Oct - American Radio History

Oct - American Radio History

Domestic receiver review amateur solid transmitter

World

1968

Three

Shillings October www.americanradiohistory.com

NOW

5

Keyswitch

is

proud to announce five new miniature relays that cleanly sweep the field. British designed, British developed, and British made,

BRITISH

MIDGETS

MAKEof or

the KMK range features high assembly contact capacity, moulded for high insulation, phosphor bronze contact springs, 99-9% silver or silver cadmium oxide contacts,

Swedish iron magnetic circuit, international contact clearance

4mm, life

in

excess of

5

million operations, connection by solder push

-on type '110' connectors, open relays mounting

in

any

1968'

and

BREAKposition,

prices

are as

plug -in relays for international plugability. Unit low

as

9/4d

(1,000 rate), substantially

less

for larger quantities.

These versatile new midgets

are

BIGGEST

changeover contacts available with one, two or three rated up to 10A

at

250Vac /6Vdc, and with coils for

6- 230Vac /2-5VA and for

6-

200Vdc

1W.

Contact

RELAYS

NEWS

Keyswitch for complete price and technical details of these exciting new all- British KM

K's.

Keyswitch

Relays Ltd, Cricklewood

Lane,

London NW2; telephone:

01

-452 3344; telex: 262754.

KEYSWITCH

R

E

LAYS

ITCH RELAYS

-WHERE

THE ACTION

IS-

KEYSWITCH

RED

-AYS

WW

-001

FOR FURTHER DETAILS www.americanradiohistory.com

Wireless

World, October

1968

Fifty

-eighth year of publication

Wireless World

Electronics, Television, Radio,

Audio

Volume

74

Number

1396

October

1968

Contents

This month's cover montage includes one of the the

Murphy "coloured" television receivers,

Ultra 6702 colour receiver,

G.E.C. portable

G832 and below it the Pye

`Piccadilly', and on the left the

Ferguson

Futura II (upper) and G.E.C. G832 table receivers.

Iliffe

Technical Publications Ltd.,

Managing Director: Kenneth

Tett

Editorial Director: George H. Mansell

Advertisement Director: George 1owkes

Dorset House, Stamford Street, London,

SEi

337

The Poor Relation

338

Domestic Receiver Survey

346

Low

-power

Transistor Transmitter by G. R.

Kennedy

351

News of the Month

354

355

Personalities

Measuring Delays Accurately by L.

E. Weaver

356

357

Announcements

WESCON

Report

359

Conferences

360

Wireless

World

Colour Television Receiver

366 A

Segmented -Fit Squaring Circuit by

R. y. Lamden

368

369

370

Electronics and the Artist

Letters to the

Editor

Switched -Mode

D.C. Regulators by

T.

D. Towers

375

376

Schmitt Trigger Circuits by

G. E.

Marshman

Meetings

377

382

Test Your Knowledge questions and answers by L. Ibbotson

378

New

Products

World of

Amateur Radio

384

Literature Received

384

H. F.

Predictions ci

Iliffe Technical Publications Ltd.,

1968

Permission in writing from the

Editor must first be obtained before letterpress or illustrations are reproduced from this journal. Brief extracts Dr comments are allowed provided acknowledgement to the journal is given.

PUBLISHED MONTHLY (3rd Monday of preceding month).

Telephone: 01

-928 3333 (70 lines). Telegrams/Telex:

Wiworld Iliffepres 25137 London. Cables:

"Ethaworld, London, S.E.1." Annual

Subscriptions: Home;

;C2

15s

0 d.

Overseas;

1 year mail privileges

0

15s

Od.

Canada and U.S.A. ;

$6.75;

3 years

Q

Os

Od.

Canada and U.S.A.; $17.50 Second -Class authorised at

New York N.Y.

Subscribers are requested to notify a change of address four weeks in advance and to return wrapper bearing previous address. BRANCH OFFICES: BIRMINGHAM: 401, Lynton House,

Walsall Road, 22b. Telephone: 021

-356 4838. BRISTOL:

8

-10,

Corporation Street.

Telephone:

0203 25210. GLASGOW:

3

Clairmont Gardens, C.3. Telephone: 041

MANCHESTER: 260, Deansgate,

3.

Telephone: 061

11

Marsh Street,

1.

Telephone:

0272 21491/2. COVENTRY:

-834 4412. NEW YORK

-332 3792.

OFFICE U.S.A.: 300 East 42nd Street,

New

York 10017.

Telephone:

867 -3900. www.americanradiohistory.com

82

Square?

Not quite, but the only rectangular

3"

tube

on

the market...

The price is

special, too!

Another example of Thorn -AEI's renowned pro- duction engineering techniques

-

oscilloscope tube at

a

minimum price.

a

top

-quality

The Brimar

D7

-200GH

is

the only rectangular tube available in the

3

inch size. The tube has

a

relatively flat screen and employs

a

mono -accelerator for reduced power requirements. With an overall length of only

18 cm, it provides

a

5

cm

x

4 cm display of waveforms or TV pictures.

Thetube can also be supplied with

a

long persistence phosphor screen, making it particularly suitable for use with medical electronic equipment.

Features developed include electrostatic deflection and focusing. Good geometry

is

ensured by specially production control spot size and focus techniques.

Small uniformity over the entire screen give good resolution at all points of the useful screen area.

H

use igh- deflection sensitivities permit the of inexpensive transistor circuits.

Applications include: alpha -numerical readout devices, medical laboratory equipment, waveform monitors, data processing equipment, voltage and power output educational equipment, etc.

-

and

of

course, popularly- priced oscilloscopes.

Val

+a3 +a4

800

Vat

Vg

50 to 150

(for cut

-

off)

-20 to

-40

1200

V

75 to 225

V

-30 to

-60

V

For lull technical data and prices, write or phone:

THORN

7

Thorn

-AEI Radio Valves & Tubes

Soho Square, London, W1V 6DN Telephone

Limited

01-437

5233

WW

-121

FOR FURTHER DETAILS

Wireless

World, October

1968 www.americanradiohistory.com

Wireless

World

Editor -in- chief:

W.

T. COCKING,

F.1.E.E.

Editor:

H.

W.

BARNARD

Technical Editor:

T. E. IVALL

Assistant Editors:

B. S.

CRANK

J.

H. WEADEN

Editorial Assistant

J.

GREENBANK,

B.A.

Drawing Office:

H. J.

COOKE

Production:

D. R. BRAY

Advertisements:

G. BENTON ROWELL (Mànager)

J. R. EYTON-JONES

Wireless World,

October 1968

The

Poor Relation?

Almost we have devoted a section the latest developments in domestic receivers. Our

As is our custom we have again devoted several pages in domestic receivers, but in doing so we have had some serious had we as, on annually since the first British National that, for the of benefit of those should publish show we of our readers unable

Now, to

Exhibition excuse, this issue to a survey heart was held attend the London even not announced this in the last issue we would in all probability reflection, there seems little worth reporting in 1923 review

Show endeavoured, after discussing with designers and manufacturers their products, to give a show a digest one or more issues

Radio at the time of the latest techniques. Indeed,

"state of the art" survey. of course, there is the annual plethora of trade shows held in

London's

-at

of the if any of we when there was not a least show to was in place searchings, in

West a needed, of a

End hotels. have omitted it technical in was latest public of fact design matters.

What has happened to the British radio industry? It relation of its now affluent offspring industry's hey day the cream development sion of the country's radio engineers were engaged on the equipment design is a good deal lower

(e.g.

-the

than in the field of professional electronics radar, computers, communications).

Is this industry's unsatisfactory situation? Although each the cause or the effect of now that the calibre of the average engineer /technician appears to be the poor electronics industry. Whereas of broadcasting equipment and domestic receivers, now we get in the radio the impres- working on domestic of the the major electronics groups has a domestic receiver section, division or subsidiary company it more than a gap -filling production activity. is in each case little that, instead of British designs being copied overseas (as in

All too often we find the past), manufacturers in this country are content to follow overseas and even go so far as to unashamedly use foreign

-made chassis competitors in so- called British receivers. imports of

Perhaps their motto is

"if

you can't beat complete receivers and parts for the

'em, first seven

£7M (45% above the figure for the same period last year), join 'em "! months of

Incidentally this year were while exports were less than £3M.

We have found, on the whole, very little evidence thinking.

Is this because set makers have got used to products which they can then adapt or adopt?

One of real initiative waiting for the active device manufacturers to take the initiative and provide circuit designs based wonders when last on a in design their latest set designer told a device products devoting space to a survey imported sets interest to report. The only encouraging sign is significant manufacturer what he wanted in terms of new components to perform particular functions. Because device that were displayed by have been bold enough to of this attitude of taking manufacturers we find there is technically little, agents we if what of the latest techniques. In fact, had is it provided would have had little by of the various groups, hence our dilemma as to the justification the any, difference between the for not been for the of technical that one or two

British manufacturers start using integrated circuits in their sets. In passing it is that, whereas Japanese, manufacturers design

German, American and Scandinavian their products for world markets, those in this country seem to be content with the home market.

In spite of what we have said we hope readers will find the technical survey and the accompanying pictorial review of new receivers of interest, and of assistance in choosing new equipment or influencing the choice of others.

337 www.americanradiohistory.com

Domestic Receiver Survey

Developments in the recent television and sound trade shows in London receiver design,

as

seen at

Although there is no longer a general radio and television exhibition open to the public, most of the receiver manufacturers and suppliers get together in late August and hold simultaneous private trade shows in various hotels in

London, at which they introduce their latest models.

This provides a suitable occasion for

Wireless

World to take a general look at the technical design of domestic receivers. The following report, compiled by mem- bers of the editorial staff, is pictorial in approach, comprising mainly photographs of new sets and items of circuitry sketched on the spot, with commentary which aims to be descriptive but may also be to some extent explanatory.

Television

All of the British manufacturing groups were displaying colour television receivers with tube sizes varying from

48cm through

59cm to 64cm.

Apart from one or two notable tions the circuitry followed what has become a excep- conventional pattern. In the monochrome television receiver field, technical innovations were few.

There was an impression given through- out the whole of the exhibitions do for the present and reserved for the commencement that last year's design would that any development effort would be single- standard receiver of the u.h.f. single- standard era starting in 1969. Most manufacturers professed to having a

"ready" but only one working model was displayed.

Bang

& standard

This was a

64cm colour receiver introduced by

Olufsen that high price, claims are made for better picture quality and a high of.circuit will retail for about £400. In justifying the stability throughout.

A quick the service manual illustrates the performance look that

B through and

O

Fig.

1.

Simplified circuit diagram employed in the Bang of the e.h.t. generator

&

Olufsen single- standard colour tv receiver, Beovision 3000. hope to achieve with this receiver in which the service engineer is asked to align the receiver north/south and to inspect various portions of the screen with a microscope during purity adjust- ments.

In order to make the task of static convergence easier and to make the adjustment less susceptible to mains voltage changes permanent magnets are not employed.

A d.c. voltage, obtained by rectifying pulses from the line convergence circuits. The amount output stage the d.c. from interfering with the operation during period, is applied to the convergence coil unit. The are also used for dynamic convergence so necessary to that insert blocking capacitors in the circuit to prevent of the scan same coils it has been the dynamic of d.c. flowing is controlled by a potentiometer accessible from the front of the receiver, are all as the convergence adjustments. No e.h.t. over -wind is included on the line output transformer, instead a separate e.h.t. generator is employed, a skeleton circuit of which is shown in

Fig.

1.

Line pulses from the line output stage are amplified and shaped in one trolled rectifier. In operation a negative charge is developed across the 100-pF capacitor which supplies the bias for the PL509 which in turn half of the PCC85, which operates as a con- provides the high level pulses which are applied to the e.h.t. coil, the output being rectified in the

GY501 to provide the required

25kV.

Low -frequency power consumption variations are fed to the grid of the other half of the PCC85 via the two e.h.t. brightness controls whilst higher frequency variations are coupled to same grid via the capacitance of the e.h.t. which is in the region of

270pF.

If

output the mean screened cable current through picture -tube increases, the voltage at the grid the the will fall (this voltage is equal to the boost voltage minus the voltage drop across the e.h.t. max. brightness control) resulting in the PCC85 taking, less current. The charge across the

100

-pF capacitor will become less negative and more drive will be applied to the PL509 resulting in the e.h.t. voltage being restored. The receiver employs a total of

14 valves, of the PCC85

53 transistors and

52 diodes.

In the non -stop quest for optimum tube dimensions there was the inevitable change in shapes and sizes and this year saw the widespread use of a new 51cm tube, near a true rectangle in shape and with the correct aspect ratio aspect of 4:3, duction which was long overdue. It is something the industry had established the anomaly of a of receiving this company introduced sets with a number of "extras

" a

4:3 ratio picture on a

5:4 aspect ratio tube for so many years.

Greater simplicity has been the main design a logical trend intro- mystery that in the

Philips colour television receivers on show this year.

Last year

-an

"Autowhite" system to give automatically the familiar bluish

- white picture when the set was receiving monochrome trans- missions, a when

"colour

-off" button to assist the user to adjust brightness and contrast correctly, a tuning indicator to show the set was correctly tuned on colour transmissions, and a

338

Wireless World, October

1968 www.americanradiohistory.com

Fig. 2. An outline of the 7.8kHz switching wave generator employed in the

Baird colour television model

718. tone control. This year the same basic circuitry is substantially the but these extra facilities have been deleted in the interests of economy. A preview of new

Philips colour receivers due to come out in January revealed mask tubes with the that these will have 56cm shadow -

"squarer " -looking screen.

Also using the 56cm

"squarer" shadowmask tube latest colour television sets from the Rank

Bush are the

Murphy group, the Bush

CTV174D and fact, the receivers that

Murphy CV2210D. These are, in have a section of the colour decoder in of an integrated circuit, as reported in our August the form issue

(p.

263).

The integrated circuit, a

20 -lead flat pack, replaces two

65 discrete components, and provides the functions of synchronous detectors, two inverters, four matrices, and three output circuits giving respectively R,

G and B signals. As explained in the earlier report, the purpose of using this position is type of c.r.t. drive without the extra complication and an i.c. in not to save space or reduce cost but to achieve an improvement in picture quality by using the superior R,G,B cost that this normally brings when constructed from discrete components.

A block schematic of the integrated circuit was shown in the

August report.

Baird's 48cm colour television Model 718, departures from circuit convention. shows some

An R.C.A. integrated circuit is employed in the a.fc. section. A

39.5MHz signal from the last stage of the vision i.f. potential to the oscillator section amplifier supplies a control of the u.h.f. tuner to correct for tuning errors and drift

(a.f.c.).

Underside view of the

ITT

/KB colour receiver trating the proliferation of cable runs resulting from wired assembly.

The system switch chassis illus- a hand

- and operating solenoid are clearly visible.

Instead of using a bistable in the 7.8kHz switch generator, a sine wave is fed from a master oscillator through a single -stage tuned amplifier to a bottoming transistor, which delivers a squared waveform as shown in Fig.

2.

To achieve 180° line-by -line phase change of the

R

-Y

modulated signal, required by the PAL system, the phase de- of the

R

-Y

signal itself is switched instead of switching the reference generator phase to the

R colour TV receiver, model CK400. This is a 48cm dual- standard which is

-Y

demodulator

ITT/KB have just released details of their unique among present generation colour the chassis is virtually hand -wired stage. first version sets in that throughout, only a small printed panel being used to support the components located round the tube base connector. The circuitry is of hybrid design with transistors occupying all except for the v.h.f. tuner, of sound the signal output stage processing output stages. Both field and line timebases are valve- stages and chrominance operated and a solid

-state tripler provides e.h.t. One departure from accustomed hybrid practice is the use by of a transistor removing a wooden cabinet panel.

A customer

(type BD119) for the luminance output stage.

In common with most other colour receivers the CK400 has a monochrome capability decoder chassis is removed for servicing.

The usual number of colour setting -up controls are provided, accessible from the front

"tint" addition to the normal saturation control, allows if the control, in exercise a the user to preference for a particular colour balance by varying the differential gain of the colour- difference pre

-amplifiers.

The circuit incorporates an electronically operated device which automatically adjusts the reference white temperature from illuminant white of

"C" required for colour monochrome pictures.

A 15

-V reception to the bluish pulse derived from

- the l.o.p.t. is applied to

Tr8 negative -going which bottoms, generating positive and squared pulses at its emitter and collector res- pectively.

Suitable proportions

R, and

R8 to the emitters of these pulses are added of the R

-Y

and B

-Y pre via

-amplifiers

Tr9

-Y and B

-Y signals at their collectors.

When these modified signals after further amplifi- to change the bias on the tube grids by an amount equal to their amplitude

(see

Fig.

3).

When the colour -killer a and

Tr,

grids reverts and are to its thus added to monochrome transmission, normal are adjusted for the correct shade the

Tr9 value. of

R cation reach the c.r.t. grids and triode clamps, the latter now operate on the tips of the inserted pulses, so that their effect is and

If

Tr, the

, o

Tr

is colour- difference signals disappear, and the bias on cut -off are disabled, on the the c.r.t. c.r.t. anode voltages white (illuminant towards the usual bluish monochrome

"C

") with the pulses present on a colour transmission, the picture will move shade when the pulses are not present.

339

Wireless World, October

1968 www.americanradiohistory.com

.F.2oV f

+ Pu/se

/3-Y oulPue

/00

/S

V

Or/te

Tr

t

+

P/se oulpur

R-y

/Ap./C r,:,i

Pe,/s.

SAGA..

r

"

Co

/.u.. d:

P,reA p/.:Gi.re a;rr.r

Fig.

3.

Circuit diagram of

part

of reference white correction circuit. the

ITT

/KB colour decoder panel showing the principle of operation of the automatic

In apparent contradiction of the generally adopted policy of marking -time on dual- standard receiver designs and in spite of the near approach of the time when all television broadcasts will be radiated in the u.h.f. bands, G.E.C. have just introduced an integrated u.h.f. /v.h.f. television tuner unit which was developed at their Slough laboratories and in their own plant. is manufactured two parallel sets

This is a solid -state unit which is virtually of tuning circuits

(one set for Bands I and III; the other for

Bands IV and

V) operating in conjunction with common r.f. amplifier and mixer transistors and a common tuning gang capacitor.

A separate transistor for the v.h.f. oscillator making a total of stage is employed three transistors. sliding switch assembly located along the centre of

A the unit provides u.h.f. /v.h.f. changeover switching.

Six press- buttons provide mechanical selection of six channels in any combination of bands and systems. Each press- button incorporates a memory

- type fine tuner and claims are made for high reset accuracy, ensured by designing -in a large mechanical tolerance.

Sound receivers

(i)

Radio receivers were divided generally tuner

/amplifiers with separate loudspeaker units for stereo reception, or with an additional loudspeaker used in conjunction with the receiver integral loudspeaker. The latter type into two categories; portable battery -operated models and

(ii) described as of equipment was invariably

"hi

-fi

", which could be a misnomer in the absence of a standard reference for hi

-fi equipment such as the D.I.N. standard in Germany. are

Techniques from high- fidelity sound reproduction equipment gradually infiltrating into the field of ordinary production.

(We are now seeing, for example, receiver the big manufac- turing groups offering the public pseudo hi

-fi of separate units required waveform

Fig.4. voltage

Fig.5. of current as

-quite

large for and this would limit the an the power unit, the purpose stabilizer the power supply a radiogram a high

-so the maximum delivered on peaks and of the stabilizer is shown in

Another function performed by it is to provide a the form alternative to the radiogram.) so

An instance of this infiltration the latest in circuit design was noted in

British Radio Corporation stereo radiograms

(the

H.M.V.

"Stereomaster" models 2401 and

2402). The chassis includes quite an elaborate voltage stabilizer circuit following on signal peaks.

7 watts output instantaneous current would have current distortion. The circuit

-15V

will be of which delivered to is the to ensure that

These amplifiers are each capable demand can for the varicap be considerable. Without the diode full two audio amplifiers internal impedance tuning in of the giving cause constant system shown in

Tuning by varicap diodes has been established for some time in v.h.f. /f.m. tuners and in some of the more advanced imported receivers, but it is still rather unusual in the ordinary run of British -designed sound receivers. The tuner in H.M.V. models

2401 and 2402 has a five

-pushbutton station selector, and each of the five buttons incorporates a variable control for tuning.

As can be seen variable from the simplified circuit diagram in Fig.5, each tuning control is a

100k,ß potentiometer, which is used

G.E.C's new u.h.f.

/v.h.f. integrated tuner unit as shown in this photograph is more electronic one. of a screened box on the underside. mechanical endeavour than

The electronic components are contained in an a

340

-01

nim

Fig.

4.

Voltage stabilization circuit used in

H.M.V. stereo radiograms.

A 15

-volt source for the varicap tuner is provided from the junction of the 820

S2 resistor and the emitter of the

D1422 transistor.

Wireless World,

October

1968 www.americanradiohistory.com

croa v. h.

P. aerial

Dune F. rmer

6fo

/,aoo p o loo

44K

E

/,000p

6-8p

--

60p o.F.i: loop on -afc

3ß,o3

Voricap cantal vorn. o

-

/2 g.

/Ok

330 l,000p

IOk

---

sACom rousr. s.l.erar bufran.s

- p.

/ kn. LLn;m9 rem rion.a'Terc

2

4k

Lz

Pzp loo

K

0 22", o

r

L3Fl6o

5.6p ace

/motor

-lsv smcu.Jo

Fig.

5.

Section between of v.h.f.lf.m. tuner in

British Radio Corporation stereo radiograms, showing use the r.f. amplifier of varicap diodes and the mixer /oscillator. for tuning for applying a variable negative voltage (total range

0

-12V) through a pair of 10kQ resistors to the anodes of varicap diodes (which are thereby reverse- biased). two BB103

As a result the capacitance between the two electrodes in proportion to the control voltage. of the diode varies

Each varicap connected in series with fixed capacitors across a diode pre is

-set inductor

(L, and

L2), thereby forming a tuned circuit. The first tuned circuit tunes the output of the BF160 r.f. amplifier while the second tunes the following oscillator

/mixer stage

(not shown). diodes is

The tuning range available from the varicap limited but is sufficient for the v.h.f.

/f.m. broadcasting band.

A been

5W amplifier fabricated on a single monolithic chip has announced by

Sinclair Radionics.

Performance figures claimed for the unit are: a frequency response ±

1dB from 5Hz to 100kHz, a power gain of

110dB and

1% distortion at 1kHz.

The distortion figure is no worse than a number of other amplifiers that carry the high- fidelity label and the amplifier should find a number of useful applications particularly when the low price,

59s

6d, is considered. The amplifier employs a class AB,

World.

"totem pole

", output stage using two n -p -n conference was covered in the November 1967 issue transistors.

The integrated circuit is manufactured by Plessey and is very similar in design to the a.f. stage of an integrated experimental radio receiver described by M. J.

Gay and M.

C.

Sucker of

Plessey in a paper given at the Radio Receiver

Conference at the University College of

Swansea about a year ago.

This of

Wireless

The receiver described had a

40dB conversion gain, 90dB of a.g.c. control and 3W a.f. output. All components, with excep- tion of the coils, being on a single chip.

What happened to this prototype receiver, we wonder?

The Sinclair amplifier contains a pre -amplifier with a sensitivity of 5mV into 2.5MQ and requires that an external volume control and tone control circuitry be added.

A

The power supply voltage can be between

8 and

18V. pioneering step in portable radio design has been taken by

Roberts' Radio who use the Mullard TAD

100 integrated circuit in their new receiver, model

RIC.1. This i.c. is a directly- coupled wideband amplifier which is capable of being connectéd in a number design it employed to provide of configurations. In the performs the functions of the mixer /oscillator, detector and audio pre -amplifier stages including the above functions. The most of discrete passive components as well as the transistors normally

r.f

Roberts. circuits follow normal practice but i.f. coupling is carried i.f., the tuning out by a tuned crystal filter. Because large values of of the limitations of producing capacitance and restricted resistor values, the a.g.c. and decoupling components are wired externally. The whole is combined on a printed circuit board. Integrated circuit amplifiers have the advantage of showing only a small spread in characteristics so in manufacture reliable. of that with a few simple precautions taken the receiver, it should prove to be very

The RIC.1 covers the long- and medium -wave bands and costs 15gn.

Sony are using piezo-electric ceramic filters in one of their latest transistor receivers in place of conventional i.f. trans- of the circuit formers and tuned circuits. The relevant part is shown in Fig.6. The main advantage of this technique is, of that since the filters are fixed components with known course, response /frequency characteristics, the time -consuming manufacturing process of lining up the i.f. section is avoided.

The receiver is the TR1000, a

10- transistor superhet giving coverage of the medium- and short

-wave bands.

The f.m. tuner front -end design employed in Hacker's

"Sovereign" portable radio is also used in radio

SAR 1000, a combined stereo f.m. their stereo audio tuner -amplifier with optional a.m. waveband. The automatic mono-to- stereo radio switching system has a single signal, which would give a tuning point.

19kHz carrier signal cancels itself.

The

A poor performance, the automatic switching to occur. At the output low -level stereo does of tuner the not cause tuner front -end the is tuned by a triple- ganged variable capacitor.

An integral f.m.

/a.m. radio receiver allowing direct broadcast recording is a feature of the Tobisonic cassette tape recorder

TCR52.

During recording from microphone or radio monitor output is available.

The tape transport a system includes button on the top of the recorder fast

-forward winding. Pressing a causes the cassette to be ejected, thus simplifying cassette changing. The microphone is equipped with an on-off switch.

Also a mains power unit is built into the recorder.

Automatic switch -over to battery operation in the event of a failure of the a.c. mains power supply is a feature of the Inter- ceptor 1858, marketed by

Elsworthy Electronics.

One of the five wavebands covered by this portable is the amateur

2

-metre band between 142 and 150MHz.

Fig. 6.

Part of circuit of

Sony TR1000 a.m. transistor receiver showing use of ceramic filters in place of conventional band - pass tuned circuits.

/sr rF

fray

I.F

s*49f

41

Wireless World, October

1968 www.americanradiohistory.com

Telefunken

Bajazzo a.m. /f.m. de luxe transistor portable radio receiver with facilities for use as car radio.

Output 2.5W (port- able), 5W (car) driving 13 x

18cm loudspeaker.

Price:

55gn.

Decca 59cm monochrome TV receiver model

DR23 with v.h.f. and u.h.f. coverage. For- ward facing

13cm diameter loudspeaker.

Mains operation. Price:

84gn, stand extra.

Pye model

6000

`Piccadilly" multi -band transistor portable ra- dio receiver. Long, medium four short wave and

-bands plus v.h.f. coverage. Also extended

I.w. band and marine band. Built

-in aerials for all bands. Visual tuning indi- cator, af.c. and variable a.g.c.

Socket facilities provided. Chassis of stainless steel. Price: £100 plus.

Sobell 48cm monochrome TV receiver model

1032. Features u.h.f. Iv.h.f. transistor tuner incorporating six press- buttons, each with memory fine tuner. Price:

72gn.

RGD model RV235 48cm monochrome tele- vision receiver. S.T.C.

VC52 series dual

- standard chassis and transistor u.h.f. tuner.

Forward facing speaker. Price:

75gn.

Ferranti

48cm colour TV receiver model

CT1167 employing hybrid dual- standard chassis. forward facing

18 x

10cm loudspeaker.

Price:

Rotary

-type multiband tuner and

,299.

National model

SC140F four

-waveband radio receiver with stereo with record player and speakers: £102

18s. separate speaker units. Price complete

Unitra model DG204 radiogram from Poland covers five wavebands including v.h.f. lf.m. B.S.R. UÁ25 record changer.

Socket for tape recording. Price: £46

4s.

142 www.americanradiohistory.com

Wireless World, October 1968

Hacker model

RP25

"Sovereign

II" transistor portable radio receiver.

Features

16 transistors and

5 diodes, a.f.c. and quiet tun- ing on v.h.f.

Power output 1.5W;

20 x

13cm loudspeaker. Price:

04

14s.

7

Nivico model

AST140E a.m. /f.m. stereo radio receiver with unusual crossdial tuning indi- cator. Automatic stereo switching.

Total output

66W.

Separate bass and treble con- trols; frequency response

20Hz

Mains operation. Price:

129gn. to

20 kHz.

Bang

&

Olufsen model 3000 single

-standard

625 -line u.h.f. colour TV receiver with

63cm tube. Available in teak or rosewood finish with tambour dogs. Price. 398gn.

1

Fidelity RAD14 a.m. /f.m. three -waveband portable radio receiver with 18 x

10cm loudspeaker.

Wooden cabinet 30 x

23 x

7.5cm. Weight: 2kg.

Price: 17gn.

Philco -Ford model T993 portable radio receiver covering medium -wave and v.h.f. Ten transistors and eight diodes employed. v.h..f

Features

Price: 14gn. include a.f.c. on r

Teleton model

12

-T 203U portable TV receiver with 31 cm c.r.t.

Power source

220 1240V a.c. or 12V d.c. All v.h.f. and u.h.f. channels covered, 300

-e balanced input.

Eighteen transistors and ployed. Weight 7kg. four diodes em-

Price: about £79.

Decca 64cm colour TV receiver model

CTV25CE. Dual- standard hybrid chassis with press- button tuning. Two 18 x

10cm loudspeakers.

Price:

342gn or 345gn accord- ing to cabinet finish.

Luxor model

B4773 a.m. /f.m. transistor portable radio receiver featuring a.f.c. and pretuned press- buttons on f.m. Audio out- put

2W. Socket facilities include input for

Price:

41 gn. battery eliminator.

Standard model

SR

-K466F a.m.

/f.m. pocket radio receiver employing

11 transistors, and featuring a.f.c. on f.m. Output

200mW; speaker 4.5cm diameter.

Power source: three mercury bat

- eries.

Dimensions:

79 x

62 x

31mm.

Price: 23gn.

Wireless

World, October 1968 www.americanradiohistory.com

Ferguson model 3163 combines a

7- transistor radio receiver with a

Battery- operated electric clock.

Long- with and medium -wave coverage internal aerial. Audio out

- put 140mW; 7.5cm diameter

'

oudspeaker.

Price: L17 10s.

343

Roberts model RIC.1 long- medium -wave and portable radio receiver employs a

TAD100

Mullard linear integrated circuit in an filter advanced design. Crystal if. coupling also featured.

Audio- output in excess of 1W driving

18 x

9cm loudspeaker.

Price:

15gn.

111

1111111111 illl

I

IIIIII

I

1111111

Illillill

Ferguson 48cm monochrome TV receiver model 3652 incorporating

B.R.C. 1400 series chassis. Price: £75

10s.

H.M.V. 59cm monochrome TV receiver model 2648.

Incorporates

B.R.C. 1400 series dual -standard chassis embodying rotary v.h.f. tuner and press- button u.h.f. tuner.

Low wattage heater circuit. Price: £82

4s.

KB. model KR607 a.m. /f.m. tran- sistor portable radio receiver covering long, medium and v.h.f. wavebands. Audio output

400mW driving

13 x

8 cm loud- speaker. Features two position tone control. earpiece

Price complete with and wallet: 14Zgn.

.4tSE1

-`';ú

__

-_

-_`-_----

_---

Ï

Sinclair IC

-10 monolithic integrated circuit power- and pre

-amplifier in dual

-in

-line pack- age measuring 2.5 x

1 x

0.5cm. Output of

5W r.m.s. (continuous) is claimed.

Price:

£2

19s 6d.

Stella model ST4311 a.m. /f.m. mains table radio receiver covering the long

-, medium

- and short

-wave bands and v.h.f.

A valve chassis is employed. Price:

32gn.

----

G.E.C. 48cm monochrome TV receiver model 2038 designed specially for rental.

Continuously tunable u.h.f. tuner and ro- tary type v.h.f. tuner. Price

7Ogn.

Ekco model A410 transistor stereo radio receiver with built

-in stereo decoder and visual stereo indicator.

Long

-, medium- and short

-wave coverage plus v.h.f.

Two 20 x

13cm loudspeakers; output

4W

Price: £53

10s. per channel.

344

Grundig Nymphenberg stereo radiogram in period style.

Transistors:

26, diodes: 23. Waveband coverage; v.h.f., s.w., m.w. and l.w.

Output power:

10W per channel. Stereo decoder fitted. Record changer: four

-speed unit with manu- al or automatic record selection.

Mains operation. Price:

437gn. www.americanradiohistory.com

Wireless World, October

1968

Pye model 63 monochrome TV receiver with

51cm

"squared" tube.

Incorporates the 368 series hybrid chassis

13cm loudspeaker. featuring

Price:

£83. solenoid -oper- ated system switch.

Forward facing 20 x

KB model CK 400 colour TV receiver em- ploying 48cm type tube.

Hybrid chassis featuring manual tint control. Removable decoder board. Solenoid- operated system switch. Solid state e.h.t. unit. Price: 275gn.

Sony 23cm battery /mains portable

TV receiver Model TV9 -90ÚB covers all v.h.f. and u.h.f. channels.

Thirty transistors and fifteen diodes employed. Weight

S.5kg.

Price: £79.

Philips model 503 colour TV receiver incor- porating a 56cm

"square" tube mounted in a

"push- through" presentation.

Available next January.

Baird 51cm

"squared" c.r.t. television receiver model 15.

Price:

72gn.

Bush model CTV174D colour TV receiver using 56cm tube with 4:3 aspect ratio. Also features integrated circuit colour decoder and R,G,B drive to c.r.t. Price: 299gn. The

Murphy CV2210D is similar.

Dynatron model

HFC4 "Geneva" transistor ster e radiogram with decoder fitted.

Long

- and medium -wave axd v.h.f. coverage,

20W power output. Bass reflex speaker enclo- sures.

Record changer with diamond stylus.

Price: L111 8s

8d plus stand.

Marconiphone

59cm monochrome

TV receiver model 4626 incorporating B.R.C.

1400 series chassis.

Price: £114 10s.

Wireless World, October

1968 www.americanradiohistory.com

Ultra

48cm monochrome TV receiver model

6656 sis. incorporating

B.R.C. 1400 series chas-

Price: £73

Iss.

345

Low

-power

Transistor

Transmitter

A

7

-MHz

3

-W

solid

-state design of very

small size by G. R.

Kennedy

The basic design requirement was for a small transmitter which could be carried in a large pocket or an ordinary brief run for a reasonable period off dry batteries, yet with enough power output to ensure at least a reasonable chance of a contact and which used readily available transistors not costing more than ios each.

These requirements were met by the basic circuit shown in Fig.

For an r.f. amplifier stage the fT of a transistor should i. be at least twice the signal frequency, and for a stable oscillator it should be several times the oscillation frequency. For the

BFY5

t,

fT(mtn) is

5o

MHz, which is seven times the signal frequency.

The voltage rating of the transistors in a transmitter should be at least

6o V three times the supply voltage:

VcE(max) for the BFY5i is and the supply voltage used in the transmitter is 18V.

The power dissipation of transistors is linked to the maximum and working at

25

°C junction temperatures. For a

BFY5r,

PToT(maz) is

5

W and 2.8W at roo

°C, the temperature being that of the case.

The oscillator and driver are run at lesss than t W, and the p.a. runs at a maximum of r

W per transistor, with the transistors in a heat sink. The

BFY5

I is, therefore, a suitable choice for the transmitter.

Circuit description

The first BFY5 r stage is a negative impedance crystal oscillator.

In this type of circuit when the emitter bias resistor is bypassed by a capacitor of a certain value and when the collector circuit is on the inductive side of resonance, a negative impedance appears between the base and the common line.

A quartz crystal is placed between the base and the common line, so that the circuit oscillates at the crystal parallel frequency. The tuned circuit in the collector must be tuned to slightly higher than the resonant frequency of the crystal to present an inductive impedance. The d.c. bias conditions are set by Rt and

R2 in a potential divider. The temperature com- pensation is not as good as would be devised for a straight class A amplifier stage since

R2 shunts the crystal, and the low value required for good temperature compensation would have damped the circuit considerably. However, more than adequate stabilisation against thermal run

-away is provided over the normally expected ambient temperature range. To further help short term stability, the first transistor is placed in a heat sink.

The supply voltage of i8

V is dropped via

R3 to

12

V at the collector of the first transistor.

The filter network,

L8 C4, isolates the oscillator supply rail as regards r.f. The oscillator output is taken via the step -down transformer, Lt

L2, which matches the oscillator collector impedance into the base impedance of the next stage.

The buffer transistor,

Tr2 a

BFY5r, operates in class B, with a tuned collector load adjusted to resonate at the crystal frequency.

Negative feedback is applied by means of the un- bypassed emitter resistor

R4.

The transistor base input is transformer coupled from the oscillator stage, with the cold end by a key jack. of drive winding interrupted

This provides a suitable method of d.c. keying the transmitter, since the buffer transistor is silicon, with an extremely small leakage current when the base is open circuited.

The efficiency of the keying is also increased by the negative feedback, since this lowers the operating point of the transistor. Keying the buffer stage means, of course, that the oscillator is free- running whilst the transmitter is switched on, and oscillator frequency shift

Fig. r

The circuit diagram of the transmitter.

The R3,

R3' combination is used to obtain the required 600

S2.

R3

1.2k

1/2W

Le

C91

01µ-

C10

50

TSVµWk.g.

©M1

C3 ceramic

R1

47k

1/2W

250pß sitveT mica

L1

6

Tri

BFY51

C4

0.1

µT

R3

,A

1.2k

1/2W

'

L2

CL

350p snyeT mica

L

3

',

Q

Tr2

BFY51

-1 i

,

L

4

L5 o+18V

,See

text

.-

1

--Co

Output

Silver mica

11

-92p

(82+10)

O5µ polyester

R2

4.7k

1/2W

7MHz

Crystat

R5 C2

C5

100

-.470

P

1/2Wceramic

0.1

µ

P

L

7

IKey jack

R4

18

6W

C v500p twin air gang spaced

Wireless World, October

1968

346 www.americanradiohistory.com

with keying

(

"chirp ") is obviated.

The filter,

L,

C5, prevents r.f. from appearing on the key leads.

The p.a. stage consists of three BFY5 is operating in parallel single

-ended class B.

The input is transformer coupled and matched by the transformer

L3 L4.

The output is taken via a

7r coupler network

C,

L5 C8 to the output capacitor

C12.

The com- bined collector currents of

Tr3 Tr4 and Tr, are monitored on the meter M1, and r.f. isolation is given by the r.f. choke

L6.

The common emitter resistor

R6 is of a low value but for tuning up purposes, a higher resistor, collector

R, is inserted at this point to limit the current and increase the negative feedback.

No thermal compensation has been provided in the p.a. stage, and consequently the three transistors

'are mounted on a common copper heat sink.

Since rail to the

BFY5 t has the collector connected to the TO5 can, the heat sink has to be isolated from the transmitter case.

The

18

V the p.a. and buffer amplifier is decoupled via C9 and

C10, to prevent instability.

A view of the transmitter showing the front panel.

Oscillator stage

A compromise has to be found in the oscillator circuit between high stability with no self -heating of the components, and sufficient output to drive the following stage. Since a power transistor is being used in both the oscillator and buffer, the oscillator output and buffer drive are more than with small signal transistors.

Hence, let the input power to the oscillator transistor be 0.25

W: then dc 13 mA.

The available voltage swing at the collector, taking into account the voltage drop across the emitter resistor, the

-

1.3

-

7.8

-

0.4

=

8.5 V series dropper and the r.f. choke is

18 r.m.s.

Assume two turns for

L2

(see later), then the number

=

(n2Vc)VIN. of turns required at the collector is n1

Now VIN will vary with the gain of the transistor, the collector load and the loop gain. However, an r.m.s. voltage of less than i V r.m.s. is advisable across the quartz crystal for stability. So taking

VIN as o85

V r.m.s., nl

2 x

8.5

-

20 turns

0.85

L1 is then 20 turns of

20 s.w.g. enamelled copper wire close wound on a o635 cm diameter former with an iron dust core.

Reference to standard wire tables gives an inducatance of

2µH

L1

=

2

µH and for resonance C

= w2L where L is in henrys, C is in farads, f is in Hz and w is

27rf.

Coil winding details

The parallel tuning capacitor for

7

MHz is

:

Cl

1012

4772

72 x

1012 x

2 x

10-e

-

257 pF

Buffer stage

Input inductor

L2:

The measured drive was 29 mW at

2.5

V r.m.s.

The average input impedance RIN

=V

IN2IPIN.

RIN-

2.52 x

1000

-21652

Let the inductor working

Q be to for good harmonic suppression, then

L2

=

RINIWQW

.'.RIN

=

216 x

108x

14V x

10

106

Referring to standard copper wire tables gives

L2 as two turns

20 s.w.g. enamelled copper wire. of

This is wound over the "cold" end of

L1 and the whole assembly is mounted inside a screening can.

The value of the collector inductor

L3 will depend on the p.a. drive inductor

L4.

Assuming for the moment that this is two turns, then n

3 n4.

1,",

VIN

2 x

16

2.5

-

13 turns.

L9 is, therefore,

13 turns of

20 s.w.g. enamelled copper wire close wound on an o635 cm diameter former with an iron dust core giving an inductance of

1°5 pH.

Ce

47r2

1012

X

49 X

1012 x

1.5 x

10-8

-

345 pF. coil

L1

L2

L3

L4

L5

L9

L,

L8 value turns wire remarks

2µH

0'5µH

1

5µH

20

2

13

20 s.w.g.

20 s.w.g.

20 s.w.g. close wound, o635-cm former on cold end of

L1 close wound, o635 -cm former on cold end of

L3 0.50H

See text

20µH

2 20 s.w.g.

. too 24 s.w.g. close wound, former

2mH1

2mH

J

Pi -wound r.f. chokes

t27

-cm

P.A.

stage

For the drive inductor

L4.

The input impedance transistors at higher emitter currents is approximately equal to the extrinsic base resistance rble

(also known as of the output the base spreading resistance of the hybrid-77 equivalent circuit). Here rbb' 120

S2.

Then

Rol

130=4052

Now the

Q of the p.a. drive inductor need not be high, since the

347

Wireless World, October 1968 www.americanradiohistory.com

collector circuit Q will be high. Let the working

Q of the inductor

(

-Ve)

or inductive

(

+Ve) reactances. For a

5o

S2 output. be 2, then

RL

Ro

=

66 x

50

=

57.5

S2

RIN

40 x

106 then

2

7 r X

106x2

-05µH

Ca

1

1012

- caXa

147r x

108 x

57.5

400 pF

-

Cb

L4 is 2 turns of

20 s.w.g. enamelled copper wire wound over the cold end of

L37 the whole enclosed in a screening can.

For the output inductor

L5.

The design of the p.a. rr network

Lc

X.

57.5 x

108

= w

=

147r x

106

=

1

3µH

will depend on the stage following the p.a.

If

a high

Q conventional aerial tuning unit is used, then a single low

Q it network can be used. The output

7r network for feeding a

5o

S2 input high -Q aerial

If

it wished to couple the p.a. directly into an aerial, then the

Q of tuning unit is shown in

Fig.

3.

The

Q is the

7r network will have to be raised.

The simplest method of doing this is to add two

Tr networks in series.

N/RL

/66

Ro

N50=1.15

Single

- network

o

RL o

I c c

Ro

This value is too low for the network to adequately suppress harmonics. In practice higher value variable capacitors may be used at C, and

C8.

This enables different aerial tuning unit loads to be used, and also allows experiment in different output network configurations to be carried out for aerial. optimium results with a given

If

the following circuit is to be used, however, the minimum capacity of each variable capacitor should be less than

40

-pF.

For the p.a. stage

RL vee

2 ti

2POUT

-

RE where

RE is

Vcc is the supply voltage, the emitter resistor.

POUT is the r.f. output power and

Assuming an 8o% transfer efficiency POUT

=

2.4 W

182

RL=2x24

1=6652

now

Za

=Ra

RL

+J

Ro cos

ß

Ro sin ß

-

1/RL Ro where

Ra is the pure resistive component fo

Za, and ß shift

(+

or the phase

-) between input and output of the network.

Assume the components to be ideal (i.e. no Ra term), and let the phase shift ß be 90

°.

Then

RL. Ro

1/RL.

Ro

-

1/RL Ro

Similarly

Zb

=Rb

+J

RL

RL

Ro sin

ß cos

ß

-

Ro and making the same assumptions

IZbI

RL

Ro

=-

1/RL Ro

-

1/RL Ro

Also

Zc

=

Re

+ j

1/RL Ro sin

ß

So

-Xa =

Zc

I

= +

1/RL Ro

-

Xb

= +

Xc, where the sign indicates capacitive

RL=66fZ o

348

400p

L5=1-3µH o

Double

77

network

o

;

Z,

RL=66C2 o

Za

T

Zb

RL=5kC21

Zd

T T

Zf

1

Ze

T

o

R0=500

From the formulae given and from the assumptions made in the previous section:

I

Za l=

-

1/RL

RL'

=

I

Zb

I

I

I

Zc

I

Zd

=

+

'VRL

I=

-

Ro

=

I

Ze l

I

Zrl

=

Ro

Then

-Xa=-Xb=+Xc

=

1/66 x 5,000

=

575

O.

Fig.

3 Single

77- network circuit.

1

Cd= Ce=

u,Xd

147r

1012 x

106 x

500

-

46 pF

Lr w

Xi

500

-

14rr

=

11

µH

The circuit of the double

7r network is given in Fig.

5.

The value of

Wireless World,

October

1968 www.americanradiohistory.com

1

Ca=

Cb=WA a

1012

-

147r x

106 x

575

40

PF

Xc

575 x

106

Lc cu

147r x

106

-Xd=-

Xe=

+Xt

-

13

µH

=

1/5,000

X

50

=

500

S2

L5 o6b6b0606b066öò

13µH 11µH

C11

R=66S2

C7

40p o

I92p

Fig.

5

Double

Tr- network circuit.

C6

R0=5of2

TbP

o

RL',

5,000

S2 was chosen to give the required Q

Q

RL

Ro

5,000

=

10

50

A Q of to is a good compromise between efficient harmonic suppression and inductor resistive loss.

The coil winding details for the single

7r network,

L5

=

P3 µH, are

6 turns of i8 s.w.g. enamelled copper wire close wound on an insulating tube of x.5 cm diameter. Alternatively

(for different loading values),

54 turns as above tapped at

6, io,

14 turns etc.

For the double

7r net- work,

L5

=

13

+

II

µH, ioo turns i8 s.w.g. enamelled copper wire close wound on an insulating tube of

1.5 cm diameter, tapped at

54 turns.

Thermal design of

p.a.

heatsink

The thermal resistance of the p.a. heatsink is given by

Bh

=

Tj

-

Tamb

Pc

Uni

-

Bi

Output

5 6 7

8 9 10

11

12 13 14 15

ó? o"

1

O o

O

O

O o o

O

0 o o

IIC y

Q o o o

O

(

f

00

O

--110

0 0. A o O

O

O

0 o o

0 0

0

0 0 0

0 0 0

0

0 0 0

0

0 0

Il

OO

O

0

0 o

0%r

O

O

O'

O

0

0

0

O O

O

ó

O

O

0

O o

ö

2

3

4

6

7 a

9

O

O

0 0 0 0

0

0 0

0 0

O

O o

00

O. o o o

O

O o o

O

/

O 0 O

O

O

O tY

0

0 0 0

0

0 0 o t

O

00

00 o o

O

O q o o

:

O

IIC4

O 0

O

O

O

0

0

0

0

O

O O O

0 0 0 0 0

o

ó o

O

0

0. o 0

O

O

COC'

O

0

0 o

0

O

O

0 0

0 0 0 0 0 0

O

0

0 0

0

.

0

0 o

O

0

O

O

O

O

O

O

0

O o' o o o

O o o

O O

O

OJ o cflo

0

00

O

O

0

0 0 0 0 ö

O

O

O o0

O O O O o

0

0 0

-

C11-

0 0 0 0 0

0 0

0 o o o o o o

O c.: a

ö:

.g::

ó

0

16

17

18

19

20

21

22

23

24

25

26

27

28

29

14

15 o

13

10 o o

11

12

0.10in pitch

Veroboard

3in

A

Fig.

6

Layout of components in the case.

Output to

Tr3,Tr4

&

Try

9 8

7

6

5?

3

2

Tr2

Heat -sink

Crystal hótder

L7 t-,6bo

Key jack

15 14 13 12

11

1

R4

,VsAr

-II-40

C5

L1

OO

L3

-1 h--

L4

Crystal

! `i

R

22I

R5

17

-

L8 fÓÚÓ

-

19

18

20

13

14

15

16

6

7

8

9

10

11

12

2

3

4

Buffer

L2

/

26

24

25 b

27 e c

28

29

21

22

Oscillator

23

3in

I

Heat-slnk

1

5in

Wireless World, October 1968

1.5in

Fig.

7

Underside of the Vero circuit board. where

OA

=

thermal resistance of the heat sink (heat sink

/ambient)

Bi

0m

= thermal resistance between transistor case and heat sink

= thermal resistance between collector junction and transistor case

Ti

= collector junction temperature

Tamb

= ambient temperature

PC

= total power dissipated at the collector junction

For the BFYSI,

Ti(max)

=

200

°C;

9m

35

°C

/W;

Pc(maz)

=

1

W

The highest ambient likely to be met is 45 °C.

If

the transistor is mounted in a copper block, with silicon grease at the interface, then

01 may be taken as

o2

°C /W

200

-

45

9h

35

-

02

=

120 °C

/W

1

The thermal resistance between the transistor case and air is approximately

220 °C /W. A heat sink is copper block therefore required.

A of minimal dimensions to accomodate three TO5 transistor cases should be adequate.

Construction details

The original transmitter was built into a small aluminium box

20

X to

X

6 cm made from a sheet of i8 s.w.g. aluminium bent

349 www.americanradiohistory.com

Less e.m.f.

for

the

Volt

This photograph can be related to the practical layout drawings to give a clean picture of the layout employed. into a squared off

"U ", with a hooded lid of the same metal fitting over the top. a

This prototype contained a single p.a. it network, but second model was rearranged to accomodate a double i network.

No special care was taken in the layout, apart from the usual precautions of using rigid wiring, and making sure that unwanted coupling and consequent feedback did not occur.

The internal layout of the first prototype is shown in Fig.

6.

The oscillator and buffer were constructed on a sub

-unit of

0.25 cm pitch Veroboard

3 x

6 cm in size (Figs.

6 and

7).

This was mounted on two pillars attached to the base plate.

The heat -sink for the three p.a. transistors is made out of solid copper. The dimensions are

4 x

1.6 x

1.6 cm.

The block is drilled at three places using a drill suitable for the

BFY5 i TO5 case

(f in, 8mm or

Letter

O), and the transistors are inserted into the holes after being smeared with silicon grease to improve heat transfer. The heat -sink is then drilled end tapped

6BA in four places on the reverse side and retained in the case by nylon screws and insulators. The other transistors in the circuit are fitted with standard round heat -sinks.

In the prototype the single p.a. ir network inductor

L5 was of

54 turns, tapped along its length to aid the original setting -up. It might well be more useful to use such a winding, instead of the calculated

6

-turn inductor, to help in matching into other loads.

Modulation of the transmitter has not been tried

-the

reader is referred to Larsen (R.S.G.B.

Bulletin, Oct.

1966, p. 639) for information and references.

Other frequency bands may be tried. Doubling in the buffer stage can be accomplished, with reduced drive to the p.a. and reduced p.a. output. The buffer inductor values and it network component values will have to be re- calculated using given. the method

The efficiency of the p.a. in these calculations should be taken as about

50 %. Other transistors could be tried in the p.a. or the p.a. could drive a further stage. Cheaper high power r.f. transistors might give interesting results.

The transistor transmitter that has been described is càpable of giving a stable and clean c.w. signal. Reports received on the air have been most encouraging. It has been found to be very economic on power requirements, and the batteries specified seem quite adequate. Construction of the basic circuit should give those new to semiconductor transmitters a footing from which to try further developments of the original idea.

It is planned to show the transmitter at this years Radio Society of Great

Britain Exhibition to be held at the

Royal

Horticultural

Society's New Hall in Westminister between October

2 to

5.

Other equipments to be shown are the r.f. signal generator by

Nelson Jones,

P. J.

Baxandall's Loudspeaker and the Wireless

World Computer and Crosshatch Generator.

350 www.americanradiohistory.com

In the near future all readings on voltmeters in

Britain are likely to be slightly low. An e.m.f. indicated as

1 volt will will in fact be slightly be higher than

1 true volt. The reason that the standard volt in

Britain is probably going to be changed, in such a direction that in future it will represent a smaller electromotive force than it does at present. In October the International

Committee on

Weights

Sèvres, and Measures (B.I.P.M.) will be meeting at

France, and is expected to make recommendations for bringing the differing volt units of ten countries into agreement. But engineers and technicians need not worry too much, for the probable change in each case is likely to be more than a few parts in a million

-in

not

Britain about

10 p.p.m. Nevertheless it is the increasing impor- tance of precision electrical measurements in engineering, to which electronics has greatly contributed, that has made necessary this more accurate definition of the standard volt.

A significant factor in this need to regularize the volt. is the general adoption of the

SI system of units, in which electrical units are defined in terms of the units for mass, length and time.

For example, the volt is defined first in terms of the ampere and the watt. defined in terms

The ampere is in turn of force and length

(force produced between wires carrying current), while the watt is defined in terms of the joule and thence also derives from force and length. Force, of course, is defined in terms of mass, length and time.

Unfortunately the absolute volt de- termined in this way is different from the volt as given by the standard cells kept by national standards laboratories, so the purpose of the October B.I.P.M. meeting will be to bring the units maintained by the various countries into conformity with the absolute

SI volt.

New measurements of the ohm and the ampere in terms of the basic

SI units have been made in various national laboratories. The ohm determinations show that the values of the standard resistors used to maintain the unit of resistance are substantially correct (to better than

1 p.p.m.). On the other hand the recent ampere de- terminations indicate that the absolute ampere is smaller by about

11 p.p.m. than the ampere determined by the standard cells and standard resistors used to maintain the standards for the volt and ohm, respectively, at

B.I.P.M.

Since these measurements indicate that no adjustment in the ohm is called for, an adjustment in the base of reference for the volt is needed to account for the results obtained in the ampere determinations.

Fortunately, it will be a long time before the average voltmeter used by electronics people is accurate enough to be affected by these minute adjustments to the standard.

Wireless World, October

1968

News of the Month

the remaining one being used for control purposes.

The normal message length is three seconds; this can be extended sired, and the message is if de- continuously repeated while the fault exists or until cancelled by the pilot.

Of the

15 messages which can be accommodated six are considered to be priority warnings and would be restricted to conditions affect- ing the immediate safety of the aircraft.

If a priority warning is called for during the replay of a non

- priority warning the non- priority warning is interrupted. In the event of two warnings occurring simultaneously they are repeated sequential- ly.

High

Power m.o.s.ts.

One has become accustomed to thinking of the m.o.s. field -effect device as a small -signal amplifier often used in input stages.

It is therefore a surprise to learn from Associated

Semiconductor

Manufacturers Ltd., of

Wembley, London, that they have a research programme aimed at developing high -power m.o.s.ts for use as linear amplifiers in single

- sideband h.f. transmitters (operating from

2 to

30

MHz). Next year, in fact, they hope to produce a device giving an output of more than

100 watts peak envelope power. Mean- while, they have actually developed an ex- perimental m.o.s.t. with an output of

30 watts p.e.p., will and as we go to press this device probably be on show at the

International

Broadcasting Convention in

London.

A feature which makes the m.o.s.t. parti- cularly suitable for single -sideband amplifi- cation is that its output current is propor- tional to the square that the level to of of odd the input products is low. In contrast to the bipolar prevent thermal runaway.

Also, voltage, of so

-order intermodulation transistor, the m.o.s. device is thermally stable and requires no complex bias circuit course, the m.o.s.t. has the advantage of a higher input impedance than that of a bipolar transistor and an exceedingly low gate cur- rent.

A.S.M. state that two factors in the development of the high -power amplifier are raising the breakdown voltage to increase the output and arranging the construction to minimize feedback capacitance. They claim that with the

30

-watt device a stable perfor- mance with little distortion is maintained over a wide frequency range.

U.K., the C.E.E. was likely to choose black for the live cord. done a great deal of

If this had been confusion between the present black for neutral and the new standard could have occurred in this country.

Air borne audible warning system

Monitoring all the equipments contained in a modern aircraft is a time-consuming busi- ness which adds to the ing system whereby a crew work -load. Most aircraft contain some form of central warn- failure in an essential service or the onset of a dangerous condition actuates some sort pilot on seeing secondary set of primary indicator. The the warning then scans a of warning indications to find out which system has failed so necessary action can be taken. cleared all

If that the the fault warnings are also cleared, but the fault persists the primary warning cation can be cancelled so that the prepared for any other failure is indi- system that if is may occur.

As a logical development of the above technique Ferranti have produced an au- dible warning system, called type FAW.3, that plays back a recorded message, via headphones or a loudspeaker, giving details that has occurred or any of any failure dangerous condition that has developed.

The tape section of the device employs a

16 track head, fifteen of the tracks are for messages,

Goodbye red, black and green in

1970?

New regulations made under the Consumer

Protection Act 1961 require coding of the power cable of that the colour domestic electri- cal appliances offered for sale in

Great Bri- tain complies with the code recently recom- mended by the International Commission on

Rules for the Approval of Electrical Equip- ment (C.E.E.).

The new coding is as follows:

LIVE

-BROWN

-LIGHT

BLUE

EARTH

NEUTRAL

-GREEN

AND YELLOW

STRIPED.

Up to a date to be specified in the regula-

tions-

January

1,

1970, has been proposed

-the

present British standard code will also be permitted.

At one time during the discussions, but for the intervention of the

Wireless World, October

1968

A technician using the microspot analyser an- nounced by

Ferranti.

The instrument which is for c.r.t. resolution measurement can be used in any one of four modes, spatial fre- quency, two width, width slit line half power line and intensity distribution across spot.

In addition the instrument can be used to measure deflection linearity, phosphor noise, up phosphor build- and persistence.

Space research spin -off aids heart patients

An instrumentation system developed by the

National Aeronautical

&

Space Administra- tion Flight Research

Center, Edwards,

Cali- fornia, for use by pilots of research aircraft, is being used as an emergency ambulance aid.

The system enables an ambulance atten- dant to transmit to a hospital the electro- cardiograms of a patient whilst en route to that hospital.

The hospital being forewarned of the patient's coronary condition

-in

de- tail

-can

make all the necessary preparations in advance. In the system slim bare wires replace ordinary clinical electrodes. These wires are applied to need the skin using a quick drying conductive silver -glue eliminating the for shaving the skin.

Information ob- tained in this way is relayed over a con- ventional radio link.

The system is at present being used in

California and Los Angeles.

Nigeria communications contract for GEC

-AEI

The Nigerian Ministry of Communications and GEC -AEI

Telecommunications have recently signed a contract in Lagos worth

£9.6m. Under the contract GEC

-AEI will supply and install a

2,500 -mile microwave and v.h.f. network which will complete stage three of Nigeria's national telecommunica- tions plan. Intelligence will be transmitted at

450MHz and

2 and 6GHz using f.d.m. while supervisory information will employ t.d.m.

351 www.americanradiohistory.com

techniques.

The it possible to supervisory channels make monitor remote stations and to display any fault occurring at such a station at one of the maintenance control centres. In some cases where the traffic is expected to be heavy auxiliary narrow -band radio links will carry the supervisory information.

A total of

17 maintenance control centres will be estab- lished to service the network; each will carry trained staff, spares and the necessary test equipment.

VASCA officers

At a recent meeting of the General Manage- ment Committee of the Electronic

Valve

&

Semiconductor

Manufacturers' Association,

Dr.

F. E.

Jones was elected chairman of the

Association in succession to

P.

H. Spagnolet- ti, who has held that office for the past two years.

V. A.

Cheeseman (managing director of the M-0

Valve Company) was elected vice -chairman.

G. P.

Thwaites has been elect- ed chairman of the group concerned with industrial valves and tubes and J.

C. Aker

- man was re- elected the chairman of semicon- ductor devices group.

The

British Radio

Valve

Manufacturers'

Association and

V.A.S.C.A. also has changed announce that the Post

Office the address of their headquarters at Mappin

House, Oxford

Street, as follows:

Mappin House,

4

Winsley

Street, Oxford

Street, London. WIN

ODT.

Wired standard frequencies

Many laboratories may wish to follow the example

Bureau of the United States

National of

Standards Laboratories in Gaiths- burg, Maryland, and wire standard frequen- cies direct to working positions on test ben- ches, in much the sanie way as the mains at present. In the American set -up a is tuned loop aerial picks up a standard frequency transmission at 60kHz from station

WWVB at California, Colorado. Using standard tech- niques the received 60kHz is used to control a local 100kHz crystal oscillator and a pen recorder keeps a record of the local oscillat- or's performance. The

100kHz is divided down to provide

1kHz, and outputs at

100, 10 and

400, 100, 60 and 1Hz and ampli- fied to provide seven spot frequencies with a stability of one part in 10n over 24 hours at

30V r.m.s. The seven outputs are then connected via balanced lines to the various working positions; one of these is half a mile away from the source.

The advantages of such a system are obvious, no need for a large number of separate frequency standards, one person only responsible for the installation and, course, convenience.

An article on of standard frequency transmissions by J. McA. Steele in

Wireless vide a

World

September starting point for anyone wishing to adopt this technique.

1967 would pro-

Defence message switching system

The Ministry largest, fully of

Defence is to have the automatic message switching system yet ment.

A ordered by the British Govern-

Lim contract has been awarded to

The Marconi Company by the Ministry of

Technology, to supply a message switching and distribution system for the Defence

Communications Centre under central

Lon- don.

The system will be used for switching military teleprinter messages and data traffic and will handle internal communications between the separate branches within the

Ministry, it will also form the hub of the military services network in the U.K. and

The millimetre -wave magnetron

Electronic Tubes factory and klystron production department formerly at Elliott at

Borehamwood, Herts, has been transferred to the

English Electric factory at

Chelmsford. The present frequency range of these components is from 18 to

80 GHz with peak power handling capabilities of up to

50 kW and mean powers up to

50

W. The photograph shows

A.B. Cutting, now manager of the department, in the millimetre wave tube laboratory.

The picture shows the auxiliary pdwer supply, a Deac rechargeable battery, being installed in a

Wayne

Kerr universal bridge model

B221

Mk.

III.

The battery enables the instrument to be used in areas where no power supply is available and enables measurements to be made on in

-situ com- ponents where a mains common link would be embarrassing. overseas. Manual routeing, or handling of messages will be completely eliminated, all messages will be accepted immediately and and stored on magnetic discs, to be transmitted automatically to the appropriate destination as lines become free. The system will accept a simple destination code at the start of each message, which defines a single address, or any one of a large number of circulation

. lists.

The preparation of messages on paper tape at the terminals is entirely eliminated, the messages being compiled on -line, directly into the system. The usual complex com- munications message format is reduced to a very simple plain -language form, suitable for use even by inexperienced operators.

This is accepted by the computer system, and expanded into the full communications format required by the networks to which the system is connected. Messages can be monitored on electronic tabular displays, and edited, without the need from the magnetic store, to extract thus reducing them tran- sit delays previously experienced in systems. manual

Television sets

on

roller

-skates!

A for new product designed to make life easier the harassed housewife has been intro- duced by a Birmingham company

P.O. Box 233,

-

Harrison

Birmingham

12.

The product, called Telemove, consists of four castors on extendible arms that are attached to the legs of any television set. The assembly is com- pleted by adding a magazine rack. The televi- sion set can now be easily moved for house- hold cleaning purposes.

Radio blackout test

A space flight to study methods for prevent- ing the loss returning to of radio signals from spacecraft

Earth has been carried out under the control ics of the National Aeronaut- and

Space

Administration from Wallops

Island, Vancouver.

The eight- minute ballastic flight test,

Wireless World, October 1968

352 www.americanradiohistory.com

RAM C -B, was a continuation of the

N.A.S.A. radio attenuation measurement project to study the problem of com- municating through the ionized gas (plasma sheath) created when a the spacecraft re- enters

Earth's atmosphere at great speeds.

The test is the second in the RAM C series.

Unlike the previous RAM test, in which water was ejected into the plasma stream to restore radio communications, this test will be confined to measuring the amount of electrons and ions which build up around the spacecraft.

The results of the flight will be used with information from the water -ejection experi- ment, laboratory tests, and theoretical stu- dies to provide a better understanding of the problem of communicating through the plasma sheath.

Digital displays for A.T.C.

An order worth about £170,000 has been placed with the Marconi Company by the

Board of Trade for

36 type X2000 graphical display systems. The equipment will be used by the Board

Control at of

Trade

College for Air Traffic

Hum airport, near mouth, to study some of

Bourne- the problems asso- dated with the large flight plan processing system currently being installed at West

Drayton near Heathrow airport,

London.

The equipment will also be used to study better ways of presenting information to air traffic controllers.

An electronic monitoring system ary patients has been installed for coron- at the

Royal

Victoria

Hewlett

Infirmary, Newcastle- upon -Tyne by

Packard. Changes in a patient's con- dition are recorded and a warning signal alerts hospital staff. As well as watching for distress conditions the equipment also in- cludes an electrocardiograph and a pace -maker. with a thin film store with a capacity to one million words with a cycle time of up of

330 ns.

The form of construction is similar to that of the 1906A, twelve -layer printed circuits and emitter coupled logic being employed.

I.C.L. also announced its intention to enhance the ex.

Leo-Marconi System

4 range with the introduction of a new processor with at least twice the power of the 4-70.

L.F. telemetry and broadcast interference

Tests conducted by the Electronic Industries

Association, Washington, has led to a formal

`Objection' being filed with the Federal Com- munications Commission concerning a pro- posal to

20 permit a.m. broadcasters to use the to 36Hz band for telemetry purposes.

If the proposal is adopted broadcasters will be freed from the necessity of maintaining ex- pensive land lines. Listening tests carried out with the co- operation have shown of local radio stations that the l.f. telemetry produces noticeable interference and distortion on broadcast material particularly where voice modulation and i music passages with sus- tained pauses are concerned.

Americans choose Decca for inter

-city flight trials

A seven -week evaluation programme is to be carried out in America to demonstrate the feasibility of using STOL (short take -off and landing) aircraft for inter

-city passenger ser- vices. The aircraft to be used is the McDon- nell Douglas model 188, known in

Europe as the Breguet 941. Due to the steep climb and descent characteristics of this aircraft accurate navigation in three dimensions is essential and a Decca /Omnitrac installation has been installed for this purpose.

New

British computer

International Computers

(Holdings) Ltd have announced a new large computer called the

1908A that is compatible with the earlier

1906A. The new computer can be supplied

Wireless World, October

1968

Navigation and the

SRN4

British Rail's Seaspeed SRN4 hovercraft now in service as a cross

-channel ferry is equipped with Decca navigational aids. At the high is operating speeds, 50 to

70 knots, it essential that some quickly assimilated readout of position be provided.

A pictorial display coupled to a Decca Mk. 8A receiver provides an instantaneous positional readout and a moving chart records the craft's track.

Other Decca equipment fitted to the

Hovercraft are true- and relative- motion radar displays.

Stereo

-radio extended

to

Northern England

The

BBC announce that stereophonic programmes on Radio

3 have been extended to a large part of Northern England

10th August. The from transmissions are from the

Holme Moss station on

91.5MHz. Good stereophonic reception is to be expected within the existing service area of the

Holme Moss v.h.f. transmitter.

S.E.R.T. and R.T.E.B. discuss merger

Successful launching and rapid growth of the Society of Electronic and Radio

Technicians has culminated in discussions for a possible merger with the Radio

Trades Examination

Board, the body which promoted the society in

1964.

Such a move would give

S.E.R.T. access to a body promoting courses and examinations in radio and electronics, and R.T.E.B. examinations would be conducted under a joint body providing membership of a professional society to successful en- trants. S.E.R.T. and R.T.E.B. have recently moved to new premises at

Faraday

House, 8-

10

Charing

Cross Rd, London W.C.2.

Educational television seminar

A

Royal

"Television in Educational Technology

-

what is its role ?" is being held in London at the Institution of Electrical Engineers,

Savoy Place, W.C.2 on

Friday, October

11th

1968 at

3 p.m. Admission is free. Appli- cations for tickets should be made to the secretary, Royal Television Society,

166

Shaftesbury Avenue, London,

W.C.2.

"The

Fuel

Cell"

In the article by G.

Beardmore in the

Sep- tember issue there are two errors.

Fig.

1 showing the percentage efficiency of energy converting devices and systems has its key missing. Cross -shaded areas represent the present range, and solid areas the projected range. In Fig.

3.

(a) both electrodes have been labelled as cathodes, whereas the left

- hand electrode, into which passes the hydrodgen gas, should be labelled "anode ".

Courses

A course of twenty -seven lectures on colour television engineering will again be given at the

Northern

Polytechnic, Holloway Road, London N.7, on Monday evenings, commencing October 7th. Fee 51s. A parallel laboratory course, requiring attendance one additional evening every four weeks, has been introduced this year. Fee 30s.

An film educational course of eight lectures on

"sound for and television" is being organized by the British

Kinematograph, Sound and Television Society for Mon- day evenings beginning October 7th at the Royal College of Arts, Queens

Gate,

London S.W.7. Registration forms are obtainable from the B.K.S.T.S., 110 /2

Victoria House, Vernon Place, London W.C.1. Fee £6.

Among the special courses arranged by the Department of

Electrical Engineering,

Hendon College of Techno- logy,

London N.W.4, for the 1968

/9 session are six - lecture courses on thyristor applications (commencing

October 17th) and colour television (October 24th) and another, of

10 lectures, on principles of feedback control

(October 14th).

Eight lectures on high- fidelity sound reproduction will be held at Wolverhampton

College of Technology on successive Wednesday evenings commencing October

16th.

Further details may be obtained from the College

Registrar.

Fee

£2.

Two short courses be held at of

Barking lectures on solid state

Regional College of physics will

Technology,

Longbridge Road, Dagenham, Essex. Commencing

October 23rd are eight lectures on electrical proper- ties of semiconductors and commencing January 22nd there will be eight lectures on optical properties of solids.

The fee for each course is 2gn. Application forms are obtainable from the college.

353 www.americanradiohistory.com

Personalities

D. East,

M.I.E.R.E., his assistant, who joined the B.B.C. in 1943. Ex- cept for three years when he was in the valve section, he had been in the transmitter section of the planning and Installation Dept. since 1954.

The new president of the Institute of Physics and Physical Society is

M. R. Gavin,

M.A.,

D.Sc., F.I.E.E., principal

Science of Chelsea College of and Technology. Until his appointment three years ago to the

College, of which he will be vice- chancellor when it becomes a uni- versity of technology, Dr. Gavin had been professor of electronic en- gineering in the University College of

North Wales, Bangor, for ten years. He received his doctorate from Glasgow University for work on valves for decimetric waves and spent eleven years at the G.E.C.

Research

Laboratories, Wembley, before entering the academic world.

His first appointment was as vice

- principal and head of the de- partment of physics and mathemat- ics of the College of Technology,

Birmingham. Dr. Gavin, who is

60, has been head of the department of electronic engineering at Bangor since it was set up in 1958.

Roger P. Towell, F.I.E.R.E., chief engineer of the recently formed

Marconi -Elliott

Microelectronics

Ltd., has been with

Marconi for about a year having previously spent

17 years with E.M.I. After seven years in the E.M.I. Engineer- ing Dept. he transferred to the

Research Laboratories and for seve- ral years was primarily concerned with the development of airborne radar. For a year prior to joining

Marconi, Mr. was

Towel, who is 41, manager (technical) of the

E.M.I. Microelectronics Division.

R. P. Towel!

354

Martin J.

L. Pulling,

C.B.E.,

M.A., post

F.I.E.E., who retired from the of deputy director of engineer- ing in the B.B.C. fifteen months ago, has joined the board and has been appointed chairman of the Ferro- graph Company in succession to

R.

W.

Merrick who remains on the board. Mr. Pulling, graduate of

King's College, Cambridge, spent five years in industry before joining the B.B.C. in 1935. He became superintendent engineer, recording,

M.

J.

L.

Pulling in 1941 and was at one time con- troller, television engineering.

W.

E.

C. Varley,

O.B.E.,

F.I.E.R.E., the B.B.C's chief en- gineer, transmitters, has retired and is succeeded by M. J. Crawt,

F.I.E.R.E. Mr. Varley joined the

Corporation in 1933 and after serv- ing at Daventry and Droitwich he joined the headquarters staff of the

Transmitter Department in 1935.

During

1943 and

1944 Mr. Varley was chief broadcasting engineer at

Allied Forces Headquarters in

North Africa. He was appointed as- sistant superintendent engineer, transmitters, in

1958 and two years later superintendent engineer, trans- mitters, later re -designated' chief engineer, transmitters. Mr. Crawt joined the B.B.C. in 1940. Since

1964 he has been superintendent engineer (transmitters

I) with re- sponsibility for all v.h.f. and u.h.f. television and sound broadcasting stations. Mr. Crawt is succeeded by

D.

A. V.

Williams,

B.A., M.I.E.E., has become chief engineer,

B.B.C. external broadcasting in succession to

Dr.

R. R. Sturley, who, as announced in July, has accepted an invitation to the Chair of

Com- munications at the Ahmadu

Bello

University, Zaria, Nigeria. Mr.

Williams, who was educated at

Marlborough

College and Trinity

Hall Cambridge, joined the

B.B.C. in 1950 at Alexandra Palace.

For two years from 1962 Mr. Williams was

New the B.B.C's senior engineer in

York and on his return from the United States he became head the external services of unit in the

Transmitter Planning and Installa- tion

Department. For the past

18 months he has been superintendent engineer (transmitters

II) with re- sponsibility for the operation and maintenance of all the B.B.C's low -, medium- and high- frequency trans- mitters at home and overseas. Mr.

Williams' successor is

C. C.

Butler,

M.I.E.E., who has been his assistant since 1964. Mr. Butler joined the

B.B.C. at Daventry in 1940.

Martin

H. Oliver,

Ph.D.,

B.Sc.,

D.I.C., A.C.G.I., M.I.E.E., head of the Radio Department of the

Royal

Aircraft

Establishment,

Farn- borough, for the past three years has beçome superintendent of the Ser- vices

Electronics Research Labora- tory, Baldock. Dr. Oliver, who grad- uated in 1926 at Imperial College,

London, received his

Ph.D. in 1939 for a thesis on

"Measurements in connection with electrical surge phe- nomena".

He spent the early war years in the Radio

Department at the National Physical Laboratory and in 1943 joined the Telecom- munications Research Laboratory.

In 1956 Dr. Oliver became superin- tendent of microwave electronic research and in 1960, superinten- dent, offensive radar

Radar Establishment. at the Royal

The Independent Television

Au- thority has announced the following three appointments to the new post

H.

French

H. N. Salisbury of Regional Engineer in the Opera- tions and Maintenance Department in preparation for the Independent

Television colour service which will start early in 1970.

H. French,

Regional Engineer

South, has been engineer -in -charge of the Chil- lerton Down, Isle of Wight, trans- mitting station since 1959. Before joining the I.T.A. in 1956 at the

Winter Hill station, Mr.

French was for

14 years with the B.B.C. as a

G.

W.

Stephenson transmitter engineer. He served in the

Royal Navy for four years.

N.

H.

Salisbury,

M.I.E.R.E., Regional

Engineer North, joined the I.T.A. in

1958. The following year he became assistant engineer

-in

-charge of the

Chillerton Down station. In 1961, he was promoted to engineer

-in- charge of the Caldbeck, Cumber- land, transmitting station and was transferred to

Winter

Hill in

1965. Before joining the

I.T.A. Mr.

Salisbury was for

14 years with the

B.B.C. During this period, he spent two years seconded as chief engineer to the Forces Broadcasting

Service in

Cyprus. G. W. Stephenson,

M.I.E.E.,

Regional Engineer

Midlands, joined the I.T.A. in 1958 at Emley

Moor and in August

1959 was appointed assistant engineer

- in-charge at the Mendlesham trans- mitting station. Since

1961 he has been engineer -in- charge of the

Stockland Hill, Devon station. Mr.

Stephenson was with the B.B.C. as a transmitter engineer for

13 years and served in the Royal Navy for three years.

Wireless World, October 1968 www.americanradiohistory.com

Measuring

Delays

Accurately

Technique using recirculation

of

pulses through network under accuracy of

+

0.2%

or better test offers

by

L.

E. Weaver,

*

B.Sc.,

M.I.E.E.

The problem of accurately

*measuring the time of transmission of waveforms through delay lines and similar networks arises in a number of branches of electronic engineering and physics.

The author's particular requirement has been to measure the delays introduced by timing elements and filters used in the signal paths in colour television studio and outside broadcast equipment. For example, the delays in the R, G,

B or Y,

R, G,

B outputs off a colour camera must be matched within very close limits.

The maximum tolerance on a lus delay might be 5ns, which would mean a measure- ment of the delay to appreciably better than

i

%. Since the tolerance is a fixed time rather than a percentage delay, the accuracy required increases with the delay.

The direct measurement of the delay of a wave form through the network by the use of the calibrated sweep circuits of an oscilloscope can be ruled out immediately since even a true accuracy of

± 2% is not easy to achieve by this method and this is far from adequate for the purpose.

The technique favoured hitherto has been to make use of the

"pi- points

", that is, the frequencies at which the phase shift through the network is

180 °, which occur at frequen- cies equal to n

/4t where tp is the phase delay through the network. Provided an oscilloscope is available which has comparable gains on both X and

Y amplifiers the network under test can be measured on an insertion basis.

The delays through the two paths are initially adjusted to equality, so that the cathode -ray tube displays a all straight line Lissajou figure at frequencies when a common signal generator is fed into the two paths.

The unknown network is then inserted into one path assumed to have a good

75 ohm impedance, and the first frequency at which the straight line figure is again obtained is measured by means meter. of a digital frequency

This is equal to

1

/4t, where tp is the phase delay at the frequency of measurement and does not necessarily coincide precisely with the pulse delay of the network since a certain variation in delay over always allowed. the band is

The permissible variations increase as the frequency increases, so it is important to keep the measured delay high, if necessary by inserting a fairly large calibrated delay in series.

The pi-point method is very reliable and easy to carry out, but uncertainties in the

*Designs Department, British Broadcasting Corporation

Wireless World, October

1968 measurement make it less attractive where the highest absolute accuracy is required. It is, however, excellent for matching the delays of samples of the same network, where the abso- lute values of delay are less important, and delays equal within lns are not difficult to obtain.

A technique which the author has used with some success in his work is based on recirculation of a test pulse through the delay network and measurement of the period corresponding to an integral number of the resultant pulses by means of a digital frequency meter. be

The achievable accuracy is believed to better than one or two parts in

103.

The circuit is essentially very simple and can probably be improvised by most laboratories without great difficulty.

The important part the loop in Fig.

1 to the right of is the two splitting pads SP, and

SP2; it consists of a pair of video delay networks in series, one of which is the unknown delay, an adjustable attenuator, a high -grade, wideband amplifier, and the two splitting pads already mentioned.

The latter take the usual form of three

75-ohm resistors in a delta connection or, alternatively, three

25-ohm resistors in a star connection.

The input and output impedances of the amplifier must provide satisfactory terminating impe- dances for the delay networks, assumed nominally

75 ohms. to be

The total attenuation around the loop is adjusted so that the net loop gain is a little less than unity, so that any pulse impressed upon the loop via the splitting pad

SP, will initiate a train of pulses which will circulate around the loop, the interval between the successive pulses being equal to the total transmission delay around the loop, and the decrement of the train being a function of the net loop gain.

Input

2 of the double -beam oscilloscope is used to view these pulses, while input

1 is used to view the initiating pulse.

These two wave- forms can be clearly seen in the photograph of Fig.

2

(a), where for convenience the lower trace has a reduced gain compared with the upper. In order to prevent mutual interference between successive pulses of the train the width of the initiating pulse must be suffi- ciently short in relation to the interval between the pulses, and it is also useful to include a

Gaussian or similar filter in series with the output of the pulse generator in order to give the initiating pulse a good shape, although this is by no means indispensable.

After an interval which may be varied by means next of a control on the pulse generator the initiating pulse arrives, and adds linearly to the waveform existing at that instant at the input to the loop. By choosing the pulse spacing suitably the initiating pulse in each instance may be made to fall precisely upon the nth recirculated pulse following the previous initiating pulse.

This is an adjustment which may be carried out with great precision by adjusting the pulse generator p.r.f. to set the nth pulse to its maximum amplitude

-all

the more obvious since the whole pulse train changes in amplitude and shape

Figs.

2 (c) during the adjustment. parison

This is clearly shown by a com- of Figs.

2 (a) and

2 (b), and also of and

2

(d).

In each instance twenty

Pulsa generator

Digital frequency meter

Attenuator

Fig.

1.

Schematic of the pulse recirculation method of measurement.

1

Double -beam c. r.o.

2

75

SPf

P2

Attenuator

Calibrating delay

Unknown delay

Amplifier

355 www.americanradiohistory.com

adjusted by

1,000ns. the pi -point technique to be

The agreement is excellent, due in this instance to the flatness of the group delay characteristic of this particular network, which improves the accuracy method. of the pi

-point

Acknowledgements. The author's thanks are due to the

Director of Engineering of the

B.B.C. for permission to publish this article.

Fig. 2. Oscillograms from the double -beam c.r.o. (Fig. 1).

Each shows the initiating pulse on the lower trace and the train of pulses resulting from recirculation on the upper trace. recirculated pulses were used and the setting in the right -hand photograph is deliberately misaligned by less than

1% of the initiating pulse period.

When this pseudo-period adjustment has been made the of the continuous sequence of recirculated pulses is evidently equal to the total delay around the loop, so that the appa- rent repetition frequency pulses is nF, where of the recirculated

F is the reading of the digital frequency meter. The total delay round the loop is the reciprocal of the pulse pseudo- frequency, that is

1

/nF. The counter is pre- ferably connected as shown outside the loop to reduce the possibility shunt capacitance. of errors from its

A single measurement of this kind, however, does not measure the value of the unknown, since the delay in the other circuit elements of the loop and the wiring are all included in the measurement. It is not in general practi- cable to measure the residual delay when the unknown delay has been removed because it is too short, so a fixed high -grade delay element, known as the calibrating delay, has been added.

The total delay with the unknown removed can then be measured once and for all and the value retained as a correction for any subsequent measurements.

As will be seen from Fig.

1, care has been taken to insert the calibrating delay into the circuit in such a way that it is buffered from the unknown delay by the video amplifier and the attenuator, so there is no possibility of interaction between the two as a result of impedance mismatching. setting inaccuracy before the twentieth pulse has obviously been passed is one -half a pulse width, say

±3

%. In fact, the sensitivity is such that a very marked change occurs in the appearance of the waveform for a variation in the initiating pulse period of less than

1%

, as is attested by a comparison of Figs.

2 (a) and

(b), where the initiating pulse had a half amplitude duration of

200ns and 20 re- circulated pulses were used. The period was measured by rheans change in of a digital frequency meter.

Tests showed that the setting accuracy in this particular instance was better than

± 0.2%

.

A difficulty can occur in measuring very short delay lines when the initiating pulse is too wide, or when the bandwidth of the delay element is unusually small by corn

- parison with its delay. In each such instance it is possible for successive pulses to fere with one another, resulting inter- in a deformation of the base -line of the train of recirculated pulses which is quite character- istic.

If taken to an extreme it can result in the conversion of this train of pulses into an approximate staircase waveform. This effect can just be seen to appear on the left

-hand side of

Figs.

2 (c) and

(d).

It can always be removed by increasing the length of the cali- brating delay.

Example

Accuracy

The accuracy is determined by the width of the initiating pulse and the number of recircu- lated pulses. on

The former has a strong bearing the sensitivity of setting, and the latter sets a limit upon the possible inaccuracy for a given number of pulses. For example, if twenty pulses are used, a number which has been found convenient in practice, the maximum

In the measurement photographed in

Fig.

2 (a) the counter indicated a frequency kHz, so of

39.99 that the period corresponding to the train of recirculated pulses, that is the total delay around the loop including the unknown and calibrating delay, was the reciprocal of

799.8 kHz or 1,250ns.

With the unknown delay removed and still using twenty recirculated pulses, the pseudo-frequency of the pulse train was

3.998 MHz, that is, a residual delay of 251.1ns.

The delay of the unknown was therefore

1,250- 251=

999ns: it had previously been

356

Announcements

Underwater

Acoustics. The Institute of Physics and

Physical Society have announced the establishment of a new award to be known as the A.B. Wood Medal and Prize. The award, which will be annual, will be made for distinguished work in physical sciences asso- ciated with the sea, particularly underwater acoustics.

A travelling exhibition, Analytex

68, of analytical, measuring and recording instruments, will be visiting several towns between October 21st and November

1st.

Techmation Ltd., will be exhibiting at hotels in

Birmingham, Bristol, Cambridge, Edinburgh, Gates- head, Glasgow, Manchester and Southampton. Invita- tions are available from Techmation

Ltd, 58

Edgware

Way, Edgware,

Middlesex.

Hirschmann Aerials.

Electroustic Ltd, of 73b North

Street, Guildford, Surrey, have been appointed sole

U.K. agents for the complete range of products manu- factured by the West German aerial company, Richard

Hirschmann.

Ultra Electronics (Components) Ltd., Microelectronics

Division,

35

/37

Park

Royal Road, London N.W.10, have been appointed sole IJ.K. agents for Melco semiconduct- or and integrated circuit products.

Environmental Equipments

Ltd, of

Denton Road, Wo- kingham, Berks, are now marketing in the U.K. the range of Tenpac semiconductor strain gauges manu- factured by

Toyota

R

& D

Laboratories in

Japan.

Data

Recognition Ltd,

7

Loverock Road, Battle Farm

Estate, Reading, Berkshire, have been appointed

11.K. distributors for the range of punch -tape perforators and readers used in data control applications manufactured by Ohr- tronics Inc., of America.

An agreement has been reached between the

Wayne

Kerr

Company Ltd.,

New

Malden, Surrey, and the

Comtel Corporation, of Michigan,

U.S.A., for the latter to market

Wayne

Kerr products in the United States.

A range of coaxial connectors manufactured by

Star -Tronics Inc., of Georgetown, Mass., U.S.A., are now available in

Europe and the British Commonwealth through

G.

&

E.

Bradley Ltd, Electral House, Neasden

Lane, London N.W.10.

B. &

K.

Instruments Ltd,

59

Union Street, London

S.E.1, will in future distribute and provide an after

- sales service for the products of Eldorado Electronics,

Concord, California,

U.S.A.

B.

H. Morris

&

Co (Radio) Ltd., of

84 /88 Nelson Street,

London

E.1, have been appointed sole U.K. distributors for the Teac

Corporation of Japan. The Teac range of products was previously handled by C. E.

Hammond.

U.S.S.R. Radio Imports.

Technical

&

Optical Equip- ment (London) Ltd,

15

/17 Praed Street, London W.2, have been appointed sole importers and distributors for all Russian transistor radio receivers.

Advance Electronics Ltd., Roebuck Road, Hainault,

Essex, have acquired the Wrexham company, Filmcap

Ltd., manufacturers of film capacitors.

The Ministry of Technology have placed an order worth more than £250,000 with G.

&

E.

Bradley, of

Electral

House, Neasden Lane, London N.W.10, for electronic multimeters for use by the Armed Forces.

Wireless

World, October

1968 www.americanradiohistory.com

WESCON

1968

New

Show

devices revealed at the

Western and Convention,

Los

Angeles

Electronic

by

Aubrey

Harris

An overwhelming amount of technical infor- mation is presented for consumption by some

50,000 electronics engineers every year at the WESCON (Western Electronic Show and

Convention). At Los Angeles this year (August

20th -23rd) approximately 130 papers were read, at

32 sessions; at certain times as many as five sessions were being conducted simul- taneously.

The show is sponsored jointly by the I.E.E.E. and the Western Electronic

Manufacturers Association.

If one has not had one's fill of information from the papers, there are one or two other events provided to remedy this deficiency: for example, a

Science

Film

Cinema where

22 scientific and engineer- ing films were shown in a gramme daily; the continuous pro-

Future

Engineers' Competi- tion at which twenty -one outstanding high school students exhibit their electronic experiments and compete for $3,400 (approx.

£1,400) in prizes; the

Industrial

Design

Award

Exhibit, at which the twenty leading electronic designs are shown; not to forget the main exhibition by some

650 companies of their latest designs of electronic equipment, components and products.

At the technical sessions solid -state devices generated much interest, not surprisingly, as with the 1969 domestic -U.S. production estimate of some ten million (colour and mono- chrome) television receivers and sixteen million radio sets there is a potentially tremendous market for integrated circuits in home -type electronic equipment. However, there are very few new designs of this equip- ment incorporation i.cs just yet.

Linear

i.cs in domestic and communications equipment

C.

H. Klasing (Mallory Co., Inc.) in his paper

"Integrated Circuits for the Entertainment

Industry" stated the aim of the i.c. manu- facturer is to design units providing in one package replacement for the largest possible number of discrete components. Present -day technology has not yet reached the stage where i.cs can be substituted for existing, well

- proven, discrete-component circuitry and still maintain the performance and reliability required at competitive pricing.

The prediction was that the incorporation of integrated circuits into domestic equipment will be a slow, perhaps

5 evolutionary process taking

-10

years.

Particular discrete circuit functions will be replaced as i.cs become

Wireless World, October

1968 available which compare favourably in performance and price.

One aim is to design i.cs to replace high cost components such as transformers and electro- lytic capacitors, although these items them- selves cannot be constructed in monolithic form. However, their functions can be per- formed by i.c. devices; for example, zener diode regulators can replace smoothing capacitors, audio driver stages using complementary or quasi -complementary transistor design can be substituted for driver transformers and a discrete amplifier stage with an emitter by

-pass capacitor can be replaced by a multi -stage amplifier (with at less cost. emitter resistor degeneration)

The question of price should not only be considered as the cost of the i.c. itself; some early attempts at designing integrated circuits for entertainment equipment necessitated more complexity and cost in decoupling components, biasing resistors and power supply changes than the circuitry they were intended to replace.

In an attempt to reduce the cost of incor- porating i.cs into home equipment circuit packages are being produced which include both an i.c. and a transducer. These packages may contain a tape head, pickup cartridge or microphone together with an associated built -in pre -amplifier. These devices are capable of providing outputs of up to one volt or so, at low impedance, simplifying the task of isolating motor noise and power supply hum from low -level

In transducer leads. the field of communications systems, the adoption of integrated circuits, it seems, will not be as rapid nor with the incorporation as easy of as was digital the i.cs case into computing systems.

This is partly because there is more functional diversity in communi- cations equipment and also because communi- cations designers are more concerned with the relationship between the system and the circuit elements than are computer designers.

The latter have been working with

"functional block" principles prior to the introduction of microcircuits.

During the transition period, before the complete adoption of m.s.i, or l.s.i, it is felt that versatile, low -cost, high- performance single- function circuits will be employed.

Robert

A.

Hirschfeld (National Semiconductor

Corp.) described one unit, the LM171, which can be used as either a tuned or untuned r.f. /i.f. amplifier, a limiter, a detector, a mixer or an oscillator (Fig.

1).

The

LM171 has improved performance over its equivalent and because transistor of the fact that one type of unit can perform many functions, manufacturing procurement problems are simplified.

Another unit, the LM172, uses a

"sub- systems" approach and can be used as a com- plete a.m. i.f. strip including detector. It replaces all components in a the superhet between output of the mixer and the input of the power amplifier.

In current discrete-component practice, the most efficient utilization of available power gain, from a limited number devices, is obtained by matched, of active tuned inter

- stage networks. Because of the limited a.g.c. range obtainable by varying, for example, d.c. emitter current in a conventional common

- emitter stage, several stages must receive a.g.c. voltage from the detector, and consequently require d.c. decoupling, to eliminate effects of changing d.c. operating points with a.g.c.

Economics usually dictate the simplest diode detectors, biased from tuned transformer secondaries, in present strips, while the large tuned gain and often marginal stability force power supply decoupling for each stage.

Attempts to simulate conventially tuned strips by using a number of monolithic amplifiers on one chip, with each input and output brought out to the interstage network, have functioned, but have neither been easy to stabilize nor economical.

The approach used in the LM172 gives a much more efficient arrangement, shown in

Fig. 2. All gain is attained in a single, lumped gain stage, which is direct coupled.

To avoid a.g.c. disturbance of the high gain amplifier, it is left running at maximum gain at all times, with a.g.c. performed by a single, efficient gain control stage, a voltage variable

2

Fig.

1

Schematic

LM171 integrated- circuit versatile

R2

2.5k

10

R,

2.5k

Q'

Q

2

357 www.americanradiohistory.com

R.E input

Fig.

The

X

F,

AG

C line

Gain control stage

F2

C attenuator, with a single decoupling capacitor

(or ceramic filter) taking the place of individual stage decoupling elements in conventional strips. Bandpass shaping is done by an external filter, ahead of the strip, which also serves to couple the strip to the receiver tuner. The lumped gain stage is designed for maximum r.f. effectiveness, and stabilized by a d.c. feedback loop, which sets all bias points at optimum level, regardless of variations in the monolithic process. Finally, an a.m. detector is used which can be directly coupled to the gain stage bias. output which is insensitive to d.c.

This reliably provides the correct a.g.c. voltage, and, by virtue of its "active detector" construction, is capable of audio voltage gain, unlike the loss characteristic of simple diode detectors.

With an active detector operational ampli- fier the audio output voltage is equal to the modulation envelope times the ratio

(R,

+R2)/R2. In the LM172 the audio output is set to about

1 volt peak- to-peak from a

100% modulated signal.

Lumped gain stage

R

R2

Active detector

2

Block diagram of the

LM172 capacitors and ceramic filters

-a

complete a.g.c. stage, if. amplifier

(XF

XF,) are external to the chip. and a.m. detector.

(Fig.

3).

The design of the unit allows for its application in a variety from the of circuitry, apart

"normal" regulator connection using external resistor sampling it may be connected for current limiting (constant current output when overloaded), "foldback" current limiting

(reduced current output when overloaded), s.c.r. overload protection

(a built -in s.c.r. drops load voltage and output current to zero when overloaded).

The usefulness of the regulator may be increased by the addition simply of

(a) one transistor (load regulation increased from

150mA to

5A, and/or

(b) one resistor and one zener diode (input voltage capability raised up to 24V), as shown in Fig.

4.

New display device

Much work has gone into the investigation of various means for producing high brightness, inexpensive, flat display devices suitable for digital displays and as a successor to the c.r.t. for television screens. Dr.

R. H.

Wilson gave details of an experimental unit developed by

Westinghouse for the display of digital infor- mation. It uses a gas discharge technique and is known as the plasma display panel.

It is a matrix display made with a sandwich of three thin plates.

The centre plate is honey- combed with either etched or ultrasonically drilled holes and thin film electrodes are deposited on the outside surface of the outer plates

(see

Fig.

5).

Air is evacuated from the

Audio output panel, and the array is filled with a neon

- nitrogen gas mixture.

Specific plasma display elements can be lit by applying coincident voltages of appropriate magnitude to selected crossed grids.

Alphanumeric characters are generated by writing a specified sequence of elements (dots).

The voltages applied to the electrodes are capacitively coupled into the cell so required. only a.c. excitation voltages are

The display is viewed through one set of highly transparent electrodes and the gas discharge is totally confined to the cylin- drical cavity formed by the glass. three sheets of

The external electrodes are not subject to ion bombardment so problems of cathode sputtering, which would decrease both the lifetime and the brightness of the cell, are avoided.

A single voltage source, capacitively coupled to the array, supplies the sustaining voltage for the whole array. Plasma discharge elements which are initially on are maintained in the on state by the sustaining voltage, while elements which are initially off, remain off.

Elements are turned on (written) or turned off (erased) by the application of appropriate voltages to select x and y co- ordinates. The device has an inherent memory much like a magnetic core.

The gas used is a neon

-nitrogen mixture although experiments have been made with krypton, xenon, carbon monoxide, water vapour and helium gases both with and with- out phosphor added.

Cell sizes have been produced between b- and

85

-thou; a typical unit with 0.015 -inch cells uses a cell density of

40 per inch and gives an average spot brightness of 2056 cd/m2.

Spot brightnesses of

3426 cd/m2 have been measured; there is no flicker on the display, as with a c.r.t. because the picture have to be continuously scanned. does not

Panel lifetime is estimated at 100,000 hours, and the power consumption is in the region 2.5 to

3 X

10-4 watts per cell.

The cost at this time

(in production quantities) is said to be about 810,000 (approx. £4,200) for a unit

2 ft

X 5 ft.

Digital coding

There are many methods of coding picture information for transmission purposes, rang-

I.C.

voltage

regulator

A linear i.c. medium power voltage regulator was described in a and

J.

H. paper by W. H. Williams

Parker

(Westinghouse). The device contained in a

TO-5 can was designed to be used as a regulator, one per circuit card, for supplying operating voltages to such items as operational amplifiers and digital modules. logic

Among the advantages of such a small sized regulator are that one (or more) can easily be accommodated on a single, large printed circuit card; a regulator for powering many cards is not therefore necessary. benefits are

Further that due to the large degree of isolation between individual cards using the regulators, and there is a minimum of interaction mutual interference between cards.

The performance of the unit is truly remark- able:' With a regulated output voltage of between four and

16 volts, a change from no load to 150mA is accomplished with 0.5% regulation in the temperature range

0 to

75

°C.

The output impedance is

0.05 ohms and the input ripple attenuation is

40dB at frequencies up to 20kHz.

The maximum power dissipa- tion is 2 watts (with a heat sink) at temperatures up to 25 °C.

The chip itself consists of seven transistors, three diodes, two zener diodes and one s.c.r.

358

Fig. 3

The

WC109 integrated circuit regulator schematic.

9

Input

12

0

10

Reference

8

Normal output

Q7

Current limit

6

OCompensation

11 o

Sample

Cathode

4Gcte

3

Anode oGround

12

Wireless World, October 1968 www.americanradiohistory.com

Vin

For greater input voltage

For larger output currents r2N3055 Vout

12

Fig. 4

Additional external components for increasing regulation range of WC109 device.

Transparent electrodes b

Glass panels

Glass panels a

Transparent electrodes

Gas cell a

Thin film electrodes

Fig. 5

Plasma display panel.

A rectangular array of gas cells with external

X and

Y addressing electrodes. ing from conventional television line

-by -line scanning to character recognition in which the specific shape of the object is determined and a special code

(or transmitted for each object partial object) shape.

The most efficient system a transmits all required information with minimum channel capacity.

A technique intermediate between those cited above, in both complexity and efficiency, known as contour coding was the subject of a paper by W. F.

Schreiber, T. S.

Huang and O.

J.

Tretiak (M.I.T.).

In the reconstruction lines it is obvious of images from out- that in the case of graphical, i.e. two-level data, the image may be recreated exactly from the outline information since all that is necessary is to fill in the spaces between the outlines with black.

It turns out that continuous also be recreated exactly from the outlines provided also that they contain information about the spatial gradients of the image.

Just as a function of a single variable can be reproduced to within an additive constant from its derivative, a function of two variables, i.e. an image, can be recovered from its gradient.

The outlines are generally found to consist of connected series of points having gradients significantly different from zero.

Thus if all the significant outlines are transmitted all the significant gradient information will be available for picture reproduction.

The information which needs to be trans-

Wireless World, October 1968 mitred about the outlines or contours in an image consists simply of the location of the contour points in the case of graphical data or the location plus gradient information or area brightness levels in the case of continuous tone images. A series available for of coding techniques is transmitting this type of informa- tion, ranging from simple and inefficient to complicated and highly efficient. Simple in this context refers to the amount of logic required for the coding and decoding and most importantly to the amount of storage re- quired, especially at the receiver.

Four possible systems were described.

The co- ordinates of all the points along all of the contours may be transmitted without taking advantage of the fact that the points are connected. In the special case where a random access picture device is available at both transmitter and receiver no storage is required. This system gives useful savings only for graphical data consisting mostly of blank paper.

Alternatively, instead of locating each point individually, they may be located with respect to previous points on the same line.

Similar transmission efficiency can be obtained by tracing the contour directly. Each contour point is located with respect to another which is only one picture element distant, except for the initial point of each contour. One may thus transmit all the points on one contour before proceeding to the next contour. This obviously also substantially increases both the storage and processing t'equirements but it permits greatly improved efficiency.

The most straightforward way of transmitting the contour information is to give the co-ordinates of the initial points of the contour plus incre- mental information from which the remaining points may be found. Since each contour point is adjacent to the previous contour takes only a limited amount point it of information

(a maximum of

3 bits) to indicate the direction of the next point. In the case tone images where also be of continuous gradient information

- must transmitted, one may also transmit the value of the gradient at the initial point and incremental information permitting the calcu- lation of the gradient at the rest of the points.

A more advanced method is to fit each contour with the most nearly matching mathe- matical curve, starting with a series of straight line segments and ing then where required bring- into use higher order

curves-

quadratics or exponential.

Conferences and Exhibitions

Further details are obtainable from the addresses in parentheses

LONDON

Sept. 30-Oct. 2

Tropospheric Wave

Propagation

(I.E.E., Savoy Pl.,

London, W.C.2)

Savoy Place

Oct. 2 -5

R.H.S. New Hall

R.S.G.B. Radio Communications

(P. A.

Exhibition

Thorogood,

35

Gibbs Green, Edgware,

Middx.)

Oct.

15 &

16

St.

Ermins Hotel

Ultrasonics for Industry

(Ultrasonics,

Dorset House, Stamford Street,

S.E.1) www.americanradiohistory.com

Oct. 30 & 31

Middx. Hosp. Medical School

Electronic Weighing Conference

(I.E.R.E.,

9

Bedford Sq.,

London, W.C.1)

BRIGHTON

Oct.

8

-10

Hotel Metropole

Intersolidus

-Nepcon-

Electronics Packaging Con- ference

(Gordon Saville

Surbiton, Surrey)

Exhibitions, 21

Victoria Rd,

KENILWORTH

Oct. 16-17

Metrication

(Inst.

...

the problems of

Chesford Grange change of Prod. Eng., 10

Chesterfield St., London, W.1)

MANCHESTER

Oct.

22 -24

Hotel Piccadilly

Electronic Instruments Exhibition

(Industrial Exhibitions Ltd., 9

Argyll St., London W.1)

OVERSEAS

Sept. 30

&

Oct.

1

Philadelphia

Computer Impact on

Engineering Management

(Instrument

Soc. of America, 530 William Penn Pl.,

Pittsburgh, Pa)

Oct. 2-4

Circuit and System Theory

Monticello, Ill.

(

I.E.E.E.,

345

East 47th St., New York, N.Y. 10017)

Oct.

5

-14

Intl.

Radio

&

Television

Exhibition

Bordeaux

(Foire de Bordeaux, 12

Plac de la

Bourse, Bordeaux)

Oct.

8

-11

Telemetering Conference

(B.

E.

Los

Angeles

Norman, P.O.

Box

56,

White Sands Missile

Range, New Mexico 88002)

Oct. 9-15

Düsseldorf

INTERKAMA- Instrumentation

&

Automation

Congress

&

Show

(NOWEA, 4

Düsseldorf

10,

Postfach 10 203)

Oct. 10

& 11

Washington

Applications of Ferroelectrics

(H. L. Stadler,

Phys. Elec. Dept., Ford Motor Co.,

Dearborn,

Mich.)

Oct. 14 & 15

Framingham, Mass.

Thermionic Energy Conversion

(J.

E. Kemme, Los Alamos

Sci. Res.

Labs., P.O.B. 608,

San Diego, Cal.)

Oct. 14-16

San Francisco

System Science

&

Cybernetics

(H. Mays, Fairchild Semiconductor

R. & D.

Labs.,

4001 Junipero Serra

Blvd., Palo Alto,

Cal.)

Oct. 15 -17

Schenectady

Switching and Automata Theory

(S. B.

Akers, Electronics Lab.,

General Electric

Co.,

Syracuse, New York 13201)

Oct. 15

-18

Reliability in Electronics

(Mrs.

A.

Valko, Scientific Soc. for Telecom.,

Haza, Szabadsag

Ter

17,

Budapest V)

Budapest

Technika

Oct.

17

-27

Bucharest

British

Industrial Exhibition

(Ind. and

Trade Fairs, Commonwealth

House, New

Oxford

St,

London W.C.1)

Oct. 19-27

Intl. Communications Fair

(Ente Autonomo Fiera Internazionale, Pizzale

Kennedy, Genoa)

Genoa

J. F.

Oct. 20-22

Tel

Aviv

Electrical

&

Electronic Engineers Convention

(I.

Cederbaum,

Israel Inst. of Tech., P.O.B. 4910,

Haifa)

Oct. 23 -25

Electron Devices

(D. A.

Chisholm, Bell

New Jersey 07974)

Washington

Telephone Labs, Murray

Hill,

Oct. 28-30

Gatlinburg

Applied Superconductivity

(Dr. W. F.

Gauster,

Oak Ridge

National Laboratory,

Bldg 9201 -2, P.O. Box Y,

Oak Ridge, Tennessee)

Oct. 28 -Nov.

1

Fiarex

-Electronic

Components Exhibition

Amsterdam

(Vereniging F.I.A.R., Minervalaan

8211s,

Amsterdam)

Oct. 30 -Nov.

1

California

Circuits

&

Systems

(Prof. Shu

-Gar Chan, Dept. of Elec. Eng., Naval Post- graduate School,

Monterey, Cal.)

359

Wireless World Colour Television

Receiver

5.

Convergence circuits and adjustments

Because the three guns of the cathode -ray tube cannot occupy the same physical position the red, green, and blue rasters produced by them and the common deflection fields are displaced from each other on the screen. The so- called con- vergence coils, magnets and circuits are necessary three rasters into register with each other. to bring the

The mains deflection fields for scanning are produced deflection

-coil assembly by a of the usual form which carries the sawtooth currents at field and line frequencies.

These fields act simultaneously on all three electron beams. To obtain register of the rasters subsidiary deflecting fields are needed which act separately on- the three beams.

The convergence assembly is fitted to the tube neck behind the main deflector -coil assembly; the deflector coils.

Three sets of that is, between pole pieces tube and have internal screens between them so the guns and are built into the produced by one does not affect the other beams. Magnetic screening is, of course, far from perfect and reduces interaction between the fields rather than and one permanent magnet. One coil carries

. that the screening eliminates

The convergence assembly comprises three sub -assemblies mounted radially with

120° between them, one for each electron beam, and they are denoted by the colours which the beams excite on the screen. Each sub -assembly has two coils current the at field it. field frequency, the other at line frequency. Usually, each coil has two windings which may be connected in series or parallel as required by the circuitry

Used.

The windings are on two strips of magnetic material and across them is at one end is mounted a disc- shaped magnet which magnetized across a diameter and which can be rotated so that the strength of the magnetic field in the cores can be adjusted. The cores at the other end come near to the internal pole -pieces in the tube. Thus by turning the magnet a static deflection can be given to the beam, and by passing currents through the coils the beam can be deflected at line and field frequencies. The two are usually known as static and dynamic convergence.

The blue components are mounted vertically on top neck of the of the tube and provide a field which deflects the blue beam vertically. The red and green sub -assemblies are mounted at 120° to each other and to the blue and they provide deflections on along the same angles. Viewed from the left and green on the right. Thus

if

a the front, red certain current is in the blue coils gives a certain deflection downwards, the same of the blue beam current in the red coils will deflect the red beam by more or less the same amount but at an angle

120° to to the right. The same thing happens with the deflection is 120° from the blue green of

"downwards ". The deflection can be resolved into two components, one vertically upwards and the other horizontally beam; but the other way, and it can be resolved one into a deflection vertically upwards combined with horizontally to the left.

360

Because of the combined vertical and horizontal movements obtained for the red and green beams nothing

Convergence can be obtained by adjusting the currents and waveforms in the two assemblies.

An additional coil and magnet are needed to give a horizontal movement to the blue beam, however, and these are provided in what further is needed. is called the blue lateral assembly. This is a small assembly which is mounted on the tube neck behind the main convergence assembly.

In addition to all this, the convergence assembly carries two purity magnets. These are mounted at the back of the main convergence

-coil assembly and comprise two flat rings which can be rotated around the neck magnetized so as to produce a magnetic field which acts across a by diameter of of the tube. The strength the tube. They are each of the field can be varied turning one magnet relative to the other; when like poles are together the field is a maximum, when unlike poles are together it is a minimum. The direction of the field across the tube can be altered by turning both together.

'The purpose of these magnets is to ensure that the red beam excites only the red phosphor dots only green dots, of the screen, the green beam and blue only blue.

Adjustment the position of the two purity magnets in conjunction with of the deflector -coil assembly enables a red, green, or blue raster of substantially uniform colour to be obtained when the appropriate gun is operating alone.

The adjustments are not usually very difficult.

Adjustment of the three convergence magnets and the blue lateral magnet readily enables perfect convergence to be ob- tained at the centre of the screen. For all convergence adjust- ments a crosshatch pattern generator must be regarded as an essential piece of equipment, and a suitable generator was described last month.

Static convergence adjustments readily give substantially perfect convergence within a circle the centre bars of some

2 inches diameter at of the screen. Outside that it gradually deteriorates and towards the edges of the screen the red, green and blue of the crosshatch pattern will be quite separate instead of being merged to produce white.

To correct for this it is necessary to pass currents at line and field frequencies of the proper amplitudes and waveforms through the convergence coils.

The usual practice is to limit the waveform to a combination of a sawtooth and a parabola.

Nothing more elaborate than this is normally attempted and because is of this perfect convergence over the whole rarely obtainable of the screen

The need for convergence arises mainly because the three guns do not occupy the same position. However, the ampli- tudes and waveforms degree by of the convergence currents which are needed to achieve perfect convergence are affected in some many other factors. For example, manufacturing tolerances in the tube and in the deflector -coil assembly, to mention only two things, affect the perfection of convergence

Wireless World, October 1968 www.americanradiohistory.com

13

G rey

Fig.1

..252-3<

August

4

Orange

Black

'Yellow.

B Lat

AT4040/58

L

Line deflector coils

R1

470

1W

RG

/L

Symm

Plessey or

405/1/05007

T a rf 3

BLat

AT1025/00

AT4040/50 pincushion

Rg

B/L Tilt

R4

25

B/L

Amp

R6

25

8

,,

",.^

R5 C2

6.8

3W

0ß1

B/L

Para

AT4040/75

L3

C3

0.068p s-

6 Blue

RG/L

Amp

7`- 000-5

B/L convergence coils

Blue

7

G/L

Green

R/L

Convergence coils

White

I

Cl

I0.047µ

R2

3.3k

2W

R3

5

006`

12

RG/ F Symm

Field

Fig.6 pt.2 deflector p.196

July coils

RG

/F

Tilt

Convergence coils

R/F G/F

12

Green

Pink

Red

Black

Red

3

Pink

RG

/FDiff

RG

/FPara

FPara

B/F convergence coils p

Fig.3

194

July

Boo

To

Al in

Fig.1 p.252 -3

August

1st anodes on tube base

Connections of

Mullard

Type transductorT

AT4041/05 a

O

5

O

3

O

1 o a o o b o

6 o

4 o

2

O b

Fig. 1.

Complete circuit diagram

(Electroniques stock number of the convergence unit.

The switches employed are

Rendar Instruments

168F27391D).

They are d.p.d.t. switches, but only one pole in each is used. miniature slide type, model

R

53005 obtainable with given convergence circuitry. In any given case it, would no doubt be possible to devise circuitry'which would give substantially perfect convergence over the whole screen.

The cost and difficulties of adjustment would rise con- siderably, however, and the usual practice is to use an arrangement which theoretically allows perfect convergence only along the vertical and horizontal centre

-lines of the screen.

This is usually adequate.

It must be understood show that imperfect convergence does not at all in uniform areas. It shows only on edges where there is a change of colour or brightness and it is evident then as a colour fringe on the edge.

It is a fact of observation that a degree of imperfection in convergence that looks intolerable on the cross

-hatch pattern is much less evident on

Test Card F, is not at all intrusive on a black- and -white programme picture and may be almost un- noticeable on a colour picture. It can always been seen

if

it is looked for, of course, but it often passes unnoticed when watching a programme. The reason is, of course, that the main interest of the picture is usually towards the centre and it is there where one's attention is concentrated. However, as well as this, there is often less detail in the picture towards the edges, especially towards the top, which may be partly a sky area.

There are a great many controls needed for convergence and they have to be adjusted while watching the screen closely.

It is very desirable, therefore, mounted that all the controls should be so that they are readily accessible while watching the

Wireless World, October

1968 www.americanradiohistory.com

screen.

The convergence panel is therefore mounted immediate- ly over the tube. In addition to the actual convergence controls the panel also carries voltage controls for the three first anode voltages of the tube and switches for rendering any guns inoperative at will.

The circuit diagram of the unit is shown in Fig.

1.

The upper part of the diagram shows the interconnections between the deflector coils. At the top left a lead from terminal

7 line

-scan transformer and one from the linearity coil, both in the line timebase unit, are brought in.

The lead from

7 is taken out again to the junction the other ends of the of the two line- deflector coils, while of these coils are brought back as two separate leads to the coil

L

which is

RG/L Symmetry. This enables the relative amplitudes coils to be of the currents in the two line- deflector adjusted. If, for example, the greater in the top coil length current is made a little of the pair than in the bottom, then the of the scanning lines at the top of the picture will be a little greater than that of those at the bottom.

Connected in shunt with the deflector coil and the symmetry

470-5

resistor R, in series with a winding of a control is a transductor.

This is a special transformer having a readily saturable core, the secondary of which is connected, with some modifying components, in series with the field- deflector coils.

The effect of the non -linear core is to cause some intermodula- tion and of the line and field scanning currents. This reduces the amplitude of the line scan at top and bottom of the picture, of the field scan at left and right to correct for pincushion

361

o o o o o o o o o o o o o o o o o o o

Grey

Blue

Yellow

Orange

Black

White

Green

Pink

Black

Purple

Brown

Red

Green

Yellow

Green

White

Grey

White

Black

Blue

Yellow

7

Green

Green

Purple

Green

1

Pink

2

1

/ s

7 on line scan transformer

Linearity coil

5 on Tine scan transformer

11

See

Fig.3,pt.2

Blue lateral coil

5

See

Fig.1 pt.3

(on fbd`

,

Cathode

Labelled

Field field

Al

, scan

Fig.1 pt.3

Deflector coils pb output valve

}Figs.384.

Line

Line convergence coils

Field convergence coils

Fig.

2.

External connections of the connector blocks are shown in this diagram. distortion of the raster. The coil

LZ provides an adjustment of the magnitude of the effect.

At the top of the diagram, leads from the field -scan output transformer are brought in from the timebase unit.

One immediately goes out again to the junction of the two field- deflector coils, the other ends of which come back as separate leads to RG/F

Symm,

R3.

This serves the same purpose in the field circuit as L, does in the line.

These two symmetry controls are used as two of the convergence controls.

The convergence circuitry proper is in the lower part of

Fig.

1.

On the right are the field circuits.

The connections on the right are all from the timebase unit.

A sawtooth

. waveform is brought in at `pink', which is from terminal

5 of the field output transformer, and a parabolic waveform is brought in at the bottom terminal, `purple'. This is actually connected to the field output

-valve cathode through a large capacitor in the timebase unit.

(C,

,B shunted by

R,913 of

Fig. 3, Part

2.) At this point the parabolic waveform is developed naturally.

The amplitude of the sawtooth applied to the red and green convèrgence coils is controlled by

R

RG/F Tilt, while the amplitude

Para. of the parabolic current is controlled by R,

3,

RG

/F

The relative amplitudes of the currents in the red and green coils is controlled by

R,,

RG/F Dill.

For blue convergence the sawtooth is fed to R,

5,

B/F Tilt, through

C4 which actually comprises two 320

-1,F capacitors in parallel.

The parabolic component of current is controlled by

R145

B/F Para.

The phase of the parabolic waveform is fixed, being derived from the valve cathode. The relative phase of the sawtooth component can be reversed by connecting to the other end of the tapped convergence winding on the field output transfor- mer. Additionally, the effect on the tube can be changed by reversing the convergence

-coil connections.

The connections given later are the normal ones, but it cannot be guaranteed that with some tube -deflector coil combinations some such changes to connections may not be desirable.

On the line side, matters are a little more complex because the inductance of the convergence coils becomes important.

As a result, the circuit is more complex.

All the coils are basically in series between

5 on the line -scan transformer and chassis, which means that they are in series with the deflector

-coil circuit. The potentiometer

R, o,

RG/L

Diff, enables the relative amplitudes of the currents in the red and green coils to be adjusted. RG /L Tilt,

L4, controls the relative amplitudes of sawtooth and parabolic components of the current, and RG /L

Amp,

R8, the amplitude.

In the blue part of the circuit

B

/L

Amp,

R6, is a shunt across the blue convergence coils.

The network

R4, R5, and

L3 enables the waveform to be adjusted.

The

B

Lateral coil is fed from terminal

4 of the line

G,

C3 output transformer through the controlling coil B

/Lat.

Included in the convergence unit, but electrically quite separate are the first anode controls for the tube. Good

'insulation for all these is necessary.

The various convergence controls tend to require operation in pairs and so as far as practicable they are grouped physically in pairs in order of adjustment. It is unfortunate, however,; that all the convergence controls without exception interact to some degree and late adjustments in the sequence throw out early ones in some degree.

It is this which accounts for the greater part of the difficulty of convergence adjustment and which makes

it

very much an art and a skill to be acquired by practice.

At a first trial some of the controls may well appear to have no effect. This does not necessarily mean that there is anything wrong. When other controls are properly adjusted they will be found to have an effect, perhaps small but sufficient.

With the exception anodes of the three connections for the first of the tube, which are soldered directly to the switches, all connections are via connector blocks. There are three of these, one

6

-way and two

12

-way. The diagrams give the arrangement and the colour- coding adopted; this last need not be adopted, one of course. However, some colour coding is almost essential to keep track of so many leads and to adopt the same facilitates reference to the diagrams.

An insulating material panel is used to carry all the controls.

The use of a metal panel at chassis potential is dangerous because the chassis is live unless an isolating transformer is employed.

If

at all possible this should be used when making adjustments to parts on the tube neck, for one can hardly do this without coming into contact with some

Watch it. part of the chassis.

Before starting to adjust convergence it is necessary to have at least the luminance output stage working, because the output of the cross -hatch generator is injected into its grid circuit. It is desirable to have the whole receiver working so that proper adjustments to picture width and height and linearity of the scans can be made, as well as to the e.h.t. supply. These things affect the scanning waveforms and their amplitudes and so the precise waveforms available to the convergence circuits. While making these adjustments, the core of the raster -connection or pincushion coil L, is adjusted to make the raster as rectangular as possible.

Wireless World, October

1968 362 www.americanradiohistory.com

Static convergence magnet o

Purity magnets

The diagram shows a rear view of the

Plessey convergence assembly.

For series coil connection, connect to tags

1 and

4 and join 2 and

3 for the field circuit and connect to tags 5 and

8 and join 6 and

7 for the line. to

3 to

For parallel coil connections, connect tags

1 and

4 and join

1 and 2 also and

4 for the field circuit and connect tags 5 and

8 and join

5 and 6 also

7 and

8 for the line.

For series connection, the have a resistance of

16852 field coils and an inductance of 1.18H while the line coils have a resistance of 2.952 and an inductance of 0.5mH.

Parallel connection is normally used but sometimes series connection of the blue line coils may be needed.

Static ileonmvearggneentce

Static convergence magnet

Assuming these to have been done, the first step is the adjustment of purity. The following is the proper sequence:

(1)

Slack the wing puts holding the deflector coils in housing. Move the coils as far back as they will go. their

(2)

Switch on the cross

-hatch generator. Switch off the blue gun. Adjust the red and green static convergence magnets so that at the centre of the screen the red and green rectangles of the pattern become superimposed to provide a single yellow pattern. Remember that the patterns can only move at 120° to each other; that is, at

30° to the horizontal. Alternate small adjustments to each control are needed, and remember that if vertical (or horizontal) bars are superimposed exactly by one control and horizontal

(or vertical) bars are not, then no adjustment to the other control will superimpose both.

(3)

Switch on the blue gun and adjust the blue convergence magnet and the blue lateral magnet to superimpose the blue reactangles at the centre of the screen on the yellow to produce white. This is much easier to do, for the blue convergence magnet moves the blue moves it raster vertically, and the blue horizontally. Unfortunately, there is some lateral interaction, and the blue adjustment will slightly upset the red and green adjustments.

At this stage this does not matter for only rough static convergence is necessary. At a later stage, however, it will be necessary to return to the red -green adjustments with off and then to turn on the blue gun and the blue gun re- adjust the blue magnets, repeating the process until all three beams have been made to converge at the centre of the screen.

(4)

In the early stages of

(2) and

(3) operate the picture -shift controls to make the edges of the rasters plainly visible to make that one is attempting to align the same rectangles in quite sure each raster at the screen centre. With everything out of adjustment it is possible to try to align the wrong rectangles, although it is unlikely

(5) and blue guns that one will succeed in doing so.

Switch off the cross

-hatch generator.

Turn off both green and adjust brightness and red

1st anode voltage for a reasonably bright raster. This will have a red patch surrounded by variously -coloured patches. Adjust the two purity magnets to get the best red patch possible in the centre of the screen.

The patch will be very ill- defined and is likely to be far from circular.

Wireless

World, October 1968

Each magnet has two tabs, one with a rounded end and the other with a square end. When two similar tabs are together the field is at its strongest, and when dissimilar tabs are together it is at its weakest. The effect seen and spreading them of turning them together is readily apart weakens or strengthens the field as the case may be.

When satisfied slide that the best red patch has been obtained, the deflector coils forward until the whole screen is a uniform red.

If

the coils have to come fully forward and the screen is still not uniform re- adjust the purity magnets endeavour to obtain a complletely uniform red screen. in an

(6)

Switch the green gun en and the red gun off.

Check that the

(7) raster is a uniform green.

Switch the blue gun on and the green gun off.

Check that the raster is a uniform blue. Adjustments to the green and blue 1st anode voltages may be needed to make these rasters visible.

(8)

There are no extra adjustments for the purity of the green and blue rasters separately from the red. Normally, none is needed.

If

the green and

/or blue rasters appear less pure than the red it usually means that the assessment of red purity has not been good enough and one must try to get the red purity better.

Convergence

We now now come to convergence. The cross

-hatch generator is kept on permanently, and for about half the stages the blue gun is switched off.

Adjust the anode voltages so and green bars are that red of about equal intensity and so that they give a clear yellow wherever they are superimposed.

(1) Adjust the red and green static convergence magnets so red and green rectangles that near the centre of the screen are superimposed.

(2)

Look at a vertical yellow line bar near the centre. It will be a single near the centre where the static convergence holds good, but above and below it will spread into separate red and green lines.

The aim is now to adjust RG

/F

Tilt and RG

/F

Para to make these red and green lines coalesce throughout their lengths into a single yellow line.

Both controls affect www.americanradiohistory.com

363

things throughout the length effect of the lines, but Para has a major at the bottom, whereas Tilt appears to twist the line about a centre pivot.

Adjust the controls alternately and only a little at a time. Do not worry if the adjustments separate the lines in the centre, but then concentrate upon obtaining parallel red and green lines.

(3) Now adjust RG /F Diff and RG /F

Symm to obtain equal separation of horizontal bars of the pattern where they cross the vertical centre bars.

If it is possible to superimpose the horizontal bars where they cross the vertical centre bars well and good.

If

it is not, concentrate upon obtaining equal spacing of the red and green bars, making quite certain that all the red bars are on the same side of the green bars.

(4)

Re- adjust static convergence for a yellow centre rectangle.

(S)

Repeat

(2) and

(3), this time endeavouring to superimpose the red and green vertical centre bars to obtain a yellow bar, and to superimpose the horizontal red and green bars at the points where they cross the vertical centre bar. From time to time slight re- adjustment

(6)

Next adjust RG of static convergence may be needed.

/L

Amp and RG IL Tilt; The first affects the separation of vertical bars on the left -hand side of the raster, the second does the same on the right. The two interact and also have some effect on the separation in the middle. The aim is, therefore, to obtain equal separation green bars on the same side of of all the red, vertical bars with so that they can all be superimposed by an adjustment of the static magnets, which then do.

Comment. There is usually little difficulty in obtaining almost perfect convergence of all the vertical bars and of the horizontal bars where they cross the centre vertical bar, by adjustment of these six controls plus the red and green static magnets. The controls may have to be gone over three or four times in succession before this is obtained. The cross hatch pattern will then be substantially yellow but horizontal bars away from the centre will probably be out

(7) of register.

There are two controls which act to bring the horizontal bars into register, RG

/L Diff and RG /L

Symm.

These are adjusted alternately to bring the horizontal bars as nearly as possible into register. In doing this, the earlier adjustments may be adversely affected, in particular

RG

/F Diff and RG /F

Symm, so it is best to best register work these four controls together for the of horizontal bars. The other controls affecting the registration found of vertical bars need be re- adjusted only if it is that they have been upset by the later adjustments.

Comment. Never attempt to correct any defect completely by the operation of one control.

If

it can be done, then that setting of the control is almost certainly wrong and will prevent the proper adjustment of some other control.

Aim at adjusting each control only to improve matters somewhat, and then by repeat- edly going over the whole process one can an gradually arrive at optimum condition. Perfect registration over the whole pattern is unlikely to be achieved.

The best that can be achieved may well be that horizontal bars go out of register at diagonally opposite corners, and the extreme left and right vertical bars may not be perfectly superimposed. Only ex- perience will tell whether a given result is the best that can be achieved with particular equipment or whether one can do better with more adjustment.

(8) Now switch on blue the blue gun and attempt to converge the pattern with the yellow.

Start with the blue static magnets to converge the blue centre rectangle on the yellow to obtain white.

Then adjust BF /Tilt and BF

/Para for the proper separa- tion of blue horizontal bars where they cross the yellow vertical bar in the centre of the screen.

Tilt has a more or less uniform affect throughout, whereas Para affects mainly the top of the picture.

As in the case to of other adjustments, the aim should be obtain uniform spacing of the blue and yellow bars with the one always on the same side of the other, rather than superposi- tion, for the latter can then always be obtained by adjusting the blue convergence magnet.

(9)

Adjust

B

/L

Amp for maximum droop of the horizontal blue bars.

The control will be fully one way. Then adjust

B

/L

Tilt and

B

/L Para to make the shape of the horizontal bars roughly symmetrical about the vertical central line of the tube.

The waveform is mainly parabolic and Para adds a second

- harmonic component which has the effect of flattening the droop will of the blue bars around the centre. Complete symmetry not be obtained, and maximum droop will be to the left of centre.

(10)

Adjust

B

/L Amp so that the shape of the horizontal blue bars is as nearly as possible the same as

(11) Re- adjust

B that of the yellow.

/F

Tilt,

B

/F Para and the vertical blue magnet to superimpose the blue horizontal bars on the yellow where they cross the vertical centre line.

Then re- adjust

B

/L

Amp,

B

/L Tilt and

B

/L Para to get the horizontal bars super- imposed as far as possible everywhere. Repeat as necessary.

(12) Adjust the static

B

Lat magnet to superimpose vertical blue bars on yellow at the centre. Adjust

B

Lat to align vertical bars at left and right; the connections to the coil may need reversing. In effect this scan so alters that amplitude of the blue line that the blue raster can be made of the same width as the yellow.

The control must thus be used in conjunction with the static

B

Lat magnet.

When all find this has been done, do not be disheartened to that the red -green convergence has deteriorated.

Because of interaction this is inevitable and it is necessary to go through the whole procedure again.

To save space, the foregoing description has been written as a complete drill, and the warning not to attempt to adjust any control for perfection should be heeded, even if it seems hard to do so.

In practice, it is best initially to go through the whole procedure of convergence adjustment quite roughly, and then

The layout adopted for the convergence controls is shown in this drawing.

The panel is screwed to the top edge of the panel which holds the cathode

-ray tube.

Switches

Timebase

Deflector coils

Convergence coils

R

G

B

43, 4 in

364

211/2 in www.americanradiohistory.com

Wireless World, October 1968

This photograph shows the underside of the convergence panel and indicates the positions of some of the chief components. The layout is

far

from critical.

L5

L3 to repeat it a little more carefully. This will bring all the controls to somewhere near their proper settings and so will reduce the magnitude of interaction between the controls at the later stages.

If

this is done, then it is likely that going over the full procedure twice more, very carefully, will suffice.

The main practical difficulty is knowing when to stop!

Perfect con- vergence over the whole screen is theoretically impossible and it is a matter of individual judgement whether or not the optimum condition has been reached in any given case.

It is not possible to indicate how much misconvergence will be found, for it depends very much on component tolerances, including the c.r. tube. Since the convergence waveforms are derived from the scanning waveforms, clearly anything which affects the linearity of scan will affect convergence, hence the need for making the timebase adjustments first.

The cross hatch pattern shows up misconvergence very clearly, much more clearly than an average television picture.

When one has reached a stage in the adjustments at which it seems hard to get the convergence any better, it is helpful to pattern by a television picture, which need not be á replace the colour picture. The newcomer will be surprised how little noticeable is a degree of misconvergence which, from the cross hatch general, the noticeable effect bad on the cross hatch pattern, fair on monochrome test card F, easily pattern, he would expect to be intolerable. In of misconvergence is judged to be tolerable on a monochrome moving picture and still less noticeable on a colour picture.

The reason is, of course, that misconvergence only shows where there are changes appears only on the edges of luminance or colour. It thus of objects in the picture and its magnitude depends on the amount or colour. Then misconvergence normally appears the edges of the change of luminance picture is usually the central area, so only towards of the picture, whereas the main interest of the that one's attention is concentrated on the part where the convergence is good.

More elaborate circuitry with extra controls would certainly enable improved convergence to be obtained but it is not that the improvement is worth the extra generally considered complication and expense. We have done no work on such of this nature. circuitry and so cannot recommend anything

The interaction between the convergence controls could also be reduced by more complicated and expensive circuits, would not eliminate it completely because some but it the tube itself, and this seems to be inevitable. It does not, therefore, seem worth while to go to a lot of of it arises trouble merely in to reduce the interaction.

Wireless World,

October 1968

It is, of course, necessary that the tube and any iron -work near it be demagnetized before any of these adjustments are made, including the purit,ï adjustments. The automatic de- gaussing circuit will normally take care will of this. However, it not deal with metal -work outside the tube shield and sometimes other measures will be needed.

A separate degaussing coil can be wound on a circular former of

12- inches inside diameter with a winding length of

i

in. and a depth of

1 in., so that the outside diameter is 14 in.

Some 800 will turns of

No. 24 enam. wire are needed and the coil take 4.1 lb.

The resistance is

60 Q with an inductance of

0.43H, and the coil take

1.5A on 240 -V, 50Hz mains.

In use the coil is slowly in side front of connected to the mains and moved about the tube and round the sides and top with its parallel to them. It is

:hen withdrawn slowly to a distance of some

6 ft. before it is switched off. Avoid getting the coil near any permanent magnets; it is not desirable demagnetized!

It is unlikely that these should be that this degaussing coil will be needed, and almost certain that it will not be if the purity adjustments can be carried out to obtain rasters of uniform colour.

Corrections

In Fig.1, Pt.4, p.322, September,

C, is drawn as an electrolytic capacitor.

In view of its voltage rating it is obvious that it should be a paper -dielectric type. On p.323, the mains trans- former secondary was inadvertently given as

35-0-35V; it should be 35

V,

12

-way

C.T., that is

17.5-0-17.5 V.

On the same page, the connector type number appeared as B751 instead of

BTS1.

Wireless World Digital Computer

The series of articles describing the Wireless World digital computer which appeared in the August to

December 1967 issues is being reprinted. The 36 -page reprint, which will be ready at the beginning of October, will cost lOs and can be obtained from Dorset House,

Stamford Street, London S.E.1.

365 www.americanradiohistory.com

A

Segmented

Fit Squaring

Circuit

A design technique for developing

a

circuit with an accurate square

-law transfer function.

by

R.

J. Lamden,

M.Sc., M.I.E.E.

A circuit with an accurate square law transfer function, that is,

Vo

= k

Vt2

(1) when k is a constant and

V

{,

Vo are the input and output voltages, has many important uses.

For example, this type of response is required for the measurement of mean power levels. Often circuits using square -law diodes or lamp- brightness measurements are used, but these are of uncertain accuracy and liable to variations with temperature and time. Moving -coil wattmeters are low- frequency devices and assume that their input waveforms are sinusoidal. circuit, a

But given an accurate squaring number of other possibilities appear.

For instance, two voltages can be multiplied in analogue fashion.

If the voltages are

VA,

VB, they can be added and subtracted in operational amplifiers, and a voltage proportional to the formed using the equation, product can be

4VAVB

=

(VA

+

Va)2

-

(VA

-

VB)2

...

(2)

Square -root function can be obtained using feedback techniques, and many functions may be approximated by using the first few terms of expansion their Taylor series

J(V)

=

Ao

+

A1V

+

AZV2 A3V3

. .

.

(3)

As A. E. Crump has recently pointed out',

R.

J.

Lamden did post-graduate research in radio astronomy at Man- chester University before joining

Ferranti to work on aircraft radar.

Moving to

U.K.A.E.A. in

7959, he now works on detection of earthquakes and nuclear explosions. any non -linear single valued function can be approximated by a series of linear segments.

The analysis of the squaring circuit is particularly simple, and the accuracy of the approximations can be calculated without recourse to graphical matching of the curve by straight -line segments.

The way in which a segment of the transfer function can be approximated is shown in Fig. t. For purposes of explana- tion suppose that the curve is to be matched

Vo

=

V12

(4) for

Vi between

-t

and

+

t volt; the alterations necessary for any magnification of this range can easily be obtained. maximum error is to be

Em then, from

Fig.

t,

this is

If the the amount by which the segment is below the parabola at its begin- ning and end. If we suppose that this is to be the nth segment of an approximation, the diode break -points will be called

Vi

(n

-1)

and

Vi

(n).

Suppose that the straight line segment follows the characteristic

Vo=AnVt

-}-

Vc

. .

.

(5)

Then at input voltage

Vi the error will be

E

=

Vo

-

AnVV

-Ve

=

Vi2

-

AnVi

-

Ve

...

(6)

This equation is also that of a parabola, and it is plotted in Fig.

2.

The graph has been made more recognisable by re- labelling the input voltage co- ordinate,

Vi'

=

Vi

-

2n

(7)

Then

E

=

Vt2

-

2

A4

+

Vn)

. . .

(8)

Fig.

2 shows also that the maximum error is present, at the centre of the segment,

Vi'

=

0, in the opposite sense to at the ends. So from equation

(8) the error

Em

= z

+

Ve

.

. .

(9)

366 or rewriting equation

(8)

E

=

V

{'2

-Em

....

(10)

If we apply this equation to the ends of the segment,

V1, from equation segment is

V where

E

=

Em, we see, to that the length of the

V4-

V1

=2

...

(11)

This shows that for a fixed maximum error, a parabola will be fitted by segments with a constant spacing between the break- point voltages. This is true for all except the first segment. Here the usual require- ment is that there shall be no output for no vo

V;

Fig.

1.

A single segment.

Fig.

2.

The error curve.

+EM

E vi v,

V4

Wireless World, October

1968 www.americanradiohistory.com

input, that is, zero standing error.

So, referring to Fig. 2, this segment must stretch over an input range

V2 to

V4, since the error must be zero at an input voltage of

V

21. shows

Substituting again in equation io that the length of this segment will be

AVI

=

17Ém(1

1/2)

..

.

(12)

So to cover t with the input voltage range from

0 to

N segments,

4

Segments,

Break -Point

Voltages.

No.

2

3

TABLE

I

Input

Range

0.5 volts.

Voltage

±1. io8

+2.405

+3.703

AVn

=

1 . . .

.

(13)

Or

Or

1

=1(1+17)+21/2(N-1)

(14a)

. . .

-

1

+V2(2N

-1) .

.

(14b)

Using this equation the number of seg- ments needed for a given accuracy can be found, and Fig.

3 shows the result graphic- ally.

This graph neglects the effects of diode imperfections, and even apart from the errors inherent in adjusting large numbers of break -point potentiometers, the assump- tion involved would probably fail for accuracies better than

o

i %. An interesting point of this graph is the quite good accuracy obtainable with a few segments. Since the first segment goes through zero, and the segments join each other, the standing voltage

Vo for each segment, is automatically inserted by the effects of the earlier segments.

The gain for the nth segment can be found from the equations

An

-

Vo

Vt

(n)

(n)

(

Vs2

(n)

-

Vo

(n

-

1)

-

V

(n

-

1)

-

Em)

Vi

(n)

=

Vi

(n)

+

V{

(n

-

(Vi

(n-1)2

-

Em)

-Vi

(n-1)

...(14a)

-

1)

(14b)

=

1

1

+

2 A/2

(n-1)

+

1/2 (2N

-

1)

(14c) when

V{

(n),

Vo (n) are the voltages at input the nth break -point. and output

Practical circuits based on this theory have been built and the outline details are shown in Fig.

4.

An accuracy of i

% of full scale was required, that is, a total output har- monic content more than 4o dB down on the second harmonic when the circuit is used for squaring a one

-volt peak sine -wave.

It is necessary to have the same for an output polarity input of either sense so a unity-gain inverting amplifier

A, was used.

Both conventional operational amplifiers and monolithic types have been used in both positions in this circuit. A capacitor C, has been inserted in the feedback path of the second amplifier to ensure an adequate stability margin.

The direct input is applied through diodes

D,,4,7,,0 while the inverted input

Wireless World, October

1968

0.1

Ill

5

Number

10 of segments

Fig.

3. A curve showing the accuracy of segmental fit.

Fig.

4.

A practical squaring circuit. the value of the feedbacks resistor

R, has been settled, by reference to equation roc.

For unity overall gain, if the operational amplifier gain can be considered very high, then with the nth diode conducting the gain will be n

An

=

(15)

Comparing

15 with 14c

R,

=

R5

1

+71/2

goes through

D,,6,9,12.

The break point voltages are fed through diodes

D2,5,8,11, which provide compensation for the effect of temperature on the diode forward

- conduction voltage, as pointed out by A. E.

Crump. also

This balancing out of the voltages means that the break- points can be set up by connecting an accurate voltmeter, most conveniently a digital type, to the sliders of the potentiometers, and adjusting these to the calculated break point voltages.

The effect of the diode non -linearity at the break -points, which although generally favourable are somewhat uncertain, can be reduced by making the input voltage swing large was in comparison with the non -linear region. An input voltage range of

±

5 volts chosen, since this is conveniently obtainable from both monolithic and dis- crete component operational amplifiers.

For an accuracy of better than i

% four segments are required, and the calculated break -point voltages are given in

Table i.

Gold- bonded germanium junction -diodes have been used to keep the switch -on region as small as possible; the effects of diode leakage- current can be reduced by making the input resistors as small as possible.

The values of these resistors can be found, once

R2

=

Rs

1

+71/2

41/

R

2

Using

=

R,

R, =

47 kS2 gives

RI

= ioo

= R, =

27o k f2.

These

kû,

values were made up from i % metal oxide resis- tors, and the potentiometer voltages were properly adjusted, using a digital voltmeter.

After these had been set to give an

RV, was adjusted output level of zero for zero input.

This is a circuit reliably capable of attaining accuracies better than i

% of full scale at any point on the characteristic, as a number of units using different amplifiers have shown. to at least

C,. io

It has kH2 a frequency response up with a suitable choice of

The long term stability is good, and the characteristics can be accurately reproduced.

REFERENCE

1

"Diode Function Generators

", by

A.

E.

Crump.

Wireless

World Dec. 1967.

Transistor

Usurps

Transmitter

Valve

An experimental solid -state device built in the U.S.A. has finally disposed of the super- old idea that transistdrs would never sede high power thermionic valves such as those used in transmitters. Developed by transistor of laminated con-

R.C.A., it is a struction capable of generating r.f. energy at

I

MHz with a power of 80o watts. device is

The formed from two separate silicon wafers, an emitter -base wafer and a base - collector wafer, which are subsequently fused, or laminated, under heat and pressure into a single structure and finally sealed in glass.

The construction makes accessible both an emitter contact area and a collector contact area for heat sinking.

367 www.americanradiohistory.com

Electronics

Artist and the

Cybernetic Serendipity, at Nash House, The Mall, London

S.W.1, is an international exhibition sponsored by the

Institute of Contemporary

Arts "exploring and demonstrating relation- ships between area the arts and technology ". Open until October

20th, it has, amongst its specific aims that of presenting

"an of activity which manifests artists' involvement with science, and the scientists' involvement with the arts ", and that of showing "links between the random systems employed by artists, composers, and poets, and those involved in the use cybernetic devices

". of

The term `cybernetic' refers to "systems of communication and control in complex electronic devices like computers, which have very definite similarities with the processes of com- munication, and control in the human nervous

-system ".

`Serendipity' here refers to

"the facility of making happy chance discoveries ". Thus we find amongst the exhibits cybernetic devices themselves, in which stimulus from the changing environment, human or machine, evokes a response in terms of sound, light or movement; and the products of computer systems in the shape

An element of line drawings, films, poems, etc. of superficiality and consequent naivety per- meates the exhibition. Upon entering the hall you are drawn into a magical scene and sparkles. of

It's all a bit like an amusement arcade on a wet afternoon at the seaside. prevailing gloom, strange noises, flashes

One of the first exhibits to be seen is rather like a one

-armed bandit combined with a

`what- the -butler -saw' machine. Clapping the hands and screaming causes pictures with pop-art motifs to change in rapid sequence under intermittent illumination. It is not very interesting and there's no pay-off so one moves on disgruntled. There are a number of devices that can be described as simple stimulus- and -response systems, besides the one just described. One with large blue lamps set in three rows responds to voices of different pitch. Different rows light up.

A more sophisticated unit incorporating a cathode -ray tube pro- duces ing diagonal striations of different colours. The most engag- of the sound -stimulated exhibits is a simple four -petal flower on a stalk of plastic tubes.

A high

-pitched sound emitted to one side causes the flower to swing round and face the source in a manner much like that of the little weed when talking to Bill and

Ben.

However, it will new in any already be apparent that nothing essentially respect has been described. When the technology behind the exhibits is reduced to block diagrams the results are very simple indeed. Thus, to talk about the "artists' involve- ment with science" is an inflated way of saying that an artist abandons certain tools and techniques, possibly because they are only associated with the immobile non -kinetic orthodox painting and sculpture, and reaches out for the dynamic possibilities of simple electronic art

-forms of and electro-mechanical feedback systems. In a curious way, fascination with the tools of a possible new means of expression has resulted in a

368 regression panel array of aesthetic expressiveness with the creation of oversize toys. Bruce Lacey, in his wall -fitting "Owl ", has used a of selenium cells to supply current to ruffle a small of feathers on each side of the creature's square body.

Two Crookes simple radiometers have become eyeballs: strikingly and equally ineffective. The prize for the most ingeni- ous non -happening must go to John Lifton for his "Analogue

Feedback Projection

System ", which was designed to involve the observer.

A rectangular image is projected on to the centre of a hemispherical screen, and secondary images oscillate towards and away from the centre.

Leaving aside the giant mobile, the numerous mechanical and electrical displays of harmonic relations, the programmed analogue computer reconstructing a drawing, and the com- puter poetry samples, we turn to the musical offerings.

A computer is described as producing music by probability and chance.

By definition the sounds being generated are random, having no points for harmonic reference and no sequential pattern.

Very different however is an experimental suite for string quartet, composed by gramme by

Hiller and Isaacson a of computer according to

Illinois a pro-

University. `Experiment

4' in the Illiac

Suite is a very satisfying, closely written piece of counterpoint. It seems quite clear that this, composed as it was in 1957, is to nevertheless by far the most advanced contribution the exhibition.

Whatever new art works result from the impact of technolo- gy on the artist, this exhibition can hardly claim to present any point for expensive. further growth. Elaborate cybernetic systems are very

They can be large, especially components are used extensively. Such systems do not do new things all on their own, but require careful programming and maintenance. Last but not in any way least there is a require- ment for new forms readable form, of if discrete electronic display. Music can be transcribed into but how can other electronic signal patterns, obeying newly conceived rules, be displayed other than by flashing lamps or on large colour cathode -ray tubes? The artist

-engineer of the future is certainly not going to be short of challenging problems.

J.

G.

Books Received

A

Guide to

Amateur

Radio, twelfth edition, by J. P. Hawker is

"intended to assist the newcomer to learn more about the hobby, and to help him or her to obtain a selection transmitting licence ". Each chapter is thick with information.

A of questions that any novice might ask is printed in the first chapter and given clear and encouraging answers. In this way the confidence of a would

-be amateur is built up. Following chapters introduce the reader to more and more general facts along with details munication receivers and transmitters. and frequency bands for

Ss. Radio Society

Valves still of circuitry predominate in of com- the more powerful designs but several transistor receiver circuits are described with full constructional details.

A chapter is devoted to explaining the licence examination, and several useful tables appear, including technical abbrevia- tion and symbols, radio circuit symbols, amateur abbreviations, morse code, amateur transmission and reception.

Pp. 86. Price of

Great Britain,

28

Little

Russell Street, London, W.C.1. a b c's of Radio and

TV

Broadcasting by F. J. Waters is a basic non -mathematical survey of transmitter equipment and its operation. The opening chapter explains the fundamentals of electromagnetic radiation, ending with a block diagram of a superhet receiver. Audio and video modulating signals, and the electronic and electrical devices for their production, amplification, and recording, are considered in the second and third chapters. The origin, amplification and modulation of radio-frequency carrier waves is investigated and explanations are accompanied by very frequent and simple diagrams. The last third supplies, transmission lines, broadcasting and receiving aerials (considered functionally), and remote transmitter of the book deals with power operating. Each chapter ends with several review questions, the answers to which are at the end of the book. The index allows the book to be used as an encyclopaedia. Pp.128.

W.

Foulsham

&

Co.

Ltd., Slough, Bucks.

Price

£1.

Wireless World, October

1968 www.americanradiohistory.com

Letters to the Editor

The

Editor

does not

necessarily endorse

opinions

expressed

by his

correspondents

I

Crosshatch and dot

read with considerable

generator

interest the article in the September issue describing an integrated circuit pattern generator, as I have made one very similar in design. However, I have achieved the same result in a more economical way and think that your readers may be interested.

A saving of some

33% of the cost of integrated circuits can be made by using

Motorola dual -in -line MC790P and MC724P in place of

2 X

µL923s and

2

X

4914s

respectively as they are electrically inter- changeable. to

9,

The first simplification to the generator is remove the buffers between the i.cs

6,

7 and and likewise 10, 11 and

13 and to use the form of counter shown in the drawing. This does not require any gating but it will not produce a single output pulse like the dividers in the article but a two to three mark -space ratio pulse. A divide

-by -five stage like this will drive an identical divider without any buffer between. I.C.14 can be eliminated in the article by using a two-input

AND gate

(fed from the 0-outputs of i.cs 9 and

13).

Also some means must be provided to ensure that the counters start from the zero position.

In my version a four input

AND gate is re- quired in order to achieve a one -line -wide pulse every twenty-five lines due to the odd wave- forms.

This could be a 4

-input integrated circuit or two

2- inputs in parallel, but it is much cheaper to use four diodes and a resistor. as

Finally, the output logic can be simplified shown to include crosshatch/dot and black/ white switching with simple switches.

Thus the complete generator now uses only. six 14L923s and three µí.914s

-or

three

MC790Ps and

2f

MC724Ps and a few extra discrete components.

Two possible further simplifications would be to remove the half

4,914 feeding the dividers as in my version, and also the buffer after the multivibrator by making the multi

- vibrator generate the correct mark -space ratio.

However, this short time constant is right on the upper limit of oscillation and is not really reliable even with selected µL914s.

A modification can be performed to my counter chain to enable it to be used for 405 - line working, this is shown in the drawing; the counter divides by

15 to give 27 horizontal lines

-some

25 being visible.

This could be arranged on a switch as shown for dual stan- dard 405/625.

For dots only, the input to the multivibrator can be the

1

-line wide output of the divider chain, so that the m.v. gives dots directly with no further gating!

A. W. CRITCHLEY

Uxbridge, Middx.

4.7k

Input

3 eV

0Dot

4.7k

220p

220p

22k

t

_J

3.6V

L

3x

µL914

47k

24

=

Dt

T

YT

1

D4

625

CG80-H

405 o

)

T

,-)PT

1 s

0405

625

Waveforms:-

5 5

-

=3

D21

3

I

Output

1

2

J

4-

D3 the same but x5 in size

4-

D4 the same but x5 in size

L

D3 (+3)

F2M11

Wireless World, October 1968

3

G

MC790P or

µL923 www.americanradiohistory.com

"Long- tailed

pair"

In his helpful article `Balanced

Transistor

D.C. Amplifiers' in your August issue, T.

D.

Towers says that the long -tailed pair circuit using transistors is sometimes known as the

Slaughter circuit, and then proceeds himself to use that name. this. Merely

I can see no justification for substituting transistors for valves in 1956 is hardly an achievement that deserves special recognition. The credit should go to the inventor of the long -tailed pair circuit, who was not F. F. Offner in

1937 as stated by

Mr Towers, but

A.

D.

Blumlein in his patent

482,740 of 1936.

The name

`

Blumlein circuit' would not do, because it would apply to about half the really important basic circuits we use, so

I urge that his own very apt name, `long -tailed pair', should be the one used for it, whether in valve or transistor version. We already have the in- justice of the name Miller (instead of Blumlein) integrator, which

I have been doing my best for the last 22 years to so fond correct'

.

Why are we of giving to Americans the credit for

British inventions?

E.g., W.W., June 1946, p.

403.

M. G.

SCROGGIE

Bromley, Kent

F.E.T. a misnomer?

It is regretted that the printers made nonsense of equation

(3) in Mr.

Hart's letter on p.

304 of the September issue. Here is the correct expression:

Ca

(t)rcA2

»

/aI\

(302

Sb

369

(3)

Switched -mode

D.C.

Regulators

How semiconductors are used to obtain

a

constant voltage from an unregulated supply

by

periodic switching of

a

series pass transistor

by

T.

D. Towers,

*

M.B.E.,

M.A.

Last month we looked at voltage d.c.

"linear" constant supplies, in which, in

- effect, transistors are used as feedback

-controlled variable resistors continuously dissipating exactly the amount of power necessary to keep the varying output voltage constant under input voltage and output load con- ditions. When it comes to high volt -amp requirements or high step -down voltage between input and output, the major problem of design in linear regulators becomes largely a mechanical one of providing an adequate heat sink for the dissipating transistors. This is the same problem that faces designers of high power audio amplifiers, who have been turning in recent years to class D (pulse- width- modulated) designs.

A similar trend is developing in d.c. power supplies, and this article gives some account of the basic circuits that have been used in discrete

- component, and more recently, in micro- circuit designs.

Switching regulators can be more com- pact than equivalent linear ones because of the reduction in heat -sink size for the same output power. To the higher efficiency of the switching mode we can add the use of economical wide tolerance semiconductors.

Finally switching regulators are of para- mount importance in battery driven equip- ment, particularly where the required is much lower than the supply voltage. output voltage

They have found a special place in missiles and space vehicles where power has to be carefully conserved.

On the other side of the coin, the greater heat -sink simplicity has to be paid for with greater circuit complexity. In general, also, switching regulators do not suppress input ripple as well as linear ones. Also, the feed- back of switching transient spikes into the supply rail can sometimes present quite a problem. Finally in the switching regulator the response to load transients (sudden changes in load current) tends to be slower than in the linear case. repetition frequency and passing the result- ant

"chopped" voltage through a low -pass filter.

A d.c. output of current,

'OUT, at voltage, VouT, calls output

POUT for a continuous power

=

VOUT

X

'OUT.

If we assume lossless switching and filtering, this means that the input line must supply the same average power. Now the source supplies input voltage current

(and therefore power) only during the "on" period,

ToN, while the load takes current (and power) during the whole period

ToN

+

TOFF.

If'IN

is the mean input current during the

"on" time, then

PIN

=

VIN x

TIN is the input energy during

TON.

Now during

TOFF the series switch is cut off and no current

(or power) is drawn from the input.

Thus

VIN ¡IN is the mean energy taken from the source over the whole switching period and must be equal to VouT x

/OUT.

With a suitable design of filter for low output ripple, the input current during

TON can be made relatively

IIN constant and equal to the constant output current,

'ouT, i.e.

=

IOUT.

Thus the input energy in one period

ToN

VIN

X

IouT x

TON must equal the out- put energy VouT

X

LOUT

(TON

From this,

VoUT

+

TOFF, given by

+

TOFF).

=

VIN TONI(TON

+

TOFF), as the illustrative waveforms in Fig. i(a) show. By the arrangement of

Fig. i(a) we can therefore, by suitable selection of the switching mark -space ratio, set the output voltage at any level from

OV up to

VIN.

However, the output level will vary if

VIN varies, which can happen either from independent variation of the input line voltage or from changes in the current taken by the load.

The output voltage can be regulated by the circuit arrangement of Fig. i(b). Here a feedback circuit samples the means of a output voltage and compares it with a reference voltage by comparator circuit. In the comparator, the difference (error signal) between the two voltages produces a control signal which is fed into the pulse

Fig.

I.

Basic blocks illustrating operation principle of switched d.c. voltage regulators.

(a)

Producing lower output voltage by periodically switching input through to output load via smoothing filter;

(b) additional regulating of output voltage by controlling

"on" time of series switch by comparing output with fixed reference.

Ton

Toff

Low -pass integrating filter

(a)

Vin

Series controlled switch

Unregulated d.c. nput supply

Fixed

- frequency pulse generator

1

Pulse drive

`Ton

/Toff

--

'cat_

Vout e Vinl

Ton on

+Toff

Load

Switched voltage

regulator principle

(b)

Series controlled switch

Low -pass integrating filter

Fig. t(a) shows the basic arrangement for obtaining a low voltage outpu from at higher voltage input by switching on and off a series controlled switch at a fixed pulse

Newmarket Transistors Ltd.

370

Unregulated d.c. input supply

Variable width pu se generator

I

Voltage reference

Comparator controlling pu se generator

I

4-

Output voltage sampler

Wireless

Load

T

World, October 1968 www.americanradiohistory.com

generator controlling the series

-switch.

The feedback signal from the comparator controls the on -off duty cycle of the pulse generator in such a way that if the output voltage across the load tends to fall below the preset level, the duration of the of the pulses increases.

"on" time

The series

-pass switch then conducts for longer periods of time, and the output voltage desired level.

Equally, if the returns to the output voltage tends to rise, the feedback shortens the

"on" time of the pulse generator to reduce the output leveL

Later we will examine the various circuit blocks in detail, but at this point the main advantage of the switched regulator over the linear regulator can be seen.

The series switch, nowadays usually a transistor, operates either fully cut -off or fully saturated

(

"bottomed "). Both of these states are low dissipation, so that dissipation losses in the series element are very much lower than in the linear regulators discussed last month. For any required output power from the regulator, therefore, much less heat -sinking provision is required for the transistor, permitting much more compact equipment design.

Low

-pass

filter

design

The low -pass filter to ensure a smooth d.c. in the is a output line of the switched regulator critical part of the circuitry.

In the interests of efficiency, no lossy elements should be included, and this rules out a simple RC filter.

Common practice is to use an

LC filter of which various forms are shown in Fig.

2.

The actual arrangement depends on whether p -n -p or n -p -n transistors are used as the series switch element.

The configurations of Fig.

2 are for n

-p

-n only, as the tendency nowadays is to use silicon power transistors (which are most commonly n

-p -n) and the configurations for p -n -p can be derived simply by reversing the d.c. polarities.

In

Fig. 2(a) the load, fed from

VouT, is in the emitter circuit of the series transistor.

The low

-pass filter has three essential elements, L,

C and D.

The diode, D, is to protect the transistor against destructive reverse voltage spikes that may be produced when the transistor rapidly switches off the current flowing through the inductance, L.

This diode acts as an energy

-recovery diode, transferring the energy stored in the inductor during the

"on" time of the transistor into the load when the transistor is turned off.

It operates much the same as the efficiency or damping diode in a tele- vision line output stage.

It substantially increases the efficiency of the circuit, and prevents the transistor emitter voltage rising materially above the positive supply line when

Because the transistor switches off. of the fast rise and fall times involved, the diode must be a fast switching one to perform the required function.

As to the LC components of the filter, the inductance is usually designed to be large enough to ensure that the stored energy, 0.5LI2, is sufficient to provide a relatively constant current flow into the load under maximum load conditions when the

Wireless World,

October

1968

Fig.

2.

Types of filters used after series switching transistor in switched d.c. voltage regulators:

(a)

In emitter circuit with choke in positive lead; (b) as

(a) but with choke in negative lead;

(c) in collector circuit with choke in negative lead; (d) as

(c) but with choke in positive lead. series is switch is off, and the capacitance, C, chosen to be large enough to provide the constant output under light load conditions when the energy in the inductance would be too low on its own. A low

L/C ratio is aimed at to reduce output voltage transients and to reduce the supply voltage transients arising from sudden changes in load current. Typical values of

L in practice are

O.5-

5.omH. With the pulse repetition rates of

3o

-iooHz commonly used, a value of

50- r000,F will be found used for C.

More specifically, to keep output ripple low, C is

1/(4n2 usually chosen much greater than f2L), where

f

is the switching pulse repetition frequency.

As for L, to keep the input current II

N relatively constant during

TON,

L is selected to make dIIN small in the formula di

/N

=

TON

(VIN

-

VOUT)/L.

The arrangement of the filter elements in the emitter circuit of the switching transistor shown in Fig. 2(a) may in practice be varied as shown in Figs. 2(b)

-(d), all configurations being shown for positive rail.

In

Fig. 2(b) the inductance is transferred to the earth rail.

In Figs.

2(c) and

(d), the filter is connected in the collector circuit of the transistor, with the inductance in the negative and positive rails respectively.

Series switch

transistor requirements

There are some very special requirements for the series -pass transistor as it has to switch substantial currents at repetition frequencies, fR, up to rooHz without undue internal losses.

Firstly the transistor should have a high current gain at the maximum load current, so that there are no substantial losses in its base

-emitter circuit. Secondly it should have an open -circuit -base avalanche voltage rating higher than the input supply voltage.

Thirdly it should have low reverse collector leakage current, to avoid undue dissipation in the cut -off condition. should have

Fourthly it should have a low collector- emitter saturation voltage in the

"on" condition, again to keep internal dissipation losses low.

Finally, it

"on" and "off" switching times short compared with the minimum

"on" pulse width

[typically r

/(rofI?)]; i.e. switching times should be of the order of r

/(roofR)

(=2oons for a 5oHz drive).

When you add up all the above require- ments, you find that only silicon diffused transistors are really suitable. Germanium alloy transistors cannot switch fast enough and have too high a leakage.

Variable

mark

-space

ratio

pulse

generators

Referring to Fig. r(b), we see that the output voltage is regulated by varying the on -off time of the series pass element. achieved by varying

This is the on -off time of the variable mark -space ratio pulse generator driving the series element, the ratio being controlled by a d.c. feedback signal from the comparator stage.

There are many possible circuits for constant -frequency pulse generators with the mark -space ratio con- trolled by a d.c. signal. Fig.

3 illustrates a few typical arrangements.

Fig.

3(a) is a unijunction RC relaxation oscillator, the added via output of which is resistively

R, with the external control voltage,

VCONTROL

When the combined voltage at its the base of the 2N13o8 exceeds forward base

-emitter starting voltage, the transistor turns on and produces a square wave output. The p.r.f. is set by the

CR3 time constant, and the mark space ratio can be varied from o to roo

°o by varying

VCONTROL from o to

12V. R3 should be kept much smaller than

R2 to keep the p.r.f. independent of the mark -space ratio, and

R, should be made equal to

R2.

In

Fig. 3(b), we find a different approach.

Tr, is a fixed- frequency

LC feedback oscillator, the via winding output from which is applied

L, to the emitter of a modulator transistor,

2'r3.

A control voltage,

VcONTROL, at the base governs via Tr2 the d.c. level of Tr,. When

VCONTROL is high, the voltage at the base of

Tr, is low and the transistor switches on early in the cycle, producing a high mark -space ratio in its output. When

VCONTROL is low, cor- respondingly the mark -space ratio is low.

Another approach to a d.c. controlled mark

-space ratio is given in Fig. 3(c).

Here,

Tr,,

Tr. is an astable multivibrator, with

C1 and

C2 its timing capacitors.

The usual base -drive timing resistors have been replaced by the output resistances of the collectors of transistors

Tr4, Try.

These output resistances are controlled by base drive from resistors

R17

R, in the collectors

371 www.americanradiohistory.com

111,414.

(a)

(b)

(c)

+Vcontrol frequency input drive, thus varying the mark -space ratio of the resultant output, while keeping the p.r.f. constant.

Feedback

control

loop

The remaining elements in the basic circuit of

Fig. t(b) form the feedback loop control comprising a circuit sampling the output voltage, a voltage reference and a com- parator circuit.

Usually the sampler is merely a potentiometer across the output and the voltage reference a zener diode.

The comparator is often some variant of a long -tailed pair of which the basic circuit is shown in Fig.

4(a). voltage phase with the

In this the control from the collector of

Tr1, moves in output voltage change as sampled by

R

Rs.

Where the 2mV /degC temperature variation of Tr, and Tr, need not be balanced out, we find a simpler single transistor comparator as in Fig. 4(b). Here the emitter of the transistor Tr is clamped to the zener diode reference voltage, and, if the output voltage tends to rise, the control voltage from the collector of Tr tends to fall and controls the oscillator mark -space ratio to bring the output voltage back to the regulated level.

12V

Cl

0.1

C3

100p

Fixed frequency input

<6V p

-p

R2

22k

Tr

Output

(d)

,

R1

100k

Vcontrol

(-121+12V) oV

+12V

Fig.

3.

Examples of constant frequency pulse generators with mark -space ratio controlled by d.c. signal

(a)

Unijunction oscillator; (b)

(c)

Astable multivibrator with long

LC

-tail oscillator with control pair;

(d) output modulator transistor;

Schmitt trigger control with fixed frequency input.

( b)

372

D. C. control to variable ratio oscillator

(a)

Control voltage in opposite change phase to output voltage

Vcc

Vout

R3 of the long tailed pair,

Tr1,

Tr,.

The base of

Tr, is held at a fixed potential by the potentiometer

R4, R5. A variable d.c. voltage applied to the base of

Tr1 alters collector voltages of directions.

Tr, and Tr, in the opposite

This in turn varies the output currents of

Tr4 and Trs in opposite directions while keeping the total constant.

As a result, the astable multivibrator,

Tr

Tr, keeps a constant p.r.f., while its mark

- space ratio changes.

A final space example of a d.c.- controlled mark

- ratio pulse generator uses a fixed input frequency from an external source as shown in Fig. 3(d).

In this circuit, varying

VCONTROL varies the switch -on point of the

Schmitt trigger

Tr1,

Tr, under the fixed

Fig.

4. Basic comparator circuits

(a)

Long - tail pair which feeds back a voltage to control series -pass transistor switching mark -space ratio to keep output voltage constant

(b) Simpler single

-transistor version.

Practical

switched

regulator circuits

One practical design for a switched regulator

(by Texas Instruments Ltd.) capable of giving a regulated

2oV

(G ±

0.2V) for o to

5A

60V output from an unregulated

3o to input is given in Fig.

5(a) as an illustra- tion of the design points discussed above, using germanium transistors.

The series -pass transistor,

Tr

is a

2N19o7 alloy

-diffused p

-n

-p power device capable of switching sufficiently fast at the relatively low p.r.f. of tokHz adopted.

The on /off drive to

Tr1 is provided by its

Tr, with input buffer Tr,. Tr, is driven in turn by the left hand collector of the astable multi

- vibrator

Tr4,

Tr4'. The multivibrator mark -space ratio is controlled from the collectors of the long -tailed pair comparator transistors

Tre, Tre' via the base- current- controlling transistors Trs, Trs'. In the comparator

Tre,

Tre', one base is tied through the potentiometer

R13, R14 to a oxed reference voltage zener other base samples the output voltage via the potentiometer

R19

-

R

V.

ZD,, while the

The low pass filter

D,,

L, C, between the switch transistor and the output is a p -n -p version of

Fig. 2(c).

Fig. 5(b) illustrates a recent design (by

Mullard Ltd.) using modern faster diffused silicon transistors in a circuit with a switch- ing rate of 4okHz with an output capable of being preset between

25 and

3oV at 5oW into the load for any input voltage between

4o and 65V with a regulation of better than

±0.3V. The overall efficiency of the circuit is around

8o% with only 2.2W dissipated in the series transistor Tr,. The circuit uses the combination of an

LC oscillator (Tr4) with a variable mark -space modulator Tr, as already outlined in simplified form in Fig. 3(b) above. The turn

-on level of the modulator Tr, is

Wireless World, October 1968 www.americanradiohistory.com

Fig.

S.

Practical circuits of switched d.c. voltage regulators controlled

(a)

Multivibrator-

2oV /5A from

3o to

6oV input

(Texas) (b) LC oscillator controlled

25 to

30V /SoW from

40 to

65V input

(Mullard).

+vvio

Di

\

R1

'--33^-'

-

/

Tr1

2N1907

T

I

D2

XR

78

Z controlled through RV, and R. by the com- parator transistor

Tr5i whose emitter is clamped via

R

to a fixed zener voltage

ZD, and whose base is connected to the output via potentiometer

R15,

RV..

A

(a) buffer driver transistor,

Tr2, is interposed between the modulator

Tr3 and the series. switch

Tr,. The transistor

Tre provides a safeguard against accidental output short circuit. When a short circuit occurs,

Tr, turns hard on and biases

Tr3 so that the series transistor

Tr, is cut off.

The series transistor requires only a minimal heatsink of some t5 °C /W by contrast with a con- ventional linear regulator where a very large and efficient heatsink would be necessary for the same regulated output power capability.

Tr2

2N1720

D3

1N751 vin s

R2

110

20W

R3

27ó

Tr3

2N1302

R4

680

R5

0511,

3W

1

I

56,42

Integrated circuit

switched

regulators

47k

22

R7

2.2k

Tr5

.,..

R9

2.2k

Trs

0.Ó01/c

Tr4

2N1304

Tr1

BD123

2 N

3055

Trá

2N1304

9II0.22/4

B

L4

2mii3O-2511

°

C oÓT

+0

A reader faced with designing a d.c. voltage regulator may find the account of the basic circuitry given above of some interest, but hitherto, because of the considerable circuit complexity, ninety -nine times out of a hundred he would fall back on the "good old" well

-tried linear regulator with its large

"hot plate" heatsink.

The coming of monolithic silicon integrated microcircuits has changed all this.

Fig.

6(a) sets out the circuit of

National Semiconductors Type the

LMioo monolithic voltage regulator, which in its entirety is fabricated in a square chip of silicon less than

otcm

square, and is comparable in size with a single silicon transistor.

Fig. 6(b) shows the external connections to the eight -lead TO5- outline can in which the circuit is encapsulated.

The single chip of the LMtoo contains the voltage reference, the feedback amplifier and other circuits necessary to form a switched regulator with the exception of the series

- pass choke transistor, the efficiency diode, the inductor, and the smoothing capacit- ance.

The availability of such prepacked complex circuitry must mean that in the future we will see many more designers turning to switching regulators.

In the circuit of Fig. 6(a),

ZD, is the basic zener voltage reference, current- supplied from the lowest collector of

Tr2. Tr2i a single transistor with three independent collectors, reflects the wide capabilities of the monolithic diffusion techniques now available.

The network Tr4,

Tr81 R11

R2,

Tr, is designed to convert the temperature dependent zener voltage of

ZD1 to a temperature- independent constant reference voltage of

1.8V at the base of Tr,. transistor pair,

The

Tr., and

Tr9 form the input

(b)

R4

820

R1

4i

3'w9k

IP

11

Tr,

3g

Tr4\

BCY70

L1

ZC2

R6

ToAL

180

(a)

Tr2

ZD1

Tr

Tr2

BFX

88

BFY52

C3

C4

2/c

R1

49k

Ra

72.2k

II

3W

Tr6

ACY39

Tr4

Tr6

RV1

4.7k

R9

2.2k

C5

0.0015/1

Tr5

ACY39

R5

20k

Tr5

rg

ZD vw

R3

1.2011,8W

11

120

R1;2

6.83

R1

2-2K

R4

14k

Trg

Rio

4.7k

R14

4.7k

ZD

BZY94-C12

Tr11

Tr1O

RV2

1k

Output voltage adjust

R

2k

R17

3.3k

R8

620

Tr12

R7

1-4k

1N2499

R2

15k

R19

3k

33

Unregulated input g

Current limit

8

Regulated output o Compensation

7

6

O

Feedback o

5

Reference bypass

Fig.

6. Some details of monolithic integrated circuit d.c. voltage regulator

(National

Semicondictor Type

LMioo). (a) Internal circuit; (b) Lead connections.

Pin connected

4 to case

Tr7

R3

22k

J

4 o Earth

Wireless World, October 1968 373 www.americanradiohistory.com

Fig.

7.

Basic principle of self -oscillating switching d.c. voltage regulator using monolithic i.c. stage of an operational amplifier, whose stage gain is collectors made high by using one of the of Tr, as a current- source load.

The output of this amplifier, from the collector of

Tr9, drives a compound emitter

- follower,

Tr the collector of and

Tr,,. The output from

Tr at pin

2 is available as the drive fur an external series switch transistor, which must be p

-n -p. transistor

Tr10 is used to limit the current drive from

Tr

The output to the external p

-n -p transistor to prevent overloading the integrated circuit. The current is set by an appropriate external resistor connected between terminal z

(current limit) and

2

(regulated output.)

In the remainder of the circuit of

Fig. 6(a),

Tr

Tr, and

Try provide bias stabilization for

Tr, to hold its collector currents at the required values. included to ensure

R

R5 and ZD, are that the regulator starts up on switch on.

Finally,

D1 is a clamp diode to keep

Tr, from saturating when it is switching.

Fig.

7 shows the principle of operation of the LM too as a switching d.c. regulator.

From the internal circuit details given above it will have been seen that the device itself comprises essentially an operational amplifier and a reference voltage. In Fig.

7, a reference voltage equal to the required output voltage is applied to one input of the operational amplifier, A, the output of which drives a separate switch transistor,

Tr,.

A resistive diVider

R, and

R,

(»R,)

provides small positive feedback at high frequencies to make the circuit oscillate.

At lower frequencies, the feedback is negative. When the circuit is first

VOUT is less turned on, than

VREF, and the switch transistor is turned on. When this happens, current flowing through R, raises the voltage on the non -inverting input of the operation- al amplifier to just above the reference voltage.

The circuit remains switched on until

VOUT rises to

VREF, when the amplifier goes into the active region and turns the switch transistor off.

The output voltage then drops back until the voltage on the amplifier non -inverting input returns to

VREF, when the amplifier switches on again. Hence the output voltage oscillates about the reference voltage, giving a ripple on the d.c.

The amplitude of this output ripple, being nearly equal to the voltage fed back through the attenuator

R

R2, can be made quite small.

Practical circuits

using monolithic switching

regulators

Fig. 8(a) a illustrates the use of the LMzoo as switching regulator to give a regulated output of

15V at oSA from an

18 to 40V input, using a small

TO5- outline silicon,

2N29o5A, p -n -p switching element. transistor as the series

In this arrangement, D,

L and

C1 comprise the usual clamped smoothing circuit, fed from the series switch

Fig.

8.

Typical practical self -oscillating switched monolithic

LMzoo microcircuit. switch

(a)

1SV,

05A from

18 to

40V using small, TOS- outline series transistor;

(b) roV,

3A from

13 to

40V voltage regulators built round using.T03- outline power transistor. transistor Tr,. Feedback to the inverting input of the operational amplifier (Pin

6) is obtained via a resistive divider,

R

R,, from the output. The resistance values shown are to regulated provide the required z5V output, but with different resistance values the output can be set any- where between

2 and

3oV. resistor which limits the

R, is the output drive current from pin

2.

R

connected from the output to the approximate zk

S2 input resistance of the reference terminal

5

[see Fig. 6(a)), corresponds to

R, in

Fig.

7 and provides the necessary feedback for oscillation. C, is an extra capacitor to minimise the output ripple by causing the full ripple. to appear on the feedback terminal. capacitor,

The remaining

C removes the fast -rise time transients which would otherwise be coupled into pin

5

R4. through the shunt capacitance of

This circuit can give output currents up to 50omA, the limit being set by the selected value of the protective resistor R,.

The optimum switching frequency lies somewhere between

3o and zookHz.

Fig. 8(b) shows a ment of higher powered arrange- the LM zoo as a switched regulator set to give

13 is toV output at up to 3A from a to 40V input.

Basically the arrangement the same as

Fig. 8(a), except that the single p

-n -p 2N29o5A is replaced by a composite connection of a

2N29o5A driving an n -p

-n

2N3o55 power transistor to produce in effect a very high gain p

-n -p power transistor. The capacitor C, across the input is to help to prevent switching transients being fed back into the supply line.

Many other circuit arrangements of the

LMzoo are possible, such as its use in a regulator with external fixed frequency drive, with separate current limiting, or with continuous

Any interested reader should consult a copy of an Applications short

Note, circuit protection.

"Designing Switch- ing Regulators" by R. J. Widlar, available from National Semiconductor Corporation,

Santa Clara, California.

(a)

(b)

374

13

-40V

2N3055

*

Mounted on small heatsink www.americanradiohistory.com

Satellite switch -off

After more than three years of successful operation the three Pegasus satellites laun- ched by the National Aeronautics

&

Space

Administration in

1965 have been turned off.

The primary task to study the density of these satellites of near -earth was meteo- roids, a meteoroid strike being registered by the discharge óf a capacitor detection panel.

For some months past the satellites, having fulfilled the collect primary role, have been used to engineering data in general and details of the reliability of electronic sub -assemblies in space in particular. It is understood that a good deal of valuable information on this later subject has been obtained. Initially the satellites had a design life eighteen of only months but were working perfectly right up until the switch off.

Wireless World, October

1968

Simplified

Circuits

Design of

Schmitt

Trigger

A method relying on the high gain and

low

leakage silicon transistors. of readily available

by

G.

E.

Marshman*

B.

Sc.

Using silicon transistors with high gains and low leakage currents, the design of many circuits become greatly simplified.

The design of a

Schmitt trigger is complicated and lengthy if the effect of base currents and leakage must be considered.

The con- siderable spread of these characteristics may also cause variations in performance if their effect is at all significant. A evolved which method will be reduces the effect of base currents to a minimum and permits rapid and simple design.

The Schmitt trigger circuit is drawn in

Fig. I(a) in its usual form.

In

Fig. i(b) it has been redrawn to show it as a long

-tailed pair amplifier, with positive feedback from the collector of Tr1 to the base of Tr,. state of the circuit depends on the voltage on the two bases.

The

The transistor whose base is more negative is turned on, and the other transistor off.

Consider the base of

Tr, to be at

0 volts.

Tr1 is set by

12, held off, and the base voltage of

Tr2 is the voltage divider chain

R1, R2 and

Emitter current flows through

R4 via the transistor into

R, the collector resistor.

Now imagine that the base voltage of

Tr1 is increased until it is very nearly equal to the base voltage of

Tr2.

This voltage is

Von.

At this point Tr, starts to turn on.

Emitter current flows via

R, and into the collector resistor

R1.

The resultant voltage drop across

R1 reduces the base voltage on

Tr, so switching the circuit into the other state.

When

Tr1 is turned off:

Tr2 base voltage

=

Von

R3

VecR3

Rt+R2+R3

cc

Von

- on

(R1

Rz)

(1)

The current through Tr,,

Ice, is given by: lee

R4

Von

-

Vbe

R4

Von

-

Vbe

Ice

(2)

"Muirhead

& Co..

Ltd.

Wireless World, October 1968

And the output voltage

Vo will be given by:

(3) is is

When the input reduced so the voltage is on

(and Tr, off) then as the such that

Tr1 input voltage current in

Tr1 decreases, raising its collector voltage.

This causes the base voltage of

Tr2 to rise also. When the input voltage reaches Voir, the base voltage of

Tr2 is also at Vert and

Tr1 will turn the circuit reverting to its other state. off,

Just before the circuit switches, the base voltage of

Tr1 is Vote and therefore the transistor current, will given by:

Ic

-

Voff

-

R4

Vbe

(4)

And

Tr2 base voltage

Woo

=

Voff

- leiR1)R3

Rt

+R2 +R3

R3

Vee

Voff

-

Ic1R1

-

Voff

(R1

+

R2)

X we

Comparing this equation with equation

(t) can see that: compared to the base current of

Tr2.

In most cases it will be satisfactory if we

R2 put equal to

R1. R3 may now be calculated from equation

(t).

From the design values of supply voltage, output voltage and current, and the two switching voltages, designed in the the circuit may now be following manner.

(a)

From equation

(2) determine

R4.

(b) From equation

(3) determine

R5.

(c)

From equation

(4) determine

Ic1.

From equation

(5) determine

R1.

(d)

Choose a value for

R2.

(e)

From equation (i) determine

R3.

Example

Using a the derived equations we will design trigger circuit with a required output of

4 volts at

I mA, a turn

-on voltage and a backlash of

2 volts.

The of

4.5 volts transistors used are

Mullard

BCY7o, minimum current gain at

I mA is 50.

Thus floc

The supply is 12 volts.

=

I2,

Vo

=

4,

Ice

=

I mA,

Von

=

4'5,

Voff

= 25.

Fig.

I

(a) Usual form of Schmitt trigger circuit.

(B) Trigger circuit redrawn as a long tailed pair amplifier with positive feedback.

Vcc

Von

Voff

-

Von Vcc

-

IcIRt

-

Voft

-

Votf)

Rl

VC e( Von

Vonlcl

In actual practice the circuit will switch when the voltage on

Tr1 base is within

o

I volt of the voltage on Tr2 base.

To compensate for this effect a figure of

o

I is added to the above expression:

R1

Vcc( Von

-

V011+01)

Vonlel

(5)

Equation

(i) states that:

R3

Vex

Von

(R1

+

R2)

R2 may now be given any value such the current in the divider chain that will be large

(b)

375 www.americanradiohistory.com

(a)

R4

-

Von

4'5

-

Vbe

!

c2

-

0.6 ki2

1

=

3.9 ki2

(b)

R5

Vo

= les

(c)

Ie1

=

4 kS2 say 3.9 kS2

=

Vof f

-

Vbe

2.5

R

4

-

0.6 mA

39

R1

=

0.48 mA

Yee( Von

-

Voft

+

0'1)

12(4'5

Von

/cl

-

2.5

+

0.1) kS2

4.5.(0'48)

=

11'6 kit say

12 kS2

(d)

Put

R2

=

12 kit

(e)

R3

=

Von

Von

(R1

+

R2)

12

4.5

-45

(12-x-

12)

=

14.4 kit say

15 kS2

16th

I.E.R.E.-

"Airborne collision avoidance sys- tems" by

S. S.

D. Jones at 18.00 at

9

Bedford Sq.,

W.C.1.

16th B.K.S.T.S.

-

Demonstrations of new professional equipment at

19.30 at the Royal Overseas League,

Park

Pl., St.

17th

James's St., S.W.1.

I.E.R.E.

&

I.E.E.-C

"System structures of modem computers" at 10.00 at Middx.

Hospital Medical School, Cleveland St.,

17th

W.I.

I.E.R.E.-

"P.C.M. for point -to-point music transmission" by E. R. Rout and

A.

H. Jones at 18.00 at

9

Bedford Sq., W.C.1.

17th Inst.

Electronics.-

"Teaching digital computers to technologists and technicians" by G. H. Stearman at

18.45 at London

School of

Hygiene and Tropical

Medi- cine, Keppel

17th

St., W.C.1.

R.T.S.

-"Ultrasonic delay lines for television" by R. W.

Gibson at 19.00 at I.T.A., 70 Brompton Road,

S.W.3.

22nd Soc. Relay Engineers.

-Symposium on "Ex- perence of colour TV on wired systems" at 14.30 at the

I.T.A., 70

23rd

Brompton Rd., S.W.3.

I.E.E.-

Discussion on "Electro-optics" at 17.30 at

Savoy PI., W.C.2.

23rd I.E.R.E.

-"Are examinations really necessary?" by Roy Cox at 18.00 at

9

Bedford Sq., W.C.1.

24th I.E.E.

-Coloquium on

"Transmission -line properties of interconnections and their measurement" at

14.30 at Savoy Pl., W.C.2.

24th I.E.R.E.

&

I.E.E.

-Colloquium on "Recent advances in the design. of biological amplifiers" at 14.30 at

9

Bedford Sq., W.C.1.

25th R.T.S.

-"A

new lightweight colour camera" by

S. C.

Tan at

19.00 at the I.T.A., 70 Brompton Rd.,

S.W.3.

28th I.E.E.

-Colloquium on

"Solid -state displays" at

14.30 at Savoy Pl., W.C.2.

29th I.E.E.

-Discussion on "Increasing the reliability of electrical and electronic equipment under high ambient temperature conditions" at

17.30 at Savoy Pl., W.C.2.

30th

I.E.E.-

"Science and profit in the electronics industry" by R. J. Clayton, Electronics Division chair- man, at 17.30 at Savoy Pl., W.C.2.

30th B.K.S.T.S.

-Discussion on "Why educational

TV?" at 19.30 at Royal Overseas

League, Park

Pl., St.

James's St., S.W.

1.

A trigger circuit was constructed to this design. the turn

The turn

-on

-off voltage voltage was 4.4 volts,

2.5 volts.

There was negligible change in performance up to an ambient temperature of

40°C.

Ó

ctober Meetings

Tickets aie required for some meetings: readers are advised, therefore, to communicate with the society condemned

LONDON

2nd

Tropical Medicine, Keppel St., W.C.1.

2nd B.K.S.T.S.

-"Sound in

Longshot ", Sympo- sium at 19.30 at the Royal Overseas

League, Park Pl.,

St. James's St., S.W.1.

7th.

I.E.E.T.E.

-"Lasers

-

light for the engineer" by

Dr. E. Eastwood at 18.00 at Savoy Pl., W.C.2.

9th I.E.E.

"Omega" by F.

Stringer at 17.30 at

Savoy Pl., W.C.2.

9th I.E.R.E. on logic

-"Large scale integration

-its

effects design" by K. J. Dean at 18.00 at

9

Bedford Sq.,

W.C.1.

11th R.T.S.

-Seminar on

"Television in educational technology

-what is its role ?" at 15.00 at I.E.E., Savoy

Pl., W.C.2.

14th I.E.E. sional

-"Propagation in a random one-dimen- medium" by

Dr. H. E. Rowe at 17.30 at Savoy Pl.,

W.C.2.

16th I.E.E.

-"Thin film transistors" by Prof. J. C.

Anderson at 17.30 at Savoy Pl., W.C.2.

376

S.E.R.T.

E. Dixon

-

"Colour television maintenance" by at

19.00 at the London School of Hygiene

&

BATH

23rd. I.E.R.E.

&

I.E.E.

-"Recent advances in mobile communications" by D.

A. S.

Dryborough at 19.00 at the Technical

College.

BELFAST

23rd. I.E.R.E.

-"Management, methods and media in electronic training at the School of Electronic En- gineering, R.E.M.E., Arborfield" by the Commandant,

Col. H. G. Frost, at 18.30

University, Stranmillis Rd. the Ashby Inst., Queen's

BIRMINGHAM

25th. S.E.R.T.

-"Colour television

W. J.

Anderson at 19.30

-the decoder" by, at the University of

Aston,

Gosta Green.

BOURNEMOUTH

31st. I.E.R.E.

-"Tunable coherent sources" by Dr.

E. L.

Thomas at 19.00 at the College of Technology.

BRISTOL

16th. I.E.R.E.

-"Integrated circuits, the present and future" by D. W. Roberts at 19.00 at the Technical

College.

CARDIFF

2nd.

R.T.S.-"A solid -state colour receiver" by S. C.

Jones, R. E. Gray and J. W. Bussell at 18.30 at Llandaff

Technical College, Western Avenue.

9th. I.E.R.E.

-"M.O.S.F.E.T. in integrated circuits" by Prof. W. Gosling at 18.30 at the University of Wales

Inst. of

Science and Technology.

11th. S.E.R.T.

-"Integrated circuits" by M. Williams at

19.00 at Llandaff Technical College, Western Ave- nue.

16th. R.T.S.

-The

1966

Fleming Memorial Lecture,

"The strange journey from retina to brain" by

Dr. R.

W. G.

Hunt at 19.00 at the B.B.C. Centre, Llandaff.

CHELMSFORD

7th. I.E.R.E.

&

I.E.E.

-"Electronics in medicine" by

Dr.

D. W. Hill at 18.30 at the Lion and Lamb Hotel,

Duke St.

DURHAM

16th. I.E.E.T.E.

-"Laser physics and developments" by Dr.

D.

Balfour at 19.30 at the University, Science

Labs., South Road.

EVESHAM

15th. I.E.R.E.

-"Pulse code modulation and its ap- plication to sound and television" by Dr. C. J. Dalton at

19.00 at the B.B.C. Club.

GLASGOW

18th. S.E.R.T.

-"Airport telecommunications" by

W. A. S. Aitken at 19.30

St., C.2. at the Y.M.CA.,

100

Bothwell

LEEDS

3rd.

I.E.R.E.

-"Digital logic and its application" by

P. McLennon at

19.00 at the University, Dept.

Electrical and Electronic Engineering. of

LEICESTER

8th I.E.R.E. tems in the oil

-"Telemetry and communications sys- industry" by

A. C. W. Bedwell at

18.30 at the University Physics Lecture

Theatre.

LIVERPOOL

16th. I.E.R.E.

-"Pulse code modulation" by P. Tyler at 19.00 at the University Dept. of Electrical Engineer- ing and Electronics.

MANCHESTER

17th.

I.E.R.E.-

"The use of thin engineering" by R. Naylor and

R.

Fairbank at 19.00 at

Renold Bldg., Inst. of Science and films in electronic

Technology, Altrin- cham

St.

31st. S.E.R.T.

-"Colour television servicing" by T.

M. Robinson at

20.00 at John Dalton College.

MIDDLESBROUGH

29th. S.E.R.T.

-"Microelectronics" by

T.

M. Ball at

19.30 at the Cleveland Scientific Inst., Corporation

Rd.

NEWCASTLE -UPON -TYNE

9th. I.E.R.E.

-"Special applications of lasers" by D.

Tuck at 18.00 at the Inst. of

Mining and Mechanical

Engrs.,

Neville

Hall, Westgate

Rd.

NEWPORT, I.O.W.

18th.

I.E.R.E.-

Discussion on

"Production manage- ment" at

18.30 at the Technical College.

READING

15th. I.E.R.E.

-"Frequency synthesizers" by K. R.

Thrower at 19.30 at J. J. Thomson Physical Lab, the

University.

SOUTHAMPTON

22nd. I.E.R.E.

-"Circuit design by computer" by E.

Wolfendale at 18.30 at the Lanchester Theatre, the

University.

SWINDON

15th. I.E.R.E.

& niques for

I.E.E.

-"Research on control tech- railways" by Dr.

L. L. Alston at

18.15 at the

College.

ESRO

1

Cleared for

Launching

Flight unit no.

1 of the ESRO

I satellite has been cleared for launching on

October 2,1968

-the

date decided upon several months ago. At a recent meeting attended by representatives of the European

Space Research Organization and the American National Aeronautics and

Space Administration the readiness state of the satellite and the launching and orbital control facilities was assessed and the go-ahead was given.

The satellite will carry eight experiments for the study of ionospheric and auroral phenomena which will measure the flux and energy of high- altitude particles. The satellite will be force so stabilised along the lines of magnetic that when over the northern polar region auroral photometers point towards earth and most of the particle detectors will be directed towards incoming particles.

U.K. experiments forming part of the pay- load include measurement of auroral lumin- osity (Queen's University, Belfast), ionospheric electron and positive -ion temperature measure- ment (University

College, London), and measurement of the electron and proton flux and energy spectrum from 40 to

400keV and

5 to 30

MeV (Radio and Space Research

Station).

Wireless World,

October

1968 www.americanradiohistory.com

Test

Your

Knowledge

Series devised

by

L.

Ibbotson,

*

B.Sc., A.Inst.P.,

M.I.E.E., M.I.E.R.E. to. Several sinusoidal waves are travellii simultaneously in both directions on a tran mission line.

At the same point on the li the voltage and current associated with a one wave have a

(a) always

Z.

(b) ratio which is

Z. unless one of the other waves is the same frequency

(c) generally not

Z.

(d)

Z. at some points on the line but n at others.

5.

Transmission lines

I.

A sinusoidal generator drives a travelling wave along a transmission line.

The phases of voltage and current each

(a) lag increasingly with distance from the generator

(b) lead increasingly with distance from the generator

(c) do not vary with distance

(d) have two values at different parts of the line.

2. A line, driven by a sinusoidal generator, as a result of its termination supports a total standing wave.

The phases of voltage and current on the line each

(a) lag increasingly with distance from the generator

(b) lead increasingly with distance from the generator

(c) do not vary with distance

(d) have two values at different parts of the line.

3 a

The wavelength of a sinusoidal signal on transmission line is

(a) directly proportional to the phase change coefficient

(b) inversely proportional to the phase change coefficient

(c) proportional to the phase change co- efficient

(d) plus a constant independent of the phase change co- efficient.

4.

For a sinusoidal input signal the characteristic impedance of a transmission line is to

(a) the impedance of a unit length of it

(b) the impedance which when connected the end of the line does not reflect back any energy

(c) the impedance seen at the input of the line when the

(d) output is short circuited the impedance seen at the input of the line when the output is open circuited.

5.

The characteristic impedance of a transmission line

(a) is always

(b) always component resistive has a significant reactive

*

West Ham College of

Technology, London,

E.ts.

Wireless World, October 1968

(c) is in general effectively complex at high frequencies and real at low frequencies

(d) is in general effectively real at high frequencies and complex at low fre- quencies.

6. A line carries a single travelling wave. indicates sinusoidal

The propagation constant

(a) the change of phase per unit length of the line

(b) the natural amplitudes unit log of the ratio of voltage distance apart on the line

(c) the natural log of the phasor ratio of voltages unit distance apart on the line

(d) the natural log of the ratio of signal powers unit distance apart on the line.

7.

If a transmission line is non -dispersive one of the following is independent of frequency:

(a) the phase change coefficient

(b)

(c)

(d) the attenuation coefficient the propagation constant the phase velocity.

II.

The input impedance of a transmissii line of length

I terminated by a load

ZL quoted in the textbooks as

Zo

¡

ZL

+

Z0 tanh y

I lZo ZL tanh y

A sinusoidal generator is switched to t input of the line at time t

=

O.

T impedance which the line presents the generator at t

=

0 is

:

(a)

(b) the above impedance

Z.

(c) ZL

(d) an impedance of a value whi depends on the phase of the generator t =O.

12.

On a mismatched transmission line t reflection coefficient at a given point defined as the complex ratio of voltage of t incident wave to voltage of the reflect wave.

The reflection coefficient is:

(a)

(h) real at all points complex at all points real at nodes and antinodes of the vo

(c) age standing wave

(d) pattern imaginary at nodes and antinodes the voltage standing wave pattern.

8.

The group velocity of a signal on a transmission line is:

(a) travels the velocity with which the energy

(b) travels the velocity with which the waveform

(c) the velocity of electrons in the con- ductors

(d) the velocity of sound in the con- ductors.

13.

On a mismatched transmission li the distance between adjacent standi wave

(a)

(b) nodes is: one wavelength half a wavelength

(c) a quarter of a wavelength

(d) an eighth of a wavelength

14.

On a mismatched lossy transmissi line

(a)

(b) there is no standing wave the voltage standing wave ratio small

(c) but uniform the v.s.w.r. is greatest near

1 generator

(d)

, the v.s.w.r. is greatest near the lo;

9.

For an ideal transmission line the equation

'32V

=

K b2

i2

x2 holds.

The us to deduce value of the constant K allows

(a) the characteristic impedance of the line

(b) the velocity of propagation of a signal on the line

(c) the largest amplitude of signal which the line can transmit

(d) the attenuation per unit length of line.

Is. A loss

-free transmission line if wal lengths long is terminated in a short circi

The input impedance is:

(a) zero

(b) infinite

(c)

Z,

(d) a small resistance.

Answers and comments, page

383

3 www.americanradiohistory.com

New

Products

QC 1062K filter shape factor at

21.4MHz has a similar but has been designed for systems having a bandwidth of

30kHz. The filters are housed in a hermetically sealed metal can and have a high resistance to vibration. Salford Electrical Instruments Ltd.,

Peel Works, Barton Lane,

Eccles, Man- chester.

WW302 for further details

High

Current Tantalum

Capacitors

Dage

(G.B.)

Ltd., are producing a series of miniature tantalum feed -through capacitors for high current and high frequency applica- tions.

The capacitors are capable of effective filtering beyond 1000MHz with low self inductance; and have a feed -through d.c. current rating of

5A at

85

°C. Each unit is hermetically sealed with glass- to-metal solder seal terminations. The leads are of tin and lead coated nickel, and can be soldered or welded. Voltage and capacity ratings extend from 60µF at

6V to 3.9,uF at 75V. The dissipa- tion factor at 120Hz is given as 6% maximum

41-

channel Marine Radio- telephone

The STR

60-A by

International Marine

Ra- dio Co, a 41- channel

20W marine radiotele- phone capable of working on the

156 -

161MHz marine v.h.f. band, is semi- conductor -built throughout. It is capable of simplex or duplex operation and employs a channel spacing of 50kHz.

A built

-in diplexer permits simultaneous transmission and reception using a single dipole aerial, with a loss in both transmit and receive directions has of only 0.7dB.

The basic version the transceiver and main control unit mounted in a cabinet (shown left in the photograph) measuring approximately

52 x

43 x 14cm.

A rack -mounting version with transceiver behind the control unit is avail- able. In addition, the transceiver may be operated from any one of up to four remote locations. Sub-control units for this purpose

(shown right in the photograph) measure approximately 36 x 18 x 11cm and are also available in rack versions.

A main and an auxiliary selector on the control unit allow for dual watch facilities. One channel is set up on each, and, with an associated switch set to

"dual watch

", the receiver dwells for

1

/10th second on the auxiliary and

9

/10th second on the main channel. When a signal is present on locks the auxiliary channel the receiver on to it and switching stops.

The receiver, with a sensitivity of 1.2µV for 20dB signal -to-noise ratio, and the transmitter are built in the Ministac system, giving high shock and vibration resistance and high re- liability, simple maintenance and minimum spares stocking.

A handset with the usual press

-to-talk button mutes the loudspeaker when taken off its rest. Provision is made for the connection of the STR

60 -A to the ship's telephone exchange. The transceiver con- sumes 130W from normal ship's supplies on full transmit power.

A special version de- signated -60 -T is available for tankers where the additional oil- refinery frequencies are required.

International Marine

Radio

Company

Ltd., Peall Road, Croydon, Sur- rey.

WW

321 for further details

Crystal Filters

Salford Electrical Instruments are now mar- keting quartz crystal filters with centre fre- quencies of

21.4 and

37.3

MHz. These units were developed primarily for the high -fre- quency communications market, as the present trend indicates a need for interme- diate frequency selectivity at frequencies at the higher end or outside the normal operat- ing range of the system.

The

QC 1062L standard filter at 37.3MHz, has a

6dB band- width of 12kHz and a 6

/60dB shape factor of

2:1. Another version is also available having an ultimate attenuation of

Spurious passbands in the stop

80dB.

-band are almost eliminated within the frequency range

± 1MHz of the centre frequency.

The

'; at

25

°C. The leads are positive and the cases negative. Dage

(G.B.)

Ltd.,

1

Penn Place,

Rickmansworth, Herts.

WW 329 for further details

Desk

Calculator

An easy -to-use desk calculator primarily intended for scientific applications has been announced by Hewlett Packard. The calculator will operate either in a fixed or floating point mode as required and has a capacity of

1

X

10-98 to

9.999999999

X

1099.

A full range of trigonometrical functions can be carried out by pushing single buttons, namely sin, cos, tan, arcsin, arcos, arctan in all four quadrants, a further single key stroke will convert the result from rectangular to polar co-ordinates or vice -versa.

The usual arithmetic keys, addition, subtraction, multiplication, division and square root

" are included as well as keys for log x,

In.x and

e.

A useful key among the group for numerical input is one that sets in the value of

77" .

Up to

196 keystrokes may be programmed by selecting

"Programme" and pressing the required keys sequentially, no code conversion is required. Programming keys available in- clude the functions if x>y, if x<y, if x

=y, if flag, set flag and go-to.

A programme, once assembled may be run automatically or step

- by -step for error detection, it may also be recorded on a small magnetic card on the

Wireless World, October

1968 www.americanradiohistory.com

amplified to give a signal proportional to the current. The membrane is driven at its resonant frequency by a signal at approxi- mately 6kHz to give an output

(d.c. verted to a.c.) at the same frequency. con-

The maximum overall length is

64.7mm and the maximum diameter

30.2mm.

Mullard Ltd.,

Mullard House, Torrington Place, London

W.C.1.

WW308 for further details. integral recorder for automatically reading

-in again for use of a later date. Peripheral equip- ment that will be available shortly include a printer and an xy plotter. The calculator costs under

,2,500.

Hewlett Packard, 224 Bath

Road, Slough,

Bucks.

WW

315 for further details

Switching

Transistors

Particularly suitable for switching circuits incorporating inductive loads are three n -p-n silicon planar epitaxial transistors, type

BSW66, BSW67 and BSW68 from Mullard.

The energy rating of these TO

-39 encapsulated transistors is 5mW.

They will all switch

1A.

The only difference between the transistors lies

VIBO in the supply voltage ratings. The and

VCEO is 100V, 120V and

150V for the BSW66, BSW67 and the

BSW68 respec- tively.

V is .6V and

VcE

O.15V.

Mullard

Mullard

House, forrington

Place,

London W.C.1.

WW

322 for further details.

Vibrating sitive

Capacitor for Sen-

Electrometers

Developed by type XL7900

Mullard, vibrating capacitor contains a membrane that can vibrate at high frequencies, and offers a new means of measuring small voltages and cur- rents. Electrometers using it have measured currents as small as

8 x 1017A, or

500 electrons per second. The

XL7900 contains four metal plates in being fixed. parallel, the two outer

The two inner plates are me- chanically linked so ween them is that the distance bet- constant. They can, however, vibrate as a rigid pair with respect to the outer plates. This movement is achieved by applying an alternating voltage between one of the vibratory plates and a fixed plate.

The d.c. supply is connected to plates, one fixed and one the other two vibratory.

Con- sequently, when the middle plate vibrates, the capacitance across the d.c. supply changes and modulates the direct voltage at the frequency of vibration. The alternating voltage so produced is easily amplified to produce a signal that is directly proportional to the voltage across the capacitor.

When measuring currents, a high resistance known value is connected across the of input capacitor of the XL7900. The voltage pro- duced across the resistance by the direct current through it is then modulated and

I.C. Mounting Board

A new circuit mounting card to to

6 dual

-in

-line (14- or

16

-lead) mount up integrated circuits now forms part of A.P.T's Lektrokit system.

3111, is

The new card, part number

LK- constructed of copper -clad epoxy glass laminate with printed power supply tracks to each i.c. It is fitted with

24 input

/output edge- connector terminations at

2.54mm

(0.lin) pitch and provision is made for the attachment of a nylon cord loop to assist card removal. Size of the card is 68mm

(2.7in) square and the price is 7s. 6d.

Cards are also available as part of the components for a complete

6

-way mounting framework.

A.P.T. Electronic Industries Ltd., Chertsey

Road, Byfleet, Surrey.

WW 320 for further details width, which is particularly advantageous when dealing with random noise. After squaring and smoothing, the result is plotted as spectral power density against frequency. Frequency accuracy is

±

1% over seven ranges, the lowest being 0.3 to

1.66Hz and the highest 4.69 to 26kHz

Smoothing times seconds are

1, 10, 25, 50 and

100 and integrator scaling is X1 and

X10. Fenlow Electronics Ltd., Springfield

Lane, Weybridge, Surrey.

WW311 for further details.

Modular Pulse Generating

System

Farnell Instruments Ltd., have produced an inexpensive generator which can be assembl- ed in either single or double output con- figuration. The modules are mounted side by side in a

19

-inch case. Power and signal interconnections appear in the form of patch- ing leads at the rear repetition frequency of the assembly.

A pulse of

1Hz to 10MHz is available with width and delay times variable between

0.1 cs and

1s.

The component mo- dules include a power supply full double pulse adequate for a generator, a trigger mo- dule with an output of

1 volt maximum up to

10MHz, a pulse repetition frequency generator, a delay generator, a pulse width unit, and an output amplifier giving

10V at an output impedance of

5051.

Farnell

In- struments Ltd., Sandbeck Way, Wetherby,

Yorkshire.

WW314 for further details.

Ferrite

Memory Stacks

Mullard offer two ferrite memory systems for use in process control equipment, machine tools and communications systems.

The first is a standard memory system, type MM1500,

Low

-frequency

Spectrum

Analyser

Fenlow Electronics have recently produced a low -frequency spectrum analyser type

SA4 incorporating integrated circuits. Operation- al amplifiers at the output provide a range of smoothing times and a true integration facility.

A neat panel layout allows any operation whether the instrument is used manually or with the Fenlow Automatic Plot- ter type MP1.

The analyser operates on a heterodyne principle using a homodyne detector with a low -pass filter to define the bandwidth of the instrument. This means that analysis is carried out at a constant band-

Wireless World, October 1968

379 www.americanradiohistory.com

with a capacity of 1024 words of

1 bit. It has a cycle time of

2 p s.

Operation is possible over the ambient temperature range 0 to + 50 °C.

The overall dimensions are approximately

9

X

28

X

13cm.

The second standard system, type MM1501 (shown in the photograph), has a cycle capacity of 1024 words of

8 bits. Its time is

4µs and access time

0.6,us.

Opera- tion is satisfactory over the ambient tempera- ture range

0 to

+

55°C.

The approximate dimensions are

7.5

X

10.75

X

20cm. Power requirements are

+6V at 2A and

-6V

at lA for the MM1500 and +

12V at 3A for the

MM1501.

Mullard Ltd., Mullard

House,

Torrington Place, London W.C.1.

WW 325 for further details aerial interests of these two companies. To be known as the dipole

Trucolour range, the folded of the new aerials is of

"fat" sectional dimension (relative to its length) which, it is claimed, is a design ponse and accurate feature giving even res- matching over a frequency band of at least 88MHz. Besides offering an increase in gain over existing types employing the same number of elements, the gain does not vary by more than

3dB over the band and by no more than

1dB over adjacent channels.

Trucolour aerials are designed to match a

75

-S2 feeder without the use of a balun and are available for channel groups

A(21

-34),

B(39 -51) and CD(49 -68). Aerials are supplied in assemblies of

6, 10,

13 or

18 elements and

10- or

18

-element broadside arrays. Prices range from

£l

17s to £7 17s.

Antiference Ltd.,

Aylesbury, Bucks.

WW 323 for further details

Modular Pulse

Generator

The

Advance

Instruments Division of

Ad- vance Electronics Ltd., has produced a new modular pulse generator, the PG52, which is assembled from the range of five signal generating and processing units available.

Repetition frequencies up to

20MHz, and output pulses of up to 20V into 50D with rise and fall times of

5ns can be obtained from the system, and its versatility enables complex pulse and ramp waveforms to be

Insulated Stand

WW317 for further details.

-offs

Thirty

-two new moulded diallyl stand -off insulators for mounting and supporting electronic equipment have been added to the range of stand

-offs by

Cambion Electronic

Products. Made in four families of mount- ing styles and a wide range of thread sizes, diameters and lengths, the new can be used in insulators temperatures up to 160 °C.

They provide a reliable way of supporting metal panels, separating boards and similar applications. Cambion

Electronic Products

Ltd., Castleton, near Sheffield, Yorkshire.

UI

'I'

_1

11

Ì produced.

The five plug -in units are all self- contained, and require only d.c. power supplies, which are obtained from the main frame. Signal paths are made externally, through

BNC /BNC connectors, to ensure maximum flexibility in interconnection. The majority of functions are selected by push- buttons. The output modules are protected against damage due to short -circuit. The clock generator uses integrated circuits, and silicon transistors are employed in all units.

Advance

Instruments, Hainault, Essex.

WW310 for further details.

U.H.F. Television Aerials

Antiference has introduced a new range of wideband u.h.f. television aerials which replaces all existing Antiference and Belling

-

Lee u.h.f. aerials following the fusion of the

Gate -protected M.O.S.Ts

Six p- channel enhancement type m.o.s.ts hav- ing a new type consisting of gate protection circuit of a diffused resistor-diode net- work, that protects each device gate from accidental damage due to voltage transients, are available from SGS

-Fairchild. The resis- tor coupled with the input capacitance forms an

RC network ing pulse so that slows down the incom- that the protection diode has time to break down.

The RC time constant

(150ns) is insignificant compared with the normal operating speed.

Gate leakage is typi- cally less than

50pA.

The simplest of the new devices are the single -channel types BSW95 and BSW95A. Intended for fast low

-current switching and high input impedance linear amplifiers, these are low capacitance units.

The BSW95 and BSW95A are the gate protected equivalents of the BSX83 and

BSX84. Very low

"on" resistance is charac- teristic of the

BSV20 and BSV20A, being the gate -protected equivalents

BSW31. of BSW30 and

The remaining two new devices, the

BSV34 and BSV34A are both dual- monoli- thic and suitable for low and high level choppers, ches.

The multivibrators and analogue swit-

"on" current is

22mA.

The

BSV34 and BSV34A are the gate -protected equi- valents of the BSX85 and BSX86. SGS -Fair- child Ltd., Planar

House, Walton Street.

Aylesbury, Bucks. ww

328 for further details

Pulse

Generator

A repetition rate continuously variable from

0.1Hz to 10MHz and width and delay variable from 35ns to 10 seconds are claimed for the A100 pulse generator introduced b)

Datapulse of California, U.S.A.Other feature include single or double pulses, <

5ns rise time, simultaneous

± 10V output, ± 400mV triggering sensitivity, and synchronous and non -synchronous gating.

Triggering featuret oscilloscope -type polarity and level controls.

Multiple decade switches plus

X

100 multi- pliers provide extended range capabilities. useful in biological and geophysical researck for educational purposes, television circuit design, etc.

Dimensions of the generator art

6

X

22

X

28cm deep and the price is

$470.

European Officer: Systron- Donner Inter- national, S.A. 447, Avenue de

Tervueren,

Brussels 15,

Belgium. ww

319 for further details

40- ampere Triacs

Low switching losses, low on

-state voltage at high current, low thermal resistance, uni- form gate current density, rapid electrical conduction, and efficient heat dissi- pation

-all

these features are claimed for four new 40A triacs from RCA

Electronic

Components. These are designed to switch from an off -state to an on

-state for either polarity of a

2.5V 80mA maximum gate signal. Types 2N5441 and 2N5444 are de- signed for 120V r.m.s. line operation and they can control a 5kW load. Types 2N5442 and 2N5445 are designed for 240V r.m.s. line operation and can control a 10kW load. The

2N5441 and 2N5442 are press

-fit types and the 2N5444 and 2N5445 are stud- mounted.

RCA Electronic Components, 415 South

Fifth Street, Harrison, N.J.07029, U.S.A.

WW313 for further details.

Monolithic

Crystal Filters

Monolithic crystal filters, type MXF, manu- factured by the Collins Radio Company:

California, U.S.A., are now available in the

U.K. from G.

A.

Stanley Palmer Ltd. The centre frequency range of these filters if from

3.5 to 20MHz with bandwidths of

Wireless

World, October 1961

380 www.americanradiohistory.com

1.005

:ally to 0.02% of centre frequencies. Basi- the filters are frequency selective de- ices, consisting vhich of coupled resonators in both the resonators and the coupling nedium are embodied in a single quartz

'late which, in turn, carries electrode pairs pn led its top and bottom surfaces; the electro- regions become resonators, and the re-

;ions between the electrodes, coupling ele- nents. When an electrical signal is fed to the nput electrode pair, the piezo -electric

'roperty of the quartz creates an acoustic lisplacement in the first electroded region.

'art of this acoustic energy, which is in the brm of a shear wave, tunnels in an exponen- ially decaying manner across the un-

:lectroded region to affect the second resonator, hereby creating an acoustic displacement. he energy continues in a similar manner hrough the array, until the piezo -electric of the last resonator turns it into an sutput signal. izes

The filters come in four case

-TO

-5, TO-8, `C' type and flat pack sousing different numbers of resonators. In dl instances passband ripple can be held to

1.25dB over an operating temperature range

'f -40

°C to

+

65°C. G.

A.

Stanley Palmer

Std.,

Island

Farm

Avenue, West Molesey

Grading

Estate, Surrey.

+VW309 for further details.

Timer

Advance

SC3 is a

-egular or four -decade instrument with clear in -line numerical display which

:an be used as a frequency meter, counter or rimer,

3anel the function being selected by front pressbuttons.

In the count mode, random signals in the frequency range 10Hz to 1MHz can be any totalled during arbitrary interval determined by either press button, external contact openings or

:losures, or external applied voltage.

Frequency is measured over the same range as on count, a time -base alternative gate times of is

0.1 provided with and

1 second, allowing accuracy measurements to be made to an of typically

0.1 %. When used as a timer or "electronic stop watch" the instrument displays in units of

0.01 sec. up to a maximum of

99.99 seconds.

Integrated circuits are used in this lightweight instru- ment. Advance Electronics

Essex.

Ltd., Hainault,

WW306 for further details

New

Counter

TV

Camera

Tube

The English Electric

Valve Co. Ltd. has produced a new type of

3

-in image orthicon

TV camera tube which can give pictures of a quality approaching that of a standard 41-in tube. It is available in two versions; the P874 which directly replaces the 8093B, and the

P875 which replaces the 7293B. In these new

Wireless World, October

1968 tubes the basic features of an ordinary Elcon image orthicon are combined with a special electron -optical design which reduces the noise in the output signal and eliminates the dynode background effect. English Electric

Valve Company Ltd., Chelmsford, Essex.

WW 318 for further details

High -surge Mains

Switch

A high -surge double -pole on

/off mains switch, specifically designed for modern TV receivers, ahs been tor

Division introduced by the Resis- of the Plessey Components

Group.

The outstanding feature of the switch, type R, is its ability to sustain instan- taneous current of about

60A. Also, it has low contact resistance. Type

RS is fitted with a steel spindle and type RIS has an insulated spindle.

The independent unit can be com- bined with one of Resistor Division's range of moulded carbon supplied to fit track potentiometers, or other manufacturers' poten- tiometers. Plessey Components Group, Re- sistor Division, Cheney Manor, Swindon,

Wilts.

WW305 for further details.

Chopper -stabilized

Operational Amplifier

Analog Devices Ltd., have encapsulated their model 230 chopper- stabilized operational amplifier in a package measuring approxi- mately

3.75

X

3.75

X

1.5cm.

In the bandwidth d.c. to 1Hz, peak -to-peak voltage noise is

1µV maximum. Stability is claimed to be better than

10µV per year, and voltage drift better than

0.1µV /degC. Open -loop gain is 10V/V, and the minimum output current at ± 10V is

4mA.

The power fall off is

6dB /octave which limits full power response to about 10kHz.

The amplifier will work from ±

12V to

+ 18V.

The unit weighs approximately

22g. Analog

Devices Ltd., 38/40 Fife

Roaçi, Kingston

- upon- Thames, Surrey.

WW

327 for further details

Smaller M.O.S.Ts

Mullard have introduced two new m.o.s. transistors, types BFW96 and BSV22. Both are electrically similar to types already avail- able but, because of their smaller encapsula- tion (TO -72 instead of TO-5), they are parti- cularly useful where space is limited. Be- cause of its very high input impedance

(typically

10-12Q) and low output impe- dance (typically

31Q with additional bipolar transistor), the

BFW96 is specially suited for use as an impedance converter. The BSV22, however, has a low offset voltage and negli- gible offset current. Therefore, in chopper circuits it will give a superior performance to that of a bipolar transistor. The maximum www.americanradiohistory.com

dissipation at

25 °C is

200mW for both types.

Mullard Ltd., Mullard House,

Torrington

Place, London W.C.1.

WW312 for further details.

Miniature Coaxial

Switches

A range of miniature coaxial switches, the

CCM series, are available from Radiall

Microwave Components

Ltd. Single pole, with a choice of

3, 4 or

6 ports, they are available in excellent manual control or electrically operated versions.

Fitted with 3mm precision coaxial connectors they are said to have performance up to

12

GHz, with a characteristic impedance of 50Q and a peak power handling capacity of

2kW.

Radiall Microwave Components Ltd.,

Station

Approach, Grove

Park

Road,

Chiswick, London W.4.

WW303 for further details

Analogue

/digital Convertor

A versatile

10

-bit analogue

-to- digital, digital

- to-analogue convertor, type APR -800 -AD/DA, is available from Lotus Electronics of

Norwich.

The module is able to convert analogue signals, in the range 0-4 V positive, into serial or parallel digital information, and conversely it is able to convert serial or parallel digital information into analogue signals from an i.c. output amplifier with a slew rate of

10V per

µsec. The module contains binary counter/ store, resistor ladder network, reference supply, clock and input and output amplifiers.

Lotus Electronics, 41

Thunder Lane, Nor- wich, Norfolk.

WW 324 for further details

Versatile F.E.Ts

Motorola's two plastics encapsulated versions of the 2N4416 junction field effect transistor have unusually low noise figures (2.0dB max. at 100MHz and 4.0dB max. at

400MHz) and will work equally well from d.c. to above 400MHz. The power gain is

18dB min. at 100MHz and 10dB min. at

400MHz. The MPF106 and MPF107 are both n- channel devices with input and out- put capacitance of max. respectively.

5.0pF max. and 2.0pF

The casing is

TO -92 pat- tern. Motorola Semiconductors Ltd.,

York

House, Empire Way, Wembley, Middx.

WW 326 for further details

1kW

Solid

State Power

Amplifier.

The Derritron amplifier described under this heading in last month's issue was erroneously illustrated by a photograph of a dissociated product.

381

World of

Amateur

Radio

current in Germany, distributed as follow

DJ, DK, DL calls, 11,997; DL4 and DL5 call

497, DM calls, 1,491.

The current annu; membership fee procal of

D.A.R.C. is

40DM.

Rec arrangements exist between Germai and a number of European countries includii

Austria, Belgium, France, Luxembourg, tl

Netherlands, Spain, Switzerland and ti

United Kingdom with agreements propose with Denmark, Norway and Sweden. Inform tion concerning the issue of short

-term licenc can be obtained from D.A.R.C.

Internation

Affairs

Office,

Muehlenstrasse 27, D -560

Doenberg, Germany.

I.A.R.U. Region

I Conference

The

Belgium

National Society

(U.B.A.) are assuming responsibility for making the arrangements for a triennial conference of

I.A.R.U. Region

I societies to be held during the period May

5

-9, 1969, at the Hotel Metro- pole, Brussels. Visitors to the conference will be entertained by U.B.A. members during the weekend preceding the opening when recep- tions will be held at the

Martin Club, Brussels and, possibly, at the Hotel de Ville (Town Hall) by courtesy of the burgomaster of the city.

Since the

1966 conference was held in Opatija,

Yugoslavia, in May of that year, the number of member societies in Region

I

Division has increased from

20 to 31.

70 cm Band Changes

The

G.P.O. has intimated that changes are to be made shortly to the existing 427

-450

MHz band allocated to radio amateurs in the

United Kingdom. On the effective date, to be announced, the frequency limits will be changed to 425

-429

MHz and 432

-450

MHz. The channel 429

-432

MHz is to be allocated to a new service which is due to come into operation in this band. Footnote 319 to article

5 of the Geneva Radio Regulations

1959, allocates frequencies in the

70 cm band on a primary basis to the radiolocation service and on a secondary basis to the amateur service.

Amateur televison takes place on frequencies in the

70 cm band (currently 427

-445

MHz) but under the re- arrangement, presumably, this will now become 432

-

,450 MHz.

Slow

-Scan

T.V.

Authorized in U.S.

The United States Federal Communications

Commission has authorized slow

-scan tele- vision in the high frequency and very high frequency amateur bands. On

3.5

MHz and higher, the operating mode is restricted to bandwidths no greater than a properly operated single sideband telephony trans- mission. On

50

MHz and higher bandwidths equal to a double sideband standard amplitude modulated signal will be permitted. cates to the purchasers of all types of trans- mitters. Radio amateurs in Italy are anxious that this new regulation should not restrict the construction and use of home -made transmitters.

Diplomatic Wireless Service at R.S.G.B.

Exhibition

The Diplomatic Wireless Service, modern version of a highly secret tion is to exhibit some of war its

-time organiza- latest communica- tion equipment on the centre stage at the

R.S.G.B.

Amateur

Radio

Exhibition to be held in the

New

Hall of the Royal

Horticultural

Society,

Greycoat Street, Westminster,

London, S.W.1, from October

2 to 5, 1968.

The exhibition is due to be opened at

12 noon on the 2nd by the Postmaster General

(Mr. John Stonehouse, M.P.).

Convention in

A

Ireland national convention for radio amateurs in

Northern Ireland and Eire will be held at the Ballymascanian Hotel,

Dundalk, on

Sunday, October 6, 1968.

The at

10 a.m. with an exhibition event will of amateur open radio equipment, followed at 2.30 p.m. by a business meeting.

The convention dinner is due to commence at

7 p.m.

Tickets and full details are available

Belmont from

S.

H. Foster (GI3GAL),

31

Park, Belfast

4.

(Complete day

£1.15.0; dinner only £1.7.6.)

New

Orleans 250th Anniversary

In commemoration of the celebration of the

250th anniversary of the foundation of the city of

New Orleans, Louisiana, in

1718 by

Jean

Baptiste Le Moyne, Sieur de

Bienville, the

Greater

New

Orleans Amateur Radio

Club is offering an attractive commemorative certifi- cate to any radio amateur who submits a log extract indicating two

=way communication on any bands, in any modes with three amateur stations located within the metropolitan area of

New Orleans. Claims, together with a stamped addressed envelope, should be sent to the secretary of the club at 2935

International

Trade Mart Tower,

2

Canal Street, New

Orleans, La, 70130.

Italy Curbs Transmitting Equipment

In order to keep a check on the sale of trans- mitting equipment to the general public and in particular the sale of 27 MHz Japanese

- made walkie- talkies, the

Italian government has introduced legislation which requires manufacturers and retailers to furnish certifi-

Amateur Radio Growth in Germany

Membership of the German National

Amateur Radio Society (D.A.R.C.) had by July

9,

1968 increased to 18,717, and of that total

10,408 were licensed amateurs. At the same date there were 13,985 transmitting licences

382 www.americanradiohistory.com

Real DX on Top

Band

Doyen of Top

Band operation Stewart

:

Perry

(W1BB), of Winthrop, Mass, in h summer 1968, 160 Metre DX bulletin, retort that

Robert Denniston (WDX, president c the A.R.R.I.), recently worked a station i

Japan and on the next morning worked G6& in England 6,500 miles westward, the

4,500 miles eastward: an achievement whit would not have been thought possible a fe years ago. parts

During the past few months mai of the world have been linked for tl first time on

Top

Band. When Stewart Per

. himself worked

CE3CZ

(Chile), on May

1 he achieved a century of countries worked t that band. Another outstanding result stan. to the credit of Roger Parsons 5Z4L

(G3RBP), in Kenya who succeeded in makii numerous contacts with the United Stat on 1827 kHz.

Earlier in the season Mr. Parso, made the first ever between contact on

Top

Bar

Tanzania and the U.S. when

1 operated as

5H3KK. He has also made number of Top

Band contacts with statioi in the U.K.

Wirral Society's

New

Headquarters

Wirral Amateur Radio Society, with

11 licensed amateurs in the peninsula and t known to be active, has solved its heat quarters problem by successfully negotiatir for the acquisition of the local

Civil Defenc building upon disbandonment of the corp

With a permanent headquarters the sociei now enters a new era which will enable a wide range of activities to be enjoyed by member

Gordon

Lee (G3UJX), was chiefly responsib for bringing about the acquisition of the ne building,

Indonesian Licences

Issued

At long last, the Indonesian governmer has decided to issue amateur licences using tr prefix YB but no official information been received from I.T.U. has headquarters y< i

Geneva that Indonesia has withdrawn ii previous objection to international commun cations by its amateurs.

New

Operating

Aid

A new operating aid, obtainable free charge from A.R.R.L. Communications Dep

225

Main Street, Newington, Connectict

06111 U.S.A., sets out the RST system, a tin- conversion chart, ending signals, the ARR. and ICAO phonetic alphabets and steps t take in an emergency.

JOHN CLARRICOATS G6C

Wireless

World, October

19E

Answers to

"Test

Your

Knowledge"

5

Questions

on

page

377

BULGIN

MOUIDEN

INSULAIION

1. (a).

Propagation of any travelling wave is always associated with a lag of phase proportional to distance in the direction of propagation.

2. (d) All points between a pair of adjacent nodes support oscillations with the same, phase. Adjacent antinodes are in antiphase.

3.

(b). points

The wavelength is the distance between adjacent supporting oscillations with the same phase for a travelling wave. Hence wavelength

=

27t

/phase change coefficient.

4. (b).

The basic definition of characteristic impedance is the input impedance of an infinite length of the line.

Such an impedance terminating a finite line absorbs all the incident power.

S. (d). per

For a line with primary constants resistance per

[mit length

R, leakance per unit length

G, inductance unit length L, capacitance per unit length

C,

Zo

=

V

(R

+ j co

L)

/(G

+jatC)

At low frequencies, because of the normal ranges of values of

R, G, L, and C it generally has a capacitive component, but at a high enough frequency and co

C w L R

s

G so that Zo

'L

/C which is resistive.

6.

-

(c).

)' I

If points

1 and

2 are separated gy distance

1 on

2 beine nearer to the load than

1, V2=

VI exp where y is the propagation constant. (a) and (b) are the phase change coefficient and attenuation coeffi- cient respectively.

7.

(d).

The term "non- dispersive" means sinusoidal signals velocity. of all frequencies travel with the same phase

If this is the case the phase change coefficient must be proportional to frequency.

8. (a). Answer (b) is sive line the phase velocity. On a non- disper- the group and phase velocities are equal, but on a dispersive line they are not.

9. (b).

K =

1

//phase velocity.

A line for which this equation holds would be non -dispersive so and group velocity are the same. that phase

10. (a). Although the steady state input impedance of a mismatched line is not generally Zo, this is due to the fact that currents and voltages in the forward and reflected waves add in a manner which depends on the length of the line in wavelengths and the phase change at reflection.

11.

(b).

The impedance quoted is the steady -state value ultimately presented to the generator when the standing wave has established itself by multiple reflections at load and generator.

12. (c). At antinodes the reflected wave voltage is in phase with the incident wave voltage, at nodes the two are in antiphase.

13.

(b).

14. (d).

Near the generator the reflected wave has suffered the attenuation of its journey to the load and back, and hence its amplitude is a smaller fraction of that of the incident wave than is the case near the load.

15.

(b).

A forms a quarter wavelength of loss -free line trans- short circuit into an open circuit. The input impedance at points any multiple of a half wavelength apart is the same.

Wireless

World, October

1968

A

WIDE RANGE

OF

SWIICHES

ACTUATING

MEANS AVAILABLE

List No. SM.270 /2 /PD

Toggle

List No. SMA46 /2

Push

-Pull

List

Ito. SM.327 /2 /PD

Biased Toggle

Push

List No. SRM.270 /2

Button, Successional

Action

List No. SMA19 /2

Biased, Push

Button

GENERAL

INFORMATION

A new range of D.P.C.O.

Chrome surfaces equally suitable switches of high quality and performance, combining highly Polished front of panel parts, with high grade Phenolic moulded body. The contact mechanism gives

'Quick Make and

Break' action and has entirely new design of fine Silver self-cleaning contact for Mains or Low Voltage uses.

Solder Tag connections are of Silver plated Copper and will accept

110 series Push/On Cable Sockets. The range comprises five forms of actuation (see above), which are mechanically interchangeable with many of our existing laminated types, but with improved performance and ratings.

ELECTRICAL &

MECHANICAL

DATA

2

Amps at

250V. A.C. 3

Amps at 125V. A.C.

1

Amp at 250V.

D.C. 1-I

Amps at 125V.

D.C.

(Non- reactive Circuit)

High Voltage Proof Test:

2KV. RMS.

Contact Resistance

(initial): 5mQ per pole at

10A. 2V. prospective.

Endurance: Typical Life Expectancy

50,000 operations

6

3

Amps at 12V.

A.C./D.C.

Amps at

50V. D.C.

SEND FOR LEAFLET No.

1531/C AVAILABLE

FREE ON

REQUEST

A.

F.

BULGIN &

CO.

LTD., BYE

-PASS ROAD,

TELEPHONE:

BARKING,

ESSEX.

01- 594 -5588

(12 LINES) Private Branch

Exchange.

WW

-126

for further deuils

383 www.americanradiohistory.com

Literature

Received

An unusual method of slide selecting power supplies is provided by a cardboar rule being distributed by A.P.T. Electronic Industries Ltd., Chertse

Rd., Byfleet, Surrey. The rule enables voltages up to 500 and data on power supplies coverin current up to

10 amps to be quickly determines

Prices and dimensions are also given on the rule.

WW 37

S for further details

Electronic bread -boarding aids of novel design for both educational an industrial use are described in leaflets produced by Howard Electroni

Industries, Manor Way, Boreham Wood, Herts.

A typical board consists of

14 connectors, each connector being capable of accepting up to twelve corn ponent lead out wires of widely varying diameter plus a

4mm wander plug desired. The boards are available in a number of configurations for discret i components or integrated circuits.

WW 376 for further details

"Reliability of Electronic

Equipment"

Pt.5 is a three

-page leaflet pro- duced by the British Standards Institution. It is intended to establish uniform criteria for an equipment reliability programme.

Some of the aspects covered include: reliability prediction; the evaluation of parts, materials and processes; and the testing of sample equipment during development and production. The leaflet is available from the

B.S.I. Sales Office, 101/113

Pentonville Road, London N.1, at

5s plus 6d postage.

Electronic Components,

Equipment and

High- fidelity Catalogue

(9th

Edition) from Henry's Radio lists more than 6000 items in 280, or so, pages.

The usual balance of kits, test equipment, components, tools, etc is maintained and Henry's comprehensive crystal list is included. The catalogue costs

7s

6d from

Henry's Radio Ltd.,

303

Edgware Road, London W.C.1.

Engineering publication

25-45 from Westinghouse Brake

&

Signal Company describes a range of silicon diffused

1.5A avalanche diodes which will withstand surge reverse voltages of up to 2kV and are encapsulated in resin.

Westinghouse Brake

&

Signal

Company Ltd.,

82

York Way, Kings Cross,

London W.1.

WW

362 for further details

The d.c.

Capsule Catalogue Selection Guide lists the range of special purpose power supplies available from Hewlett Packard. The folding guide classifies each power supply in terms of regulation,

0.1 0.01 and 0.001

%, in the first instance and then according to voltage and current ratings.

Hewlett Packard Ltd., 224 Bath Road, Slough, Bucks.

WW

363 for further details

We have received from the Spectral Electronics Corp., Italy (1) a trimming potentiometer short-form catalogue listing a comprehensive range of trim- mers of various shapes and sizes;

(2) two leaflets describing a range of miniature printed circuit switches typically

1

-pole 10-way or

2 -pole

5

-way and (3) a leaflet describing the company and some of its products. The U.K. agents for Spectral are: Guest Electronics Ltd., Nicholas House, Brigstock

Road, Thornton Heath,

Surrey.

(1) WW

364 for for further details further details

(2) WW

365 for further details

(3) WW

366

Product Data

(1) is published by Technical Indexes Ltd., Index House

Cross Lane, Marple, Cheshire, and performs three main functions: It act as an accurate buyers' guide, it is an index for the Electronic Engineerin.

Index

(2) and it also acts as an index for the Electronic Micro Data Servio

(3).

The product data book is reprinted three times a year and costs

250s pe annum. Product data supplies names and addresses of manufacturers of large number of components and test equipment.

(1) WW 377 for further details (2) WW

378 for further details

(3) WW 379 fa further details

H. F.

Predictions

-October

The predicted value of the Ionospheric Index (IF2) used for the charts is the same as that predicted and later measured for October 1967.

The trans -equatorial routes show little change from the previous few months except that the effect of shortening northern hemisphere daylight i becoming apparent. Hong Kong promises much improved working

(0600

-1200

G.M.T.) compared with the previous six months. The Montrea daylight MUF is up to 30MHz so operation above 20MHz should now become consistently good.

The LUFs shown were drawn by

Cable

&

Wireless Ltd. for reception in the

U.K. of commercial telegraphy; curves for high -power broadcasting would lx similar.

MONTREAL

15

20

40

30

MHz

BUENOS

A

RES

NNMI!\

Switches, signal lamps and signal lamp holders are the main subjects of the latest catalogue

(No. 137) from Arcolectric Switches Ltd., Central

Avenue, West Molesey, Surrey. Each component is illustrated and described in some detail.

WW

367 for further details

Two catalogues describing (1) r.f. coaxial cables and (2) coaxial con- nectors are available from Radiall Microwave Components, Station

Ap- proach, Grove Park Road, London

W.4. Both catalogues are well produced and give comprehensive technical information.

(1)

WW 371 for further details

(2) WW

372 for further details

Marconi -Elliott Microelectronics is the title of a brochure that describes the facilities available at the new microelectronics plant at Witham,

Essex.

The brochure is well illustrated and some of the manufacturing processes peculiar to integrated circuits are explained.

WW

373 for further details

An instrument that checks the speed accuracy of tape recorders when used in conjunction with a test tape is described in a leaflet from Calan

Electronics Ltd.,

6

Croft Street, Dalkeith, Scotland. The instrument can be used for speed checking at

19 or 9.5cm /s and has full -scale deflections of

± 10,

3, or

1%.

WW 374 for further details

384

0

0

4 i J

1

4

8

With auroral correction

8

12 16

JOHANNESBURG

12

G.

M.T.

16

20

20

10

8

6

5 aN11111N

4

030

1122NNNNNNN

12 16

20

MHz

40

30

HONG KONG

20

15

8

6

5

4

10

-----

======57Eil===I

-1111111111----U11111%

MMMMEMMOINE/Sil

....

12 16

20

G.M.T.

-

--

Median

Optimum standard

M U

F traffic frequency

Lowest usable

H F

Wireless

World, October 1968 www.americanradiohistory.com

INSTRUMENT

TO MAKE IT. THEY

ARE MADE

TO SOLVE PROBLEMS

SEND NOW FOR

THE

LATEST

CATALOGUE.

NEED AN

ADCOLA

SOLDERING

NO, WE o o

r

DON'T

KNOW

WHAT IT

IS

EITHER, BUT IF

SOLDERING

IS

IN-

VOLVED WE

KNOW

THAT

YOU'D

f

,

'

-

,

,

N

°

\

.

°

/

/

N '`

1

...

(.

.,

4

\

°

/\o

°

o

.:.

,

.

o\

o

.I.

ADCOLA rf

/

%j

lI

'

,

`

¡;/I

(

I

i

,/Oft

iy

I

/

/4044,

0

.

,

1ri

,

.....owe

.

DCO

PRODUCTS

LIMITED

(Regd Trade

Mork

)

Wireless

World, October

1968

ADCOLA

HOUSE,

GAUDEN ROAD,

LONDON, S.W.4

Tel.

01

-622 0291;3

Telegrams: SOLJOINT LONDON S.W.4 www.americanradiohistory.com

NOW!

the first oxide-free

high

purity extruded solder for printed circuit soldering machines, baths

und

pots

EXTRUSOL

is a new concept in solder for solder machines, baths and pots used in the electronics industry.

EXTRUSOL

is a very high purity solder

which

is also substantially free of oxides, sulphides and other undesirable elements.

The percentages of impurities in

EXTRUSOL

are considerably lower than those quoted in national or company specifica- tions, thus providing a solder more suitable for use in the electronics industry.

EXTRUSOL

of course can be released under AID authority and conforms

with

USA QQS

571

D.

ADVANTAGES

OF

EXTRUSOL

1.

Less

dross

on

initial melting

2.

3.

4.

5.

More soldered joints per pound of solder purchased

Less

reject joints

Improved wetting of electronic components

and

printed circuit boards

More uniform results

ALL

EXTRUSOL

IS

COMPLETELY

PROTECTED

BY

PLASTIC

FILM

FROM THE

MOMENT

OF

MANUFACTURE

UNTIL

IT

IS USED

A section of a typical cast solder bar.

Note the surface dross and general contamination.

A section of an

EXTR

USOL

bar

with

the plastic coating removed show- ing no dross or contamination.

Solder chips

which

are usually made by dripping solder alloy on to a metal surface.

EXTRUSOL

precision made solder pellets

individually

cut to a standard size from extruded rod.

EXTRUSOL does from oxide and much not only possess contamination, better physical appearance, but the alloys lower percentage of impurities. are

with

freedom formulated to contain a

EXTRUSOL

is

supplied

in

1

-lb. and

2

-Ib. Trapezium

Bars and

Pellets

in

different alloys with strictly controlled tin contents to suit the appropriate soldering machines, baths and pots.

Bars are

available for automatic solder feed.

Write for details of

EXTRUSOL and and literature about

Ersin

Fluxes

Chemicals for printed circuit soldering machines, baths and pots.

Multicore

Solders Ltd.

Hemel Hempstead, Herts.

Tel: Hemel Hempstead 3636 Telex: 82363

Wireless

World,

October

1968

WW

-002

FOR

FURTHER DETAILS 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