department,
Victor Division, of the Radio Corporation of America, at
Camden, New Jersey.
During this period a study was made
of the effects of unstable synchronizing pulse frequency,
a
method of measurement was devised, and a circuit was
designed v/hich permits continuous monitoring of the synchronizing generator frequency behavioPo
This work was undertaken at the suggestion of Mr.
W, J. Poch, Group Manager of the Television Terminal Equip-
ment Group,
Mr, Poch stated the need for a simplified
measuring method which utilized a monitor scope as the
indicating device.
The author wishes to further acknowledge the advice
and assistance given in connection with this work by Mr.
J. M.
Brumbaugh and other members of the Advance Develop-
ment Group under Mr. H. N, Kozanowski.
I
ia\
CONTENTS
Preface
ii
List of Illustrations
iv
Introduction
1
CHAPTER I
THE TELEVISION SYSTEM
Limitations
Scanning
Synchronization
3
3
3
8
CHAPTER II
SYNCHRONIZING SIGNAL FREQUENCY STABILITY CONSIDERATIONS
Receiver Requirements
Studio Requirements
Standards
Synchronizing Generator Lock-in Circuit
Effect of Erratic Power Line Frequency
Effect which Maximum df/dt of Horizontal
Synch May Produce in a Television Picture ...
14
l/<
15
15
IS
20
2/».
CHAPTER III
....
FREQUENCY MEASUREI^IENT OF SYNCHRONIZING SIGNALS
Average Repetition Rate of Synchronizing
Pulses
Rate of Change of Frequency of Horizontal
Synchronizing Pulses
Recommended Operational Method of Measurement of Average Repetition Rate and Rate
of Change of Frequency of Horizontal Synchronizing Pulses
Bibliography
30
30
30
35
44
(iii)
ILLUSTRATIONS
Figure
1
Title
Page
Odd-line interlaced scanning system with
13 lines
7
Effect of horizontal motion on vertical
edges in a 2-to-l interlaced system
.
«
7
3
Picture line amplifier standard output.
...
9
U
Detail of Figure
5
RCA TG-IA Synchronizing Generator Lock-in
2
.
.
10
3
Bridge Circuit
19
Curve showing relationship between tolerable
line frequency surge and duration,
23
7
Deviation Curve
27
8
Deviation Curve
28
9
Equipment set-up for rate of change of
frequency measurement
31
Masking of cathode ray tube face for rate
of change of frequency measurement
32
Graphical determination of the rate of
change of frequency of horizontal synch
pulses from oscillogram
32
Block diagram of circuit for measuring
drift rate of horizontal drive pulses
34
6
10
11
12
...
...
13
Frequency measurement circuit schematic
14
Idealized waveform-s of frequency measurement circuit
37
Frequency measurement indication of face
of monitor scope
39
15
(iv)
36
INTRODUCTION
Synchronism between television transmitters and receivers Is perhaps the most iF.portant element of a tele-
vision system,
V^ith the
incorporation of frequency stable
horizontal scanning oscillators in receivers the frequency
stability of transmitter synchronizing signals becomes a
matter of fundamental Importance.
It will be the purpose of the following pages to give
a complete presentation of the problems Incident to the
requirement for frequency stable synchronizing signals.
The first few pages are devoted to a brief review of
those parts of the television system pertinent to the problem of synchronization so that the reader may have a more
comprehensive insight into the reasoning behind the choice
of synchronizing signal frequency standards,
S3''nchronizlng signal frequency standards are discussed
and the effects these standards have on studio and receiving
equipment is analyzed.
It is shown that the present stand-
ards are a fair compromise of ideal requirements.
Not much thought was given to possible methods of
measurement when the standards were proposed or when later
adopted.
Available methods of measurement of synchronizing
signal frequency and rate of change of synchronizing signal
frequency are discussed.
In general these methods have
been unsatisfactory from an operational viewpoint,
A suggested method for making operational measurements
of synchronizing signal frequency and rate of change of
(1)
frequency is presented along with a circuit developed for
this purpose.
(2)
CHAPTER I
THE TELEVISION SYSTEM
1,
Limitations
Noise Is the most serious limitation in the television
system as in the sound system, and in general arises from
2 sources;
(l)
shot noise in vacuum tubes and thermal ag-
itation in vacuum tubes and other circuit elements, and (2)
pickup from associated or remote electrical apparatus.
Resolution of a T.V, system is affected by noise but
is more directly affected by the frequency bandwidth avail-
able for the transmission system.
The Federal Communica-
tions Commission has set this bandwidth at 6 megacycles
per second.
Lastly, there is the limitation imposed by the tech-
nological development of pickup and reproducing equipment.
Economic considerations have prescribed in general that
quality be stressed in transmitting equipment to provide
reliability and reduce the need for including in receivers
expensive corrective circuits.
2.
Scanning
The standard scanning system in television requires
the aperture to traverse the scene in essentially horizon-
tal lines from left to right, and progressively from top to
bottom.
The aperture moves at constant velocity during
actual scanning periods because this is a simple type of
motion to duplicate in the reproducing aperture and because
it provides a uniform light source in the reproducer.
(3)
Daring the retrace, or flyback period between lines the
scanning beam moves extremely rapidly from the end of one
line to the start of the next with a motion which need not
be linear.
Complete traversal of the scene mast be accom-
plished at a rate high enoagh to avoid the sensation of
Because most power systems in the United States
flicker.
are 60 cycle systems, and synchronization with the pov/er
system minimizes the effects of hum, and simplifies the
problem of synchronizing rotating machinery in the television
studio, 60 cycles has been chosen as the vertical scanning
frequency.
To accomplish synchronism between the scanning aperture in the pickup and reproducing parts of the system, syn-
chronizing information in the form of electrical pulses is
provided during the otherwise wasted retrace interval between successive lines and successive pictures.
Synchroni-
zing pulses are generated at the studio and control the
scanning of the pickup equipment as well as the scanning of
the aperture in the receiving equipment when they are re-
ceived as a part of the complete composite signal transmitted*
Vertical and horizontal resolving power of the system
is,
in part, a function of the number of scanning lines.
Obviously, the ability of the system to resolve fine detail
in the vertical direction will increase with the number of
lines.
However, as the number of lines increases, the band-
width of the system must also increase to accomodate the
(4)
greater resolution required in the horizontal direction.
Consideration of the related questions of channel width
and resolution has resulted in the present day usage of
525 lines.
A factor which influenced the choice of this
particular number of scanning lines is the need for an
exact integral relationship between horizontal and vertical scanning frequencies.
This relationship is attained in
the television system by the use of a series of electronic
counters made stable by limiting the characteristic count
of each to a small integer.
For example, R.C.A. synchro-
nizing generator equipment uses four counters, counting 7,
5,
5,
525,
and 3«
The combined product of these 4 numbers is
the number of lines per frame.
The product of 525
and 60 is 31,500 which is the master oscillator frequency.
Another counter divides the master oscillator frequency by
2 to
yield the required horizontal scanning frequency of
15,750 cycles per second.
To conserve bandwidth without sacrificing freedom from
flicker a standard 2-to-l interlaced system of scanning is
the present practice.
The sensation of flicker is related
to the frequency of illumination of the entire scene, and
has no relation to the number of lines nor to the frequency
of the lines.
Interlacing gives greater freedom from
flicker by scanning half of the lines, uniformly distributed
over the entire picture area during 1 vertical scan.
Thus
the effect of doubling the frequency of picture illumination
is achieved without changing the velocity of the scanning
(5)
beam.
The aperture scans alternate lines consecutively
from top to bottom after which the remaining lines, that
fall in between those included in the first operation, are
scanned consecutively from top to bottom,
A 13 line inter-
laced scanning system is illustrated in Figure 1,
Each
group of lines called a field consists of 262^ lines, 2
consecutive fields constitute a complete picture or frame
of 52$ lines.
Field frequency is 60 cycle, and frame fre-
quency is 30 cycles per second.
One fault of interlaced
scanning is horizontal break-up illustrated in Figure 2.
The series of waves and pulses generated during the
actual scanning line periods is the basic part of the tel-
evision signal.
During retrace periods the pickup tube may
produce spurious signals which along with retrace lines detract from the reproduced picture.
It is therefore desir-
able to add blanking pulses to the picture signal during
retrace periods.
Actually, blanking pulses are applied to
the scanning beams in both the camera pickup tube and in
the receiver reproducing tube.
Camera blanking pulses are
used only in the pickup tube and serve to close the scanning
aperture during retrace.
Picture blanking pulses are some-
what longer than camera blanking pulses and are made an integral part of the signal radiated to the receiver.
Picture
blanking pulses are simple rectangular pulses having a time
duration slightly longer than the actual retrace periods in
order to trim up the edges of the picture.
They are pro-
duced in the synchronizing generator from the same basic
(6)
Fig.
1
Odd- line interkceJ sainnin^
s:ysteffi
with
13 lines.
Con".ecutiv£ i\t\6s e^re mdTcdteci by solid and
dotted
lines, i-cs|>2ctively.
1~L
\
FiS.
2
Effect of
in
d
bnzonUl
Z-to-1
stdtn,iar_y.
(7)
inu!'i-.cr.]
Motion
in /bO sec.
motion on verticti ec!^es
'-'i^^.
i}\^\^^-^'
Lower otect movm^
^tjject
to ri^ht.
tiF:ing circuits which produce the scanning signals and
are therefore accurately synchronized with the retrace
periods.
Horizontal blanking pulses recur at intervals
of 1/15,750 of a second and are no longer than 18 percent
of the period between the start of successive scanning
lines.
At times corresponding to the bottom of the pic-
tures, horizontal blanking pulses are replaced by vertical
blanking pulses which are similar but have a duration ap-
proximately 15 scanning lines long.
Both blanking pulses,
with some detail, are shown in Figures
3
and 4.
The width of the vertical blanking pulse is not limited by circuit considerations as is that of the horizontal
pulse.
The limitation in this case is the requirement of
television film projectors of the intermittent type that
the scene be projected on the pickup tube only during the
vertical blanking period.
Hence, the blanking period must
be of sufficient duration so that there is adequate storage
of photo-electric charges on the sensitive surface of the
pickup tube,
3.
Synchronization
The horizontal and vertical scanning systems in tele-
vision receivers are two separate and independent circuits,
each requiring accurate synchronizing information to keep
them in step with the corresponding scanning system in the
camera which originates the signal.
The synchronizing
system employed utilizes horizontal and vertical pulses of
equal amplitude but of different wave shape.
(8)
By this means
©
®
^i
,M
i
g
^
£.
C
O
«j
I'l
*}
<J
Ni
1 1
w
y
X
-1
0-
^
^
-<;
•w.
«-j
J;
H
%t
c
Vj
*^
\)
f
^
V
C:
1
i3
>1
t-
*3
^
^
^
*
M
^2
c
V
i4
K.
V
V.
«
.^
4^
>
1=^
M
tf)
V
.«
.43
^^
I
«h
^1
•~5
>0
'^5
^
5
^
<i
.
:?
s
^ 1
1
•X'
h
1
C-<
Ki
I
^
•h
(9)
HI
>>
^
>,
"^
£
'
«
^
e/-^^^)
«s
V
^
^.
^
r^-
-^
r
•Cv.-
ti'O
1^
<:
^
•<i
^'^
^*
«>i"
^
^
•n
o
•-
-*
^1-0
o
c>
"2^ M
^
1-
"^
**^J:
tfe
u
^^^
X
i
k
5^
«
^1
1
^
^t
^v,-
\
'•9
2:0
OCT
ceo
»o
^
^
1j
4
2
LU
<
o
^*>
<c
no
(10)
of synchronization the pulses may be added to the picture
signal without overloading the transmitter or reducing the
power available for the picture signal, and simple frequency
discrimination may be used to separate the horizontal and
vertical pulses.
Figure
3
shows the addition of synchroni-
zing pulses in the composite signal.
All synchronizing pulses appear below black level and
hence can have no effect on the tonal gradation of the picture.
Horizontal synchronizing pulses, except those occur-
ring during the first portion of the vertical blanking period, are all simple rectangular pulses appearing at the
negative base of the horizontal blanking pulses and during
the latter portion of the vertical blanking pulses.
The
synchronizing pulses are considerably shorter then the blanking pulses, and their leading edge is set back from the
leading edge of the blanking pulse forming a step in the
composite pulse called the front porch.
This off -set at the
leading edge is made to insure that blanking has occurred
before the synchronizing pulse initiates horizontal retrace,
and to insure that any discrepancies which may occur in the
leading edge of the blanking pulses do not affect either
the timing or amplitude of the synchronizing pulses.
The
choice of pulse width of the horizontal synch, 8 percent of
the period between the start of successive lines, was in-
fluenced by several factors.
The width should be as great
as possible to make the energy content of the pulse large
when compared to the worst noise interference.
The width
should be no greater than necessary to meet the above con-
(11)
dition so that average power requirerrents of the transmitter
may be minimized.
Finally, the pulse width should be as
narrov; as possible to facilitate discrimination between the
horizontal pulses and the wider vertical pulses.
Vertical synchronizing pulses are basically rectangular in shape and are much longer than horizontal pulses.
In addition, each pulse is cut by 6 slots making it appear
to be a series of 6 wide pulses at twice horizontal pulse
frequency.
The slots contribute nothing to the vertical
synchronizing information but do provide a means of providing continuing horizontal synchronizing information.
Pre-
ceding and follov;ing the vertical pulse are groups of 6 narrow equalizing pulses at twice horizontal frequency whose
purpose is also to provide continuous horizontal synchronizing information.
Doubling the frequency of the equalizing
pulses and the slots in the vertical pulse provides an
arrangement whereby the choice of proper alternate pulses
makes available some type of horizontal synchronizing infor-
mation at the end of each horizontal scanning line in either
even or odd fields.
This arrangement also makes the verti-
cal synchronizing interval and both equalizing pulse inter-
vals exactly alike in both even and odd fields.
It should
be noted that the leading edge of the horizontal synchroni-
zing pulses and of the equalizing pulses, and the trailing
edge of the slots in the vertical pulse are responsible for
triggering the horizontal retrace in the receiver,
A most difficult problem in synchronizing is that of
maintaining accurate interlacing.
(12)
Discrepancies in either
the timing or amplitude of the vertical scanning of alter-
nate fields will cause displacement in space of the interlaced field.
This effect known as pairing causes non-uni-
form spacing of the scanning lines, reduction in vertical
resolution, and intensification of the line structure in
the picture.
The presence of minute 30 cycle component in
the vertical scanning will cause pairing.
Since alternate
fields are displaced with respect to each other by half a
line, the horizontal synchronizing signal will have an in-
herent 30 cycle component.
It is this situation and the
need to prevent the transfer of the 30 cycle component into
vertical deflection v/hich accounts for the introduction of
the double frequency narrow equalizing pulses.
Vertical
synchronizing pulses are separated from the composite synchronizing signal by suppressing the horizontal synchro-
nizing pulses in an integrating network.
The equalizing
pulses cause the integrating network to "forget" the dif-
ference between alternate fields by the time the vertical
synchronizing pulses begin*
Equalizing pulses are made
with half the width of normal horizontal pulses so that at
the transition from normal horizontal pulses to equalizing
pulses at double frequency, the a-c axis of the synchroni-
zing signal does not shift.
(13)
CHAPTER II
SYNCHRONIZBIG SIGNAL FREqUEr^ICY STABILITY CONSIDERATIONS
lo
Receiver Requirements
Horizontal scanning circuits of most receivers are
equipped with automatic frequency control in order to neutralize the deleterious effects of noise.
In contrast to
triggered circuits where each scanning line is individually
initiated by a pulse in the incoming signal, A.F.C. scan-
ning is governed by stable oscillators which in turn are
controlled by voltages obtained from phase comparison of
the incoming synchronizing pulses and the scanning signals
themselves.
The time constant of the comparison circuit
must be made long compared to the scanning period so that
random noise has little effect on the resulting control
voltage.
Such A.F.C. circuits are keyed to provide further
ixcmunization against noise by eliminating all noise pulses
except those occurring during the short keying interval.
As a result of A.F.C. scanning circuits in receivers,
a
very high order of frequency stability is required in the
horizontal synchronizing and blanking signals.
Frequency
modulation of the horizontal pulses is intolerable because
it causes the right and left hand edges of the blanked ras-
ter as well as vertical lines in the scene to assume the
same shape as the modulating wave.
Since retrace timing is
controlled by a stable oscillator which is not responsive
to short time changes of the synchronizing timing, horizon-
tal retrace begins along a straight line.
(U)
Variation in
synchronizing timing shows as a displacement of vertical
objects in the televised scene and in the edges of the
raster.
2,
Studio Requirements
As was stated in the foregoing chapter, the vertical
scanning frequency is locked in v/ith the 60 cycle power
mains to minimize hum effects and to simplify the synchro-
nization of rotating machinery.
The effect of hum pickup
in studio equipment, principally in the low level stages of
the video amplifiers, is to put a horizontal sinusoidal in-
tensity gradation in the televised picture commonly referred to as a hum bar.
If the vertical scanning frequency
is locked to the 60 cycle power supply frequency, the hum
bars v/ill remain fixed in phase and consequently they will
remain stationary across the picture.
They can then be
virtually eliminated by the addition of corrective sawtooth
and exponential voltages to the composite signal, i.e., the
hum bars are made to fade into the background shading.
How-
ever, if there is no lock-in, hum bars will slip across the
picture with a speed dependent on phase difference and corrective shading is impossible.
Even without the benefit of
corrective shading, stationary hum bars are much less objectionable than those in motion.
Hence, power main lock- in
makes hum pickup at receivers less noticable in those cases
where transmitter and receivers operate on the same or synchronized power systems,
3.
Standards
Because of the conflicting requirements of high hori-
(15)
zontal synchronizing pulse stability and synchronizing!: generator lock-in with power line frequency, minimum standards
for synchronizing signal tolerance have been adopted.
These
standards are as follows:
It shall be standard that the time of occurrence of the leading edge of any horizontal
pulse "N" of any group of twenty horizontal
pulses not differ from "IIH" by more than O.OOIH
where **H'* is the average interval between the
leading edges of horizontal pulses as determined by an averaging process carried out over
a period of not more than 100 lines.
It shall be standard that the rate of change
of the frequency of recurrance of the leading
edges of the horizontal synchronizing pulses
appearing in the picture line amplifier output
be not greater than 0,15 percent per second,
the frequency to be determined by an averaging
process carried out over a period of not less
than 20 or more than 100 lines, such lines not
to include any portion of the vertical blanking
signalo
It shall be standard that the frequency of
horizontal and vertical scanning pulses not vary
from the values established by the standards
of frame frequency and number of scanning lines
by more than tl percent regardless of variations
in frequency of the power source supplying the
television station.
The standard for the change in timing between suc-
cessive horizontal pulses is in need of revision.
This
standard was written in the days when synchronizing pulses
were picked off of a rotating scanning wheel.
cuits in present use are not restricted by it.
Counter cirThe standard
for rate of change of scanning frequency is considered a
fair compromise.
It is low enough so that satisfactory
automatic frequency control circuits with good noise immunity can be used, and it still permits the scanning frequency
(16)
to be shifted rapidly enough to follow the variation in
frequency of a large power system without more than a few
percent variation in phase relationship between the 60
cycle power wave and the vertical blanking signal of the
synchronizing generator.
The standard specifying the lim-
its of drift of the synchronizing frequency principally
provides a manufacturing tolerance on receiver automatic
frequency control circuits.
Under normal conditions large
power systems will not exceed this frequency standard.
Should unusual circumstances create a condition which made
power frequency run for extended periods outside the limiting frequencies of 59.4 to 60,6 cycles per second, it would
be necessary to switch the sjmchronizing generator from 60
cycle to crystal control.
The standard for rate of change of scanning frequency
is the
critical operating standard.
To maintain the least
phase difference between the vertical blanking or scanning
pulse, and the projection equipment exposure pulse, which
is desired during vertical blanking,
the synchronization
lock-in time constant should be made as fast as possible.
The time constant setting must be fast enough to keep the
phase difference between vertical synch, derived from the
master oscillator, and power supply frequencies within half
a cycle or the lock-in circuit v;ill lose control.
However,
it is not uncommon for power line frequency to vary quite
erratically within the limits established for maximum drift,
and the time constant setting must not permit the master
oscillator to follow so rapidly that the maximum allowable
(17)
rate of change of synchronizing pulse frequency is exceeded,
h*
Synchronizing Generator Lock-in Circuit
To facilitate an understanding of synchronizing gener-
ator lock-in control, a brief description of a typical lockin circuit employed in R.C.A, TG-lA Studio Synchronizing
Generator equipment can be given with the aid of the simplified diagram of Figure 5,
The frequency of the master
oscillator is determined by plate current in the reactance
tube.
Bias for the reactance tube is obtained from the 60
cycle lock-in circuit, which compares the 60 cycle output
pulse from the generator counters with the 60 cycle supply
voltage, and converts any phase difference into a d.c,
voltage*
The 60 cycle supply voltage, after clipping, is applied
to one corner of the lock-in bridge through the A.F.C, time
constant switch, 3-1, as a square wave.
The opposite cor-
ner of the bridge is connected through the 60 cycle position of the frequency control switch, S-2, to the grid of
the reactance tube.
The 60 cycle pulse from the synchroni-
zing generator is applied to the 2 remaining corners of the
bridge through a transformer, T-1, which has its secondary
connected in series with the parallel combination of
R/4.-C4,
When the 60 cycle pulse occurs, all the diodes are
caused to conduct, thus making possible a transfer of current in either direction between input and output terminals
of the bridge.
The 60 cycle pulse also creates a charge
across the combination
R4-C/4.
which is negative toward the
double plate corner of the bridge, and keeps the diodes
(18)
r3
u
o
1.
o
2^
•oo
£
c
i-
u
<
CQ
(19)
non-conducting during the interval between pulses.
master oscillator frequency is
ad.i'usted to
The
31,500 cycles
per second when the voltage on the reactance tube is zero.
If the frequency is exactly 31,500 cycles, the square wave
voltage applied to the bridge will be passing through zero
when the pulse from the counters causes the diodes to conduct.
No current will pass through the bridge circuit un-
der these conditions,
VThen the
frequency is slightly high-
er than 31,500 cycles per second, the resultant pulse will
occur sooner, while the square v/ave is negative.
Current
will be passed through the bridge, placing a negative charge
on condenser C5, and therefore on the grid of the reactance
tube, thereby reducing its mutual conductance and bringing
about a reduction in the frequenc3'' of the master oscillator.
A similar action takes place when the frequency of
the master oscillator is slightly lower than 31,500 cycles
per second.
The speed at which the charge on the reactance
tube grid follows changes in the relation between supply
voltage and master oscillator frequencies depends on the
size of the R-C combination in the grid return circuit of
the reactance tube, hence on the position of the time con-
stant switch S-1,
5,
Effect of Erratic Power Line Frequency
It is not uncommon for power line frequency to make
sudden shifts with changes in load.
As a matter of academ-
ic interest assume the power main frequency to suddenly
jump from one value to another.
If the frequency jumps its
maximum allowable limit from 59.4 to 60,6 cycles per second,
(20)
and if there is continuous raaxlraum df/dt, rate of change of
frequency of the synchronizing generator, 13.3/f seconds
will be required for lock-in.
Since normal horizontal scan-
ning frequency is 15,750 cycles per second, the allowable
maximum df/dt is equal to 0.0015 times 15,750 or 23.6 cycles
per secon(? per second.
Then from,
f 2 = f -L
where,
-h
(
df/dt )t
fg = final frequency
=
f]^
initial frequency
df/dt = rate of change of frequency
t = time
13.34 seconds.
t =
During these 13.34 seconds, a total of 8 hum bars would move
across the televised scene.
in a given time, t,
The number of lines, n, scanned
is given by the equation,
n
=
/fpdt
•'0
n
=
fit t
*(
df/dt )t*
The number of hum bars, N, which will move across the scene
is given by the expression,
fpt
N = _e
where, fg
fit
-
i
-
i-(
df/dt )t*
average frequency, or in this case the nom-
=
inal standard frequency, 15,750 c.p.s.
to
time for 1 field, I/60 sec.
=
(Z15)(15.?4)
N
=
-
(l/2)(2?.6)(l?.g4)*
262.5
(21)
=
^
However, consideration of lock-in circuit action discloses
that the discriminator output to the reactance tube changes
polarity after a slip of 1/2 a hum bar and gives a negative
correction.
The synchronizing generator could never lock-in
under these conditions.
Hold-in of control will be main-
tained, and proper phase restored on sudden shifts of line
frequency only if full frequency correction can take place
within a slip of 1/2 hum bars.
Thus, the maximum sustained
instantaneous shift in power line frequency for which lockin can be maintained may be calculated from a simultaneous
solution of the following
f1 = f
-
flti
then,
and,
t-i
f-].
-
=
-h
2
equations:
(df/dt)ti
fQti
- -|(df/dt)t*i
=
131.25
3.33 seconds
fQ = 78.5
cp.s.
The percentage frequency jump is therefore,
^1
-
^0
X 100 = 0,497?6
^0
Hence, lock-in can be maintained for sustained jumps in
line frequency of 0,29^ cycles.
Increasingly larger jumps
in line frequency can be tolerated for shorter periods than
that established in the above calculation provided the phase
difference between the
60 cycle pulse from the synchroni-
zing generator, and the comparison voltage from the power
supply does not exceed a half cycleo
6
The curves of Figure
show the relationship between tolerable line frequency
surge and surge duration.
While the conditions specified in making the fore(22)
J
m
li
44t
kmm
liL
§
:.t
lil^
wmi^^04^ri^:
2t-^
D ji
I
i!'
1.
V^A f
^,«;
'
1
'!
-rM^ ""•^t
•OlS;.
i;
^'
I.}
1
.
r^
-t-
,
t
t.
tS'
r.
o
'
ik-''^2fe'^.?!Ji:!
:*^:^':
3ii::
iio
^iti^'/lo
1
;t=1536
F-j,-
V-
^.
fr
t.
H^Wf^ij-J^rf4l4-i^f
H-^-L ni-rMi^-H{i^4t'- ir4riH;M :r:iHnH-;^jl-v1i^H^
(23)
rlfrirfniij-
j'
.
going calculation may never occur in actual practice, they
are a fair enough approximation that the results demonstrate
the need for maintaining as fast a time constant setting on
the synchronizing generator lock-in circuit as is possible
within tolerance.
Should power line frequency surges exceed
tolerable limits shown by the curves of Figure 6, it would
be necessary to increase the follow-up speed of the lock-in
circuit and exceed the minimum standard established for rate
of change of horizontal synchronizing pulses, or switch the
master oscillator to crystal control,
6,
Effect V^hicb Maximum df/dt of Horizontal Synch May
Produce in a Television Picture .
To assist in understanding the effect the maximum rate
of change of horizontal synchronizing pulses may produce in
a received television picture,
the following calculation has
been made to show the effect produced by sinewave frequency
modulation applied to the horizontal synchronizing pulses
with a peak amplitude corresponding to the maximum allowable
rate of change given by the standard.
Let
df/dt = 2Z.6 cosMt
where, f
=
<o
=
horizontal scanning frequency
2IT X frequency of applied sinewave of
frequency modulation, f^.
Integrating this expression one gets an equation for the
horizontal scanning frequency,
f =
fo+
(24)
%^
sln«rt.
where, fo
average horizontal scanning frequency.
=
Assume the television receiver in use to have a horizontal
oscillator locked in with an A.F.C, circuit having a time
constant sufficiently long so that its frequency always
corresponds to f^.
The result will then be that normally
straight vertical lines in the picture will have a displacement from a straight line varying in time at the frequency
of the applied modulation.
This displacement from a straight
line is found by integrating the second part of the above
expression for frequency,
sinot dt
2g.6
coswt,
—-J—
CO
The following table shows the value of the maximum displacement, D, obtained for several values of the modulating frequency, f^:
£m
Max. D
60
,000166 cycle
30
0OOO664
20
.00149
10
,00598
,0239
5
One of the most troublesome types of frequency modu-
lation encountered in synchronizing generator output signals is frequency modulation traceable to the power supply
frequency.
From the above table it is seen that the depar-
(25)
tare from a straigrht line allowed by the present standard
is only 0.0166% of the period for 1 horizontal line.
suming the receiver to have
a 4.5
As-
megacycle bandwidth there
will be approximately 570 picture elements per line.
There-
fore, the maximum allowable displacement for the case of
60 cycle modulation is approximately one tenth of a picture
element.
It is questionable if this small displacement
could be visually detected on the best receivers.
In the
case of 30 cycle modulation however, the maximum displace-
ment would be four tenths of a picture element, and
2
suc-
cessive fields would have the displacement in opposite directions so that at one point in the raster a total displace-
ment of eight tenths of a picture element would be possible.
The curves of Figures 7 and 8 have been calculated to
show the effect continuous maximum uniform rate of change
of horizontal synchronizing pulses will have on a received
television picture.
The receiver horizontal scanning oscil-
lator and received synchronizing pulses are assumed to be
in phase and at the nominal standard frequency of 15,750
c,p«s, at time equal to zero.
At time equal to zero the
synch frequency begins to change frequency at the maximum
allowable rate of 0.15 percent per second while the receiver
oscillator holds constant at 15,750 cp.s. because of long
time constant A.F.C, discriminator action.
The phase dif-
ference between the two frequencies is calculated from the
equation,
d = i(df/dt)t*
(26)
f~^^T^~r"T:
(^
-
r
I
i-
I
I
.
-
K!
I
t.
N-i'^i
!-^i
I
I
I
i
t'
!
—
J.:;.
•t.,
I.
'T
-|_4_j
-
:::..t-:.t
i:::l
IILU
l—i.
Ui
-rt
i;:-";;
--H
l^t:i:tM:l:::
;;•'••
I
Ui:
I-
mm
:|:
I
I
i-Hr;r*- .:N-f -t
•!•
-f l^n^^H^-t-^
;::
.;
:
-t:-
:j-:::
-ill:|:
ST.
)-
•--^3
E
'
^
3
3
-:
:-;t--
»
...
•..It
§{
g
*^
:c
H
_-.
..^.^.^
!_
cJ
^
Ji
-tt.-oc
— o - ^-b
^ _. in
I
ol
^
^-,
^
^ ^
b^
^S ^
^
.
1
;'
1
1
*-^
ifi
.=
ts
Si.
,
i
«
__p <
_f_i
O
J|L!>ad .S,tllKUVOi: lVXrslCIiaOH_ JNO: iO XN30. b3d N)
H-4-
—X
-J.
(27)
NQIIVIAIQ
'
.I
!
,
1
•ri"-i
:Li^i:..|.L
:(::
^-+"4
F-t:'i-
r '-^
1
1-.:
;
:-rj.-:
.:_;i
— f^"
V
-l-j—
;!
:.
—
——
i-
:
^
I
-
T••f
t
r^
IT"..
IM
'n
t-^
I
:
!
:
.
.J
<
.,
'H~N
......
i_.:^l^
Tl-*'
-rrrtM
Tig
^ ^
3i
^-f-":HT^
UJ
^ >
o
-^
S I
1
--
^iJ
:"
'
1
1
—
[[':
•;
-I
'
,•!.:•:
^ a
"*
..1.
.JZ-
-r
J..
i-
J-,;
..;.!.
^F-^-L
1-.
MS ^
f^
—J
2j
m5
i
o o
-: -^
to
-
ii
^
;.
'
r— rt
::r;|
•
—
t-'-" -.-Tr:
.,.-/
—
;
^1-
—
:.
:
_1_
i
I
:t~
Oi
^^
—
1
»
^ Tj
]±1
L
(28)
.
.t
where "d" is the deviation in cycles per second.
From the curve of Figure
7 it can be seen that a
nor-
mally straight line in the televised picture will be displaced at the bottom of the first field by approximately
0.33 percent of a horizontal scanning line, or by approx-
imately two picture elements.
This displacement will be
evident between successive fields and result in a loss of
horizontal resolution.
The curve of Figure 8 shows long
time deviation and assumes no correction of phase difference by the receiver oscillator.
If the studio should
switch pickups at such a time that the horizontal synch
shifted 50 percent of a horizontal scanning line, it is
seen from this curve that the maximum time for lock- in with
the new synchronizing signal is approximately
(29)
5
seconds.
CHAPTER III
FRE^UEITCY MEASUREIIENT OF SYNCHRONIZING SIGNALS
!•
Average Repetition Rate of Synchronizing. Pulses
Measurement of the repetition rate of the line and
field synchronizing pulses in the output of a television
transmitter serves as a check on the frequency dividing circuits in the synchronizing generator, and as a check on the
minimuin operating standard that the frequency not vary more
than tl percent.
This measurement Is relatively simple and
the standard method is to compare the horizontal and/or
vertical synchronizing pulses with a sinewave signal from
a signal generator by means of a cathode ray oscilloscope.
For continuous monitoring purposes, each divider stage in
the synchronizing generator may be equipped with a small
cathode ray tube to indicate its frequency division, thus
making any change in division of any stage immediately
apparent,
2,
Rate of Change of Frequency of Horizontal Synchronizing,
Pulses
The measurement of rate of change of frequency of hor-
izontal synchronizing pulses is a more complicated problem
than that of measuring the average or instantaneous frequency.
It is questionable if a completely satisfactory
operational measurement has been devised.
One method of
measurement which has been reduced to practice is outlined
below but it can be readily seen that it is most useful as
a
laboratory procedure.
An independent reference oscillator
of very high frequency stability is set at or very close to
(30)
a submultiple of the average horizontal synch frequency
and is used to synchronize a cathode ray oscilloscope sweep,
The connection of apparatus is shown in Figure 9.
Linear
\
Sv/eep
Oscillator
The
Stable
Reference
Oscillator
/
Composite
Synchronizing
Signal
Figure 9.
Equipment set-up for rate of change of
frequency measurcFient.
cathode ray oscilloscope face is masked to block off all
but the tips of the synchronizing pulses as illustrated
in Figure 10,
A camera with a motor drive is used to pho-
tograph the trace on 35 millimeter film.
V^hen the
film is
moved at a constant speed in a vertical direction, a line
is traced on the film by the top of each horizontal pulse.
The position of the line will vary in accordance with the
shift in phase of frequency of the pulse relative to the
output of the reference oscillator.
Vertical synchroni-
zing pulses interrupt the horizontal pulse traces, but provide convenient time marks.
The slope of each trace rela-
tive to the vertical on the oscillogram is inversely pro-
portional to the frequency difference between the submultiple of synchronizing pulses and the reference oscillatoro
(31)
The change in slope is proportional to the rate of change
S/6/VAL
Figure 10.
Masking of cathode ray tube face for rate
of change of frequency measurement.
of this frequency.
If the two slopes
during a period of time,
At,
S-j^
and S2 change
then, as shown in Figure 11,
the average rate of change of frequency over the period,
At,
is
^1 " ^2
.
At
7^Ar£
At
X__
I
c7/=-
C^AA/OS
-
S«C.
2^_
L-i^ia;-
Figure 11.
Graphical determination of the rate of
change of frequency of horizontal synch
pulses from oscillogram.
(32)
The block diagram of Figure 12 outlines a scheme which
has been devised for making an operational raeasurenent of
horizontal pulse frequency drift.
This system monitors the
horizontal drive signal rather than the synchronizing signal, because the former is a continuous and periodic wave
without extra components at the field frequency, and its
lock-in with synchronizing pulses is sufficiently tight that
its drift may be used to measure synch drift.
The horizon-
tal drive signal is heterodyned down to an average frequency
of 400 cycles per second and applied to a frequency meter
which gives a d-c voltage output proportional to synch freUsing this voltage to drive a cathode follower
quency.
provides a source of current v;hich is proportional to fre-
Differentiation of this current in a large inductor
quency.
gives
e
«
di/dt, which is thus proportional to df/dt, or
rate of change of frequency.
This output voltage can be
indicated on a meter, an electron eye tube, or other suitable indicating device after amplification.
In experiments
with this set-up, a high speed VU meter with the rectifier
removed was used as the indicating instrument, and while
the response was too slow for completely satisfactory meas-
urement, the set-up was ad.ludged sufficiently accurate for
practical measurements.
Calibration was obtained by taking
steady state reading of horizontal drive frequency from the
synchronizing generator for various values of d-c voltage
on the reactance tube grid, and plotting the resultant*
This grid potential was then varied at a constant known
(33)
I
V-
ra
(_)
r--
Ul
^'
J-
at
in
Q^
~>
—o
«/^
^O -^^
li.
CM
:
n
'-»
'..u
>
"3> a^'
O
<1
»r
a
T*"
•
I—»
.
<^::x
•
*
*1
J
<£
'P
iili
I
ft:
<
H^
4'
tu
A.
>-
o
^v
_l
fi
Q^
^
u.
a:iLi_c^
00
0,0
^
>
.1
/I-
<15
to
O^'
t
)
/
>kX
J.
>^
j
3'-^
"^
-q
'dzz
-H ->
CO
CD
(i)
<Jj
o
>^
'
'
<u
J
'^;
(34)
1^
rate, by means of a very lonp; time constant discharge circuit, over a linear portion of the frequency vs. grid-poten-
tial curve, and the meter indication noted.
In this way,
meter indications were found for various values of df/dt.
A third possible approach to the problem of rate of
change of frequency measurement would be to arbitrarily
shift the phase of the 60 cycle power wave used for locking the synchronizing generator to the power supply a cer-
tain definite amount, and then to measure the time necessary
for the synchronizing generator to reach stability.
If this
time interval is less than a certain miniraum, the tolerable
rate of change of frequency will be exceeded.
Such a pro-
cedure would give data on the follow-up characteristic of
the synchronizing generator and its lock-in circuit, but
would be useless as an operational measurement,
3»
Recommended Operational Method of Measurement of
Average Repetition Rate and Rate of Chanp^e of Frequency of Horizontal Synchronizing?: Pulses ,
The schematic diagram of Figure 13 is that of a cir-
cuit designed by the author to be used for operational meas-
urement of horizontal synch pulse frequency and rate of
change of frequency.
This circuit is used in conjunction
with a studio monitor scope as the indicating device.
The
input at Jl is the vertical drive pulse from the synchroni-
zing generator.
Stages including vacuum tubes VI through
V5 select alternate vertical drive pulses and use its stable
leading edge to generate a narrow adjustable width pulse.
The waveforms of Figure lU indicate the action of these
(35)
(S6)
V
t.^^s
t
»
1
\/lbV.
V.^5
Uv/;b5
r
L
r
n.
^.v..
J5:.i
'r -
•.Tb^:
J
f
*
tiTpe
^
Tirne
mter^ah
are
tft;.,^^^.5
not to scile but are as follows
YTiicio
seconds
,
defv='xleT>t or»
not
ti'tc-ZlOOO
5\/nc Gen. f^C^'jer^cy
drift,
ci-itical
Vt8,i500
Voltiie dni^htjcles
WivsTorm
FIS.
K
Ideal izeJ
iiot
<irc
identiticat,OT\
Wdveforms
is
of
to
scile.
as Tbllows-. Vlfjl; tube
1.
f^l-^'^e,
Frequency Measjrement Cu'cu\t.
(37)
|:»"n
Z,
stages.
V6 is a stable keyed oscillator tunable by means
of variable air condenser C12 from approximately 15,500 to
16,000 cycles per second. The oscillator is keyed off for
approximately
3
cycles and on with a fixed phase relation-
ship with the first horizontal pulse of each frame,
R20
varies the width of the keying pulse, and hence makes this
phase relationship adjustable,
V/ith SI in position 1,
the
oscillator output is amplified, clipped, and peaked in V7
and V8.
V12 provides a low impedance output to the mixer
stage VI3, and by means of R68 provides a pulse amplitude
control.
The stable pulse from the oscillator is mixed
with the composite picture signal in VI3 and the combined
output at J3 is fed into the studio monitor scope.
The resultant addition to the monitor picture is a
narrow black line, less than
1
picture element in width. If
the horizontal synchronizing pulse frequency, and the keyed
oscillator frequency are the same this line will be solid
and vertical, see Figure 15, line A-B.
Point A can be var-
ied across the top of the picture by adjusting the width of
the oscillator keying pulse.
If the synchronizing pulse
frequency is higher than the oscillator frequency the solid
vertical line A-B will degenerate into two dotted sloping
lines such as shown by the dashed lines A-C and D-E, Figure
15.
The slope is a function of the difference in frequency.
The lines are dotted because of interlaced scanning, i.e.,
A-C and D-E are generated by successive fields.
will be displaced to the right of point
C a
Point D
distance pro-
portional to the slope of A-C and the duration of the ver-
(38)
Figure 15.
Frequency measureirient indication
on face of monitor scope,
tical blanking period.
Lines A-F and G-H represent the
case of synchronizing pulse frequency lower than oscillator
frequencyo
By tuning the keyed oscillator for the solid
vertical line indication the frequency of horizontal syn-
chronizing pulses can be read from the calibrated tuning
dialo
Deformation of this vertical line is an indication
of frequency modulation in the synchronizing generator.
As
discussed in the preceding chapter, bowing of the line is
indicative of 30 cycle modulation.
Calibration of the os-
cillator is made by comparison with crystal controlled syn-
chronizing signals
The discussion of the preceding paragraph is applicable to those periods when the synchronizing generator is
oscillating at a constant frequency.
(39)
Normally, the syn-
chronizing generator output frequency will vary about the
nominal 15,750 c.p.s., maintaining stable oscillation at
any one frequency for only short intervals.
The line A-B
will appear to sv/ing to the right and left with a speed
dependent on the rate of change of the frequency.
In ad-
dition, the lines A-C, D-E, etc. will no longer be straight
lines, but will be slightly curved.
The maximum tolerable
curvature is illustrated in Figure 8.
A kinescope record-
ing of the monitor scope and subsequent analysis of the
line curvature permits accurate determination of the rate
of change of synchronizing pulse frequency over any desired
number of lines.
However, because of the time factor in-
volved, this type of measurement is not practicable as an
operating method.
Several limitations are involved in the use of this
circuit as an operational measurement of the rate of change
of horizontal synchronizing frequency.
The device will in-
dicate only tolerable or excessive df/dt of the synchronizing frequency as defined by the standards.
The measure-
ment can be made only over a period of 525 scanning lines,
or 1/30 second.
The assumption is made that the rate of
change of frequency is approximately constant.
This as-
sumption is deemed valid on the basis of experimental evidence, and consideration of the smoothing action of the syn-
chronizing generator lock-in circuit.
The line generated on the face of the monitor scope by
the pulses from the stable keyed oscillator may be expressed
(40)
by the equation,
X
where,
x
=
Xq =
= xo -h f ot -H |(df/cLt)t
deviation from vertical reference line,
initial deviation,
fo " initial frequency of synch pulses,
t
-
df/dt
=
time
rate of change of synch pulse frequency.
Assuming df/dt to be constant, it is seen that the path
generated is a parabola.
It is a simple matter to show that
the second difference of a second degree equation is a con-
stant.
Hence, under conditions of uniform rate of change
of synchronizing pulse frequency, point B of Figure 15 will
move to the right or left a constant amount each successive
frame.
If df/dt is made equal to the maximum allowable
rate of change of frequency, 23.6 cycles per second per second,
it can be shown that this distance on a 10 inch monitor
scope with 8 inch horizontal lines is equal to 0,2096 inches.
In time this corresponds to 1.15 microseconds, assuming a
line to take 55 microseconds.
The above discussion forms the basis for the rate of
change of frequency measurement.
The lines generated by
the stable oscillator are made 0,2096 inches wide and blanked
out except on the bottom two or three horizontal scanning
lines of alternate fields.
The presentation on the monitor
scope is then a short black horizontal line at the bottom
of the monitor picture which moves horizontally as the hor-
izontal synchronizing frequency changes.
(41)
So long as the
rate of chanpe of frequency is v/lthin the tolerable limits,
each successive line overlaps or meets the preceding one
as
it moves across the
bottom of the picture.
If the limit
is exceeded a light space is left void between successive
lines which can easily be noted by an observer.
Thus an
immediate indication is given that the limit of rate of
change of horizontal synchronizing frequency is being exceeded, and remedial action can be taken.
The measurement of rate of change of horizontal syn-
chronizing pulse frequency is made with Si in position
Figure 13.
2,
VIO provides a fixed delay from the oscillator
keying pulse, and Vll a gate pulse which enables the peaked
pulses from the stable oscillator to trigger V9,
The dura-
tion of the pulses from the one shot multivibrator, V9, is
adjusted to 1.15 microseconds.
These pulses are then clip-
ped and mixed with the composite signal in VI3 to be fed to
the monitor scope.
Calibration of the pulse length can be
made by a simple measurement of the pulse length on the
monitor screen.
It should be pointed out that the circuit of Figure 13
was not developed beyond the breadboard stage and many refinements are unquestionably possible.
The circuit did how-
ever serve to verify the feasibility of the measurement
methods last discussed.
There is no question of the ability
of the method to give accurate data on the instantaneous fre-
quency of synchronizing signals, or of the presence of frequency modulation in the synchronizing signal,
(42)
V/hile the
measurement of the rate of change of horizontal synchro-
nizing pulse frequency is not made within a period specified by the standards, the accuracy of the method seems
adequate to meet operational needs.
(43)
BIBLIOGRAPHY
!•
Engstrom, E. W,, and R, S. Holmes, Television
synchronization. Electronics, November 1938
2,
Fink, Donald G, Principles of Television engineering. New York, Mc-Graw-Hill, 19A.3.
3«
Institute of Radio Engineers, The, Standards on
television, methods of testing television
transmitters, 19^+7. Proc. I.R.E., 1947
(Special Edition)
4.
Kiver, Milton S. Television simplified.
D. VanNostrand, 1948.
5.
National Television Standards Committee, Edited
Donald G. Fink. Television standards and
practice. New York, McGraw-Hill, 1940.
6.
Radio Corporation of America, IB-36OO8, synchronizing generator Instruction book. R.C.A.,
1946.
7-
Radio Manufacturers Association Engineering
Department. Proposed electrical performance
standards for television studio facilities,
R.M.A., October, 1946.
8.
Roe, John H.
New York,
by-
The philosophy of our T.V. system.
Broadcast news. Nos. 53, 5A-, 55, February,
March, April, 1949
<.
9«
Synchronization of scophony
television receivers. Proc. I,R.E, 27:494-495,
August, 1939
V'ikkenhauser, P.
(44)
^>
DAOS DUB
--
f
1311
Vanderburg
^^unitv
Uty
Frequency stab
televistandards for
sion synchronizing
methods
generators and
of measurement.
13119
Thesis
Vanderburg
V16
Frequency stability
standards for television synchronizing
generators and methods
of measurement.
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