THE

RADIO CONSTRUCTOR

Vol. 23 No. 6 JANUARY 1970 THREE SHILLINGS

HIGH SELECTIVITY T.R.F. RECEIVER

A 2-valve (6BR7, ECL82) all-wave design with super-efficient
regeneration providing performance comparable with a
superhet

sr
 THt
Radio Constructor

Incorporating THE RADIO AMATEUR

JANUARY 1970

Vol. 23 No. 6
CONTENTS

Published Monthly (1st of Month) TORCH BATTERY POWER PACK 330

First Published 1947
EXPERIMENTAL RECORD PLAYER
AMPLIFIER (Suggested Circuit No. 230) 333
Editorial and Advertising Offices
57 MAIDA VALE LONDON m QSX 335
NEWS AND COMMENT 335
Telephone Telegrams SPACE DX 33s
01-286 6141 Databux, London
LOW-COST REGENERATIVE RECEIVER 339
ELECTRICAL FAULT-FINDING
© Data Publications Ltd., 1970. Contents
may only be reproduced after obtaining 'DOWN UNDER' 34!
prior permission from the Editor. Short
abstracts or references are allowable VERSATILE A.C. MILLIVOLTMETER 342
provided acknowledgement of source is
given. DESIGN FOR A CLASS 'A' BATTERY
Annual Subscription 42s. (U.S.A. and AMPLIFIER 349
Canada $5) including postage. Remit-
tances should be made payable to "Data HIGH-SELECTIVITY T.R.F. RECEIVER 352
Publications Ltd.". Overseas readers
please pay by cheque or International CAN ANYONE HELP? 359
Money Order.
CONSTANT CURRENT GENERATOR FOR
Queries. We regret that we are unable OHMMETER CONVERSION 360
to answer queries other than those
arising from articles appearing in this CURRENT TRENDS 332
magazine nor can we advise on modifi-
cations to equipment described. Queries
should be submitted in writing and METER SHUNTS
accompanied by a stamped addressed (Understanding Radio) 353
envelope for reply.
CURRENT SCENE - TOP BAND DX 368
Correspondence should be addressed to
the Editor, Advertising Manager. Sub- IN YOUR WORKSHOP 339
scription Manager or the Publishers as
appropriate. RADIO TOPICS 373
Opinions expressed by contributors are
not necessarily those of the Editor or LAST LOOK ROUND 377
proprietors.
RADIO CONSTRUCTOR'S DATA SHEET
Production.—Letterpress. No. 34 (R.M.S., Peak and Average Values) ill

Published in Great Britain by the Proprietors and np%a ■ a - __ ._ .

London^W g5313 Pub,icatl0ns Ltd' 57 Ma'da Vale, FEBRUARY ISSUE WILL BE PUBLISHED
The Radio Constructor is printed by Kent Paper OM PPRDIIADV 1e»
Company Ltd, London and Ashford, Kent. IPCDriU/^riT I Si
 TORCH BATTERY

POWER PACK

by

R. L. GRAPER

Quite a few constructors, including in particular the younger enthusiasts, prefer

to power transistor receivers and similar equipment with torch batteries rather
than with the standard batteries that are specifically intended for transistor
work. Apart from the possibility of initial cost saving, this approach certainly
makes it possible for the batteries to serve a double function during their
useful life. The article published here describes a simple home-built battery case
with external connectors which can accommodate two 1289 torch batteries

The battery unit described consists of a simple to reveal construction and the positions of the bat-
home-constructed case in which two 1289 torch teries. Above these appears the contact cover and
batteries are fitted. These are automatically con- lid. This is shown partly assembled, illustrating the
nected in series when the case lid is fitted, with the contact strips at the ends for positive and negative
connections, and the central springy contact fpr con-
result that a 9 volt supply becomes available for tran- necting the batteries in series. The end strips are
sistor radios and similar equipment. One advantage wrapped around the plywood and then glued. Ihe
with the unit is that it is not necessary to solder top panel is glued to the plywood panel, and has
any connections to the battery terminal strips, and clearance holes for the plug and socket connections.
the batteries can at any time be employed for other The latter are taken from an old PP9 battery or
uses, if desired. similar, the insulated panel on which they are mounted
ASSEMBLY being cut, as required, to fit the present assembly.
The various parts forming the unit are shown in
The completed assembly is illustrated in Fig. 1. Fig. are
2. Six lengths of iin. square planed stnpwood
cut and glued to form two U-shaped sections.
This diagram shows one side panel partly cut away These are then glued together, in their turn, to give
a single cradle lin. wide. On to this are glued the
two side and two end panels, these being cut from
i Formica laminate V^in. thick. Due allowance has been
I given in the sizes of these panels so that the edges
can be bevelled at 45° and a knife edge fitting secured.
There is no Formica panel at the bottom of the case.
The bottom consists of the four bottom pane edges
and the bottom surfaces of the wood, these all being
cleaned up flush. . , ^ 3 ■„
The top cover is formed from a rectangle of Tsin.
plywood, size 4||in. by lin. This is carefully sanded
down to fit into the top opening of the box. A recess
is cut out for the battery connector panels. After
soldering appropriate lengths of wire to them, these
are glued in position as also are the two end contacts.
1 Note that a groove is needed for the wire to the
negative connector, and that the wires must be
soldered to the end contact strips. A V-shaped piece
of brass foil, as illustrated in Fig. 1, is also glued
at the centre of the plywood. This contact connects
The unit with its lid removed together the inner terminal strips of the batteries,
THE RADIO CONSTRUCTOR
330
 Clearance hole for J5BA nuts
6BA threaded rod ,6BA plain washers
glued Central springy
shorting contact

Glued Clearance holes

Groove for wire
connection to
negative socket (SI

Positive contact,
^'wide brass
foil strip
0

Bevelled edges ^Recess slot

at 45°
'/2 square wood
% 50
5R
?5OCiy x
Brass countersunk
screw
5D
89

Formica panel "Formica panel

Cshown cut away)

'72^ square wood 3 spacers

Fig. 1. Illustrating the final assembly of the unit

causing them to be in series. No external connection

is made to this centre contact. Finally, the top surface mm
of the contact panel is treated with Evo-Stik, as also
is the Formica top panel above, and the two brought
into contact to form the complete lid. The output
connectors are then marked "+" and " to show
their polarity. This should, of course, correspond to
their polarity when fitted to a standard battery.
Two |in. lengths of 6BA brass studding act as
fixtures for the lid. These lengths are sunk into suit-
ably drilled holes in the wooden cradle as shown.
They project by about -J-in. and are securely fixed
at this height by short countersunk brass wood-screws
which engage them through the end panels. See Fig. 1.
The three spacers illustrated can be of cardboard 'iJWi
or plastic laminate, and are pushed in after the bat- ■■
teries have been lowered into the case. They prevent
excessive movement of the batteries. . mm
The top strips of the batteries should be bent as
shown in Fig. 1 and should be nearly level with the
top of the box. Very good contact will then be ob-
tained when the cover is fitted and the 6BA nuts When the lid is secured in place a 9 volt output
is available at the two top connectors
JANUARY 1970
331
 I'/s

3
/4 centres
5/8 dia
SCS
5/16
Formica panels
C'/32*thick)
6l/0,x3l/2''(2off)

Drill for
6BA
Formica panels
('/324thick)
372 x I'/S"
(2 off)
72 square wood

fl6
Glued

Slightly Ibevel 372

to 'S/lb"
2V2 Groove
25/l6 Recess
Hole for wire 732 deep
4i5/I6 Hole for wire
l3/4
/8
Spacers, stout cardboard or Bevel slightly with sandpaper
Formica (3 off) to fit

Fig. 2. The component parts of the unit have the dimensions given here

screwed down. The form of construction described Although this unit was intended for a 9 volt output,
gives a very robust unit, and replacement of the two there is no reason why a larger model could not
batteries takes but a few moments. be made for higher voltages. ■

SGS PUBLICATIONS
Currently available from SGS (United Kingdom) Ltd. are two manuals giving full ^^"m'iSed on
vices These are entitled respectively 'Professional Discrete Devices' and Consumer Devices and are noin priineu
an IIJ by 8Jin. format. Both manuals contain a 'Planar Selector' guide which cnables specific . j !, |lstcd
th e d e
from clearly laid-out tables, the guides being followed by vfry comprehensive mformat.on f A u ^ 1969-
The latter sections constitute the major parts of the manuals. Both manuals are fully up to date as 01 Augusi jjo*.
Uenis in 'professional Discrete Delict include transistors for a f ^
sistors, phototransistors, reference diodes, high speed diodes and1 »hyr,sto.rs-^ contents
elude general purpose transistors, u.h.f./v.h.f. transistors, high voltage and video output transistors, and mtegraieo
''Readers of The Radio Constructor may obtain either or both of these publications from SGS (United Kingdom)
Ltd.,' Planar House, Walton Street, Aylesbury, Bucks, at the price of 21s. each. the radio constructor
332
 Every now and again the The input signal from the crystal
writer includes, in the Suggested tion as intended, even though the
Circuit series, a design which pick-up is applied to the volume result is that TRI is operating at a
has to be categorised as being control, Rl, and is then fed, via R2 very low collector current. The rather
"strictly experimental". The circuit and Cl, to the base of TR1. The wide voltage swing from the crystal
to be discussed this month falls into collector of TRI couples direct to pick-up is changed to a correspond-
this class because both the tran- the base of TR2, which functions as ing current swing at the base of TRI
sistors it employs are operated under an emitter follower feeding into the due to the presence of R2 plus, at
conditions which are well removed output transformer primary. The settings other than full volume, what-
from those in which they would latter is one half of the primary of ever track resistance happens to
normally function. For this reason, a standard "Eagle" LT^ output appear above the slider of RI. Some
the circuit is not recommended for transformer, the secondary of this non-linearity in TRI is inevitable
the beginner who wishes merely to component coupling to a 3H speaker. but, as was found experimentally by
follow a tailor-made design which Base bias for TRI is provided by the writer, any resultant distortion
requires no setting up and which R4 and R3, and it will be noted that in the output of the amplifier appears
employs conventional circuitry. On these components have much higher to be well within acceptable limits
the other hand, the circuit will be values than are usually encountered for a low-cost design such as this.
of interest to the more experienced in circuits of this nature. It was found Capacitor Cl is a d.c. blocking
experimenter who likes trying out empirically, nevertheless, that these capacitor and it is essential that it
unconventional applications. The fact values enabled the amplifier to func- be a good quality paper or plastic
that the transistors in the circuit are
operated under unusual conditions
involves the risk that not all tran-
sistors of the types quoted may offer On-Off
the same results. Such a risk should +18V
not be particularly high in the
present case but its existence has to
be mentioned and accepted. blk.
"4 white
THE CIRCUIT IO-25Mn hite
(see text) green
The unit to be described is an -WW—
experimental record player amplifier (not used)
intended to be coupled to a low- TR, Cg
cost crystal pick-up, and its circuit ^2 BCI68C 200pF
appears in Fig. 1. The r.m.s. output 560kn 0 5'jF TRa 20V wkq
available is of the order of lOOmW. —I OC36
However, when coupled to a reason- Input
ably sized speaker of the order of IMn
6in. or more, the audible signal is log "3
more than adequate for a small IMn
room. The quality of reproduction
is not in the high fidelity class, but
it is certainly comparable with that
offered by a conventional transistor
amplifier having a low power Class B Tj - Eagle LT^^'nco
output stage. The main advantage of 0
the circuit is its exceptional sim- ecHoo R2to ffy-'/Bwatt (00/o
plicity. The circuit complications in-
herent in a more conventional ampli- BC168C OC36
fier employing a Class B output stage
are completely absent. Fig. 1. Circuit of the experimental record player amplifier

Experimental

A
Record Player

Amplifier

by G. A. FRENCH I

JANUARY 1970
333
 To emitter TRo
2-2Mfv
BUY THIS

BEST SELLER
2-2Mn

T.V. FAULT FINDING 2-2Mn

405/625 LINES
2-2Mn
2Mn 2-2Mn 2-2Mn lOMn
To base
TR,
- Fig 2. A suggested method of finding the value required in R4.
The flying lead may be terminated in a crocodile clip
am. m
foil component. If an electrolytic passing very high currents if the
8'6 value of R4 is accidentally made too
capacitor is used in the C1 position
the amplifier will not function. low, and so the setting up of R4
The output stage draws a standing must be carried out very carefully.
current of 20mA, which means that The method employed by the author
TR2 dissipates 360mW all the time consisted of wiring up a single 10MS2
that the amplifier is switched on. The resistor in series with a string of
REVISED & ENLARGED 2.2MS2 resistors, as shown in Fig. 2.
circuit offers no protection against
thermal runaway in TR2 (due to the The resistance was then reduced,
Edited by J. R. Davies small resistance of T1 primary, the from its maximum, in 2.2MJ2 steps
level of d.c. feedback via R4 is until the desired 20mA was drawn
124 pages only 8/6 negligibly low) and the solution to from the supply. In the writer's case
this problem consists, quite simply, it was found that about 17Mil (10MS2
Over 100 illustrations, including of choosing a much larger transistor in series with four 2.2M12 resistors)
60 photographs of a television for TR2 than the signal conditions produced the desired current, where-
screen after the appropriate warrant. In the prototype, TR2 was upon the circuit was finally wired up
faults have been deliberately operated continually for several with a lOMfl and 6.8 M£2 resistor
introduced. hours without any heat sink at all connected in series.
Comprehensive Fault Finding and, whilst growing perceptibly
warm, still passed exactly the same THE SUPPLY
Guide cross-referenced to emitter current at the end of this
methods of fault rectification period as it did at the beginning. As has already been stated, the
described at greater length in It would be wise, nevertheless, to current drawn from the 18 volt
the text. mount TR2 on a small heat sink in supply is 20mA. This current is not
order to provide extra thermal stabil- so high as to preclude the use of
Price 8/6 from your isation. Since the collector of TR2 batteries to power the amplifier, even
Bookseller is at chassis potential, the heat sink though the efficiency is not as good
can be provided by bolting this tran- as is given by an amplifier with a
or post this Coupon together sistor direct to a small metal plate Class B output. If a mains supply
with remittance for 9/2 or chassis on which are mounted all unit is employed there is no need
(to include postage) to the other components. During early to pay attention to voltage regulation
experiments the author used a small so far as signal handling perform-
DATA PUBLICATIONS LTD. TO-5 transistor in the TR2 position ance is concerned since the output
and this gave definite evidence of transistor operates in Class A. Apart
57 Malda Vale, London, W.9 thermal instability. It is necessary, from small deviations during over-
therefore, to employ the large tran- load, the current drawn remains
sistor specified. steady at 20mA. The supply voltage
The transistor used in the TR1 should not be allowed to rise above
Please send me the 4th re- 20 volts, however, as this is the maxi-
vised edition ot TV Fault Find- position is a BC168C (available from
Amatronix Ltd.). It is essential that mum collector voltage rating of TR1.
ing, Data Book No. 5. Should the unit be powered by a
TR1 be a silicon transistor with a
I encloie cheque/crossed postal order collector voltage rating in excess of poorly regulated mains supply it
lor 18 volts. would be advisable to connect an
18 volt zener diode across the supply
NAME SETTING UP lines. The function of this diode
would be merely that of ensuring
ADDRESS The circuit is set up by finding the that excessive voltage is not accident-
value in R4 which causes a standing ally applied to TR1, and it need not
current of about 20mA to be drawn draw any zener current when the
from the 18 volt supply. A word of amplifier and power supply are oper-
^R.C. Block Letters PleaseJ warning is necessary here. The OC36 ating under normal conditions. ^
in the TR2 position is capable of
THE RADIO CONSTRUCTOR
334
 Amateur band listeners tend to • BROADCAST BANDS
Q crowd on to the 14 MHz band, Several readers have written
this being the most favoured range stating they are Broadcast band
of frequencies for the reason that beginners and have not, up to the
they produce the most varied and time of writing, heard an In-
S choicest Dx signals throughout the
/» year than any of the other bands. donesian station, although they
Comparatively few listeners are have logged some of the signals
by X primarily interested, or even from the Indian sub-continent. The
operate to any extent, over the Indonesian stations on the l.f.
FRANK A. BALDWIN lower frequency Amateur bands. It bands are all comparatively low-
is a truism that Dx is relative, powered and conditions must be
(All Times GMT) among other considerations, to the good before any hope can be held
band in use - and many choice out of successful reception and
• TOPIC catches may be made on the lower recognition.
Short wave listeners, like any frequencies. Moreover, these trans- As a tip, the most consistent
group of human beings, tend to missions are reported upon by very Indonesian signal here is that of
be a highly individualistic lot - few s.w.l.'s, and this implies that YDK6 Djambi on 4927kHz - listen
some eventually obtain an amateur a comprehensive report to such a around 1530 to station close down
transmitting licence whilst others station would stand a very good at 1600. The receiver must be
are quite happy to remain s.w.l.'s chance of obtaining a reply. selective, this channel often being
and to thoroughly master the craft. To be an all-rounder s.w.l., oper- surrounded with CW and teletype
Others drift in and out of the hobby tion over the Broadcast bands QRM. The station closing ceremony
over a period of some years. should also be included in the - for want of a better term - com-
There are those, of course, who spread of interest. Broadcast band mences at about 1550 with an
remain in the hobby for some operation makes a complete Eastern type chanting song ren-
time, become quite well known in change and a break from the dered by a young woman and this
the world of short wave listening Amateur bands. Many Dx stations is followed by identification in In-
and then, for reasons best known are there for those who care to donesian and a most haunting
to themselves, vanish completely search for them and their recep- melody rendered on a Hawaiian
from the scene, never to return tion often represents a major feat guitar.
to the fold. The writer can recall of short wave listening. Other Far Eastern stations
s.w.l.'s of the past whose names logged were:-
became a byword in the hobby, • AMATEUR BANDS 4725kHz 1612 Rangoon, Burma,
even prior to the last war. Others Top band has shown some pro- with drama programme in ver-
have been on the short wave listen- mise of improvement on last year nacular.
ing scene for many years and these with signals being heard from 4740kHz 1620 Radio Maldives, with
old timers tend, as a group, to be across the Atlantic since early American religious programme.
conversant with all aspects of the October. The dates of Trans-
hobby, constructing their own Atlantic Tests still to be held are - 4790kHz 1610 Penang, Malaysia,
equipment - either from published 11th January and 1st, 15th of Feb- radiating recorded 'pop' music.
or their own designs - testing and ruary from 0500 to 0730. 4877kHz 1545 Saigon, S. Vietnam,
evaluating the results, and operat- 1.8MHz with drama programme in ver-
ing over both the Broadcast and CW; DL2PB, DL9KRA, EI9BJ, nacular.
Amateur bands. On the Amateur EI9J, GM3BGW, GM3HLQ, GM3YOR, 4907kHz 1434 Radio Cambodia,
bands, both phone (a.m. and GW3RHC, GW3VPL, GW3XJC, with a drama complete with
s.s.b.) and c.w. stations are logged, GW3XST, HB0NL, OK1ATX, OKIJR, many gong and cymbal clashes
the ability to read c.w. being neces- OK2BLH, OK3CJE, OL2AIO, - most impressive!
sary to become an all-rounder. Still on the lower frequencies,
it is probably true, however, that OL2AKS, OL5ALY, OL5AMT,
the majority of short wave listeners OL6ALT, OL9AY. Dx'er Alan Thompson tells me that
tend to specialise in one aspect Trans-Atlantic Dx signals logged Radio Botswana is heard on
of the hobby or another. In the to time of writing are:- K1BPW 4845kHz with BBC news at 1800
main, most listeners operate over (0525, 1804kHz): W1EXI (0545, and station identification in English
the Amateur bands, using a.m. or 1803kHz): W1BB/1 (0515,1804kHz) at 1809. He also mentions that
s.s.b. phone transmissions as their and VP9GW (0530, 1802kHz). Brazzaville (ex-4795kHz) is - or
prime targets. Unable to read c.w. 3.5MHz was at the time of writing - on
they are confined to the phone CW: K3JH, VE1AJK, VEIAVN, 4801.5kHz.
portions of these bands, thus re- WA1FSU, WA2CPQ. 17770kHz 0532 Karachi, Pakistan,
stricting their sphere of activity SSB: CN8HD, KV4FZ, VE3DJE, with sports commentary in
and interest and, in - the writer's V01AC, XE1KB. English.
opinion, only gaining a small part 7MHz 21535kHz 0800 Tokio, Japan, with
of the interest which the short CW: HP5CFW, HP6ASD, K4DJN, identification and news in
waves offer. If a continued and K5AZX, VE3DHC. English.
lasting interest is to be taken in 14MHz 21545kHz 0800 RSA, South Africa,
the hobby, such a listener is ad- CW: CR6AL, CR6KB, CX1JM, station identification and music.
vised to iearn the dots and dashes. HC2HM, HP1BR, KP4DDY, PZ1AV, Whilst dealing with the higher
The ability to read c.w. widens the VP8JV, VQ8CC, ZE1DL, ZM3ABC, frequencies, Alan Thompson also
horizons of the operator who has ZS1JJ, ZS5SG, ZS6GG, 9J2XZ. brought to my notice the following
previously only used the phone SSB: CP6HQ, CR4BP, EL6UN, - 11855kHz Saudi Arabia in
sections of the bands, it following ET3USA, HC1RF, HC20A, KP4AC, English, good signal between jam-
that an increased report/reply KZ5NG, PZ1CR, TR8MC, VK2SG, ming of Tirana (Peking relay) on
ratio is the result, especially when VK7KX, VP2LA, VP2VI, VP7DL, same channel, try around 1900-1930
morse transmissions are reported VP7NF, VP8KO, VP9BO, ZM3FT, when channel is clear. 15540kHz
upon. Licence holders who operate ZP5GJ, 6Y5RA, 9K2BF. Tanzania with English programme
on c.w. tend to receive far fewer For those interested, 9K2BF at 1700-1830. 15723kHz at 1730
s.w.l. reports than do their phone gave his QTH as P.O. Box 1083 Omdurman in English/Arabic.
counterparts. Kuwait.
JANUARY 1970 335
 TRANSISTORISED SPACE DX

Have you ever had secret dreams of receiving signals

CRYSTAL MARKER from another civilisation in our galaxy from way across
the other side of the Milky Way? If you have then you
are normal; if you haven't then you must be unimagina-
by tive!
Has a listening watch ever been made for such
S. G. WOOD, G5UJ signals? Yes, the first and only attempt to receive in-
telligent signals from another planet was made in 1960
and termed the Ozma Project - during which a listen-
This simple and inexpensive circuit can ing session of 150 hours was made using directional
aerials beamed to our nearest solar type stars, Tau
produce detectable 100kHz harmonics up Ceti and Epsilon Eridani, at a distance from Earth of
to 28MHz some 11 light years, and around which planetary sys-
tems may be orbiting. The project was dismally un-
successful.
AlOOkHz CRYSTAL OSCILLATOR CAN BE A USEFUL
adjunct to any "ham-shack" and it need not WHICH FREQUENCY?
be expensive or difficult to build. Apart from
the cost of the crystal - and these have been on the From experience gained in the field of radio astro-
market for as little as 7s. 6d. - the total expenditure nomy the choice of wavelength would lie between 3 to
can be very low. 30cms at which the effects of cosmic noise and
atmospheric absorption on a signal would be at a
Good results have been achieved with the circuit minimum. It has been suggested that the best fre-
shown in the accompanying diagram, using general- quency would be 1,420 MHz — this being the emitting
purpose germanium transistors such as the Mullard frequency of neutral hydrogen clouds around the spiral
OC70 or OC71. Even the old Red-spot types have arms of our galaxy - the reasoning being based on the
been found to function satisfactorily. assumption that other galactic civilisations may also be
The oscillator is of the Pierce type and sufficient conducting surveys of hydrogen distribution.
r.f. output for most purposes is available when the
circuit is powered by a 4.5 volt battery. Resistors R1 | SUCCESS OR FAILURE?
and R2 can be i or i watt 10% components. Switch At the present time, success is miserably zero. We
SI may be a normal toggle or slide type; alternatively need more sensitive equipment and improved tech-
a push-button can be employed, this being pressed niques, much time and investigation before the odds
to switch on the oscillator when required. Choke LI against success are anything like evens. Careful listen-
can be a small 2.5mH component such as the Repanco ing will have to be maintained out into space, to dis-
CHI. However, the writer employed a home-wound tances of hundreds of light years. One povyerful radio
choke, this consisting of 45 turns of 30 s.w.g. enam- astronomy transmitter, the 1,000ft. dish in Arecibo,
elled wire pile-wound on a short length (about lin.) pulses Puerto Rico, is claimed to be able to transmit radio
of sufficient power to be detected at a distance
of ferrite rod. The choke inductance is not critical. of 1,400 light years from Earth.
Pick-up wire
WHO WILL LISTEN?
How many intelligent civilisations are there in our
•)! I galaxy of sufficiently advanced technology able to
receive our signals? Well, astronomers have calculated
iiiLl that the number of single solar-like stars that have
accompanying planetary systems formed during the 10
47kn billion years of the existence of our galaxy, is some-
lOOKHz -±-
CZi C| B| , thing like 50 billion. With these sun-like stars, the form-
crystal ing of accompanying plants is a common event. Of
O OlpF 4'5V J- the planet total however only a relatively few will be
i TR
able to support intelligent life with the required con-
ditions - atmosphere, temperature, water, etc. With the
remainder, not all will see the emergence of life, and,
ISkn On-Off where it does emerge, not all species will be tech-
nically competent to take part in space communica-
tions - and some that are may not be interested in
Circuit of the 100kHz marker oscillator. Suit- making contact with others.
able components for L1 and TR1 are discussed
in the text One authority has calculated that there ought to be
i something like four million civilisations capable of com-
The various uses to which this little unit may be municating with each other.* Of course, all these
civilisations will not attain a culture peak at the same
put need hardly be enumerated here. It is possible - period. Nevertheless, there should be sufficient overlap
with care - to pick out the harmonics on a sensitive of civilised periods to make interstellar communication
receiver at 100kHz intervals up to as high as 28MHz. possible.
Should any difficulty be experienced due to receiver With a total of four million communications minded
screening, etc., a short length of "pick-up" wire may galactic civilisations, we may yet - at some future
be introduced into the collector circuit of the tran- date - be holding regular Dx schedules with another
sistor, as shown in the diagram. This can then be far across the vast distances of space!
positioned near the receiver aerial lead or aerial
terminal to increase coupling. *L. Berman, K6BW, writing in CQ, November, 1969.
THE RADIO CONSTRUCTOR
338

«
 LOW-COST
COMPONENTS
REGENERA TIVE Resistors
(All fixed values i watt 10%)
R1 47kS2
R2 1 MSI potentiometer,
RECEIVER linear, with switch
SI
R3 3.9kS2
R4 560kSI
by R5 5.6kn
Capacitors
A. SAPCIYAN Cl 365pF variable, air-
spaced
C2 0,01/iF, paper or
plastic foil
C3 0.01 /iF, paper or
This simple little medium wave receiver can be built plastic foil
as a 2-transistor set driving an earphone, or as a C4 5|uF electrolytic, 10V
wkg. (see text)
3-transistor set driving a speaker at moderate volume
level. Few components are required, and there is Inductors
LI, L2 See text
opportunity for experiment with some of these to Transistors
obtain best results with the particular transistors TR1 AF126
TR2 QC75
employed
Battery
B1 9-volt battery
Switch
Regenerative receivers are has a relatively high gain at medium SI s.p.s.t., part of R2
ideal for those who wish to wave frequencies. R.F. transistors
build simple and reasonably such as the OC44 have also been Miscellaneous
sensitive sets. It has to be re- tried, but they did not offer the 1 ,OOOS2 magnetic earphone
membered, however, that they same degree of sensitivity, particu- with jack plug and socket
are rather tricky to operate when larly at the high frequency end of (not needed with 3-transis-
compared with commercially made the band. When set up for best tor version)
sets, and that their performance can amplification, the collector current 2 knobs
never match a superhet in terms of for TR1 is around 1mA.
sensitivity and selectivity. Neverthe-
less, satisfactory results can be
obtained in areas of good signal
strength and local stations can be U/
tuned in without interference, which
is what most people want. Provided,
therefore, that their limitations are bee Earphone
appreciated, receivers of this class ;ext jack
can be instructive to build and can
offer a very useful performance. "2
The receiver described in this L—
article uses a simple circuit and may
be built in either a 2-transistor or
a 3-transistor version. If desired, the
2-transistor version can be con- 721 9V
structed first and the third stage TR,
added at a later date. The 2-transis- AFI26
tor receiver drives an earpiece, A
whilst the 3-transistor receiver f■ Shield TR2
drives a loudspeaker at moderate OC75
level. —WvV C3
2-TRANSISTOR RECEIVER
The circuit of the 2-transistor
receiver appears in Fig. 1. The fer- Shield
rite rod aerial coil, LI, is tuned by
Cl, its tap being coupled to the
base of TR1. Detection takes place O Q oe bo co > dot
due to non-linearity in TR1 and
TR2, with the result that the primary AFI26 OC75
function of TR1 is to amplify at Lead-outs Lead-outs
radio frequency. The transistor in
the first stage is an AF126, which Fig. 1. The 2-transistor version of the regenerative receiver
JANUARY 1970
339
 tory here since they have low im- off switch and is ganged with R2.
Coil L2 provides regenerative pedances, of the order of 10-7- Their R2 should be wired up such that it
feedback. The variable resistor R2 inserts maximum resistance into cir-
controls the base current to TR1 use will cause disappointment. A cuit when SI is switched off.
and hence operates both as a volume measure of stabilisation, which has When the receiver has been com-
control and as a reaction control. proved adequate in practice, is pleted and the adjustment procedure
It is adjusted normally so that TR1 given by taking the base bias cur-
rent for TR2 from its collector via (to be described later) carried out,
is just below the oscillation point. R4. The total current consumption it is worth experimentally changing
R2 has a high value in order to C4 for a 0.25/iF capacitor. With
accommodate transistors having of the first two stages is approxi- some AF126 transistors it may be
both low and high gain figures. mately 2.5mA. found that this gives a very useful
The second stage incorporates an If the constructor wishes to add a increase in sensitivity. If it does not,
OC75 and drives a high impedance third stage of amplification, the cir- the electrolytic capacitor should, of
(1,OOOS1) magnetic earphone. The cuit shown in Fig. 2 is employed. course, be retained in circuit. It may
miniature earphones used with com- The components to the right of the even be found, in the 3-transistor
mercially made sets are not satisfac- dashed line in Fig. 1 (including the version, that there is an improve-
6 ment in performance when a 0.25/iF
WW capacitor is similarly connected into
circuit in place of C6.
FERRTTE AERIAL WINDINGS
7 The aerial coil, LI may be wound
on a ferrite rod 4in. long and with
TR a diameter of lin. using 28
C6 s.w.g. enamelled copper wire. The
ID AC 128 9V total number of turns is 72 close-
wound, with the tap at nine turns
from the end which connects to C2.
Be prepared to add or take off a
Cs c few turns at the end remote from C2
7 in order to get the range exactly
right. In this respect, it will prove
1 helpful to wind the coil on a paper
sleeve which can be moved along
the rod to provide small changes in
AC 128 inductance, if desired. In general,
Lead-outs best results will be given with LI
at, or near, the centre of the rod.
Other ferrite rods of about the
Fig. 2. This extra stage is added to provide the same dimensions may be employed,
3-transistor version remembering that a longer rod will
require fewer turns, as also will a
earphone and jack socket) are not thicker rod. The writer checked the
then required, and the circuit of receiver with a ferrite slab which
Fig. 2 couples to that of Fig. 1 at was 4in. long, lin. wide and iin.
COMPONENTS points X, Y and Z. TR3 is an AC 128 thick. This slab required 64 turns
and its base couples to the collector with a tap eight turns from the C2
Resistors of TR2 via C6. R6 and C5 are end. The position of the tap in the
(All fixed values i watt 10/) decoupling components, and pre- coil is not very critical and accept-
R6 IkR vent unwanted couplings to the pre- able results are given when the num-
R7 27kn ceding stages via the negative supply ber of turns between the tap and
R8 3.9kn line. Without these components the C2 end of the coil is approxi-
R9 15012 there would be instability, evident mately one-eighth of the total num-
as "motorboating", when the battery ber of turns.
Capacitors ages and its internal resistance rises. The beginner will find it quite
C5 50/iF electrolytic, Total current consumption of the 3- easy to adjust the number of turns
10V wkg. transistor version is approximately on LI. Provided that approximately
C6 5pP electrolytic, 10V lOmA. the correct number are initially
wkg. (see text) It is desirable to use a fairly large wound on, he will find it possible
C7 IOOjuF electrolytic, speaker, say 4in. in diameter or to receive stations near the middle
6V wkg. more, to take greatest advantage of of the medium wave band. He may
the fairly low audio output avail- then, later, search for stations at the
Inductor high and low frequency ends. If
T1 See text able. The output transformer, Tl,
should have a ratio which presents stations at either of these ends are
Transistor an impedance of some 150R to outside the tuning range, the num-
TR3 AC128 25(K1 to TR3. A suitable type, for ber of turns can be adjusted accord-
use with a 3R speaker, would be the ingly.
Speaker Radiospares transformer Type T/T4. The feedback coil L2 is also
See text (Radiospares components may only wound on the ferrite rod (or slab)
be obtained via retailers.—Editor.) and consists of three turns of 28
Battery If a speaker with an impedance of s.w.g. enamelled copper wire. It
B1 Included with Fig. 1 130SI or so can be obtained, this should be wound on a paper sleeve
Components List or former which is free to slide
may be connected direct in the col-
Switch lector circuit of TR3, whereupon no along the rod.
output transformer is required. General construction of the re-
SI Included with Fig. 1 ceiver presents few problems, and
Components List Switch SI, in both the 2-transistor
and 3-transistor versions, is the on- the stages of either the 2-transistor
THE RADIO CONSTRUCTOR
340
or 3-transistor version may proceed oscillation occurs. Should there be of the receiver is at its greatest when
along the board or chassis on which no oscillation with R2 at "maxi- R2 is just below the oscillation
they are assembled in the same order mum", move L2 towards LI until point. If a howl is evident in the
as they appear in the circuit dia- oscillation commences. If there is 3-transistor version when the re-
gram. The wiring around TR1 and still no oscillation, even when L2 is ceiver goes into oscillation, trans-
TR2 should be kept reasonably very close to LI, reverse the connec- pose the connections to T1 primary.
short. In the 3-transistor version, tions to L2, whereupon the oscilla- Should the constructor so desire,
transformer T1 should be farthest tion should commence. a short aerial may be coupled to the
away from the ferrite rod, and its If oscillations are too strong they receiver via a lOOpF trimmer as
laminations should be at right angles may be reduced by adjusting R2 shown in dotted line at the left of
to the rod. towards "minimum" and/or moving Fig. 1. This trimmer and L2 posi-
L2 away from LI. L2 should be tioning are then adjusted experi-
ADJUSTMENT AND finally positioned so that oscillation mentally for the best compromise
OPERATION is available over all of the band at between sensitivity and selectivity.
a setting in R2 which gives best The external aerial can, incidentally,
When all wiring has been com- performance with received signals. be the springs of the bed if the set
pleted and checked, the battery In general, oscillation should begin is to be used as a bed-side receiver!
should be connected with correct with R2 about a quarter of the way As a final point, R2 is adjusted
polarity and the receiver switched from the "minimum" end when a to its optimum position for each
on. It should be possible, with either new battery is fitted. However, much station received. It will need to be
version, to receive at least one depends on the specific transistor advanced further towards its "maxi-
station by adjusting Cl. If no used for TR1, and final adjustments mum" setting as the battery ages
station is received, adjust R2 to- in the position of L2 should be for and its voltage falls.
wards its "maximum" position the best overall results, as adjudged
(minimum resistance in circuit) until on received signals. The sensitivity

ELECTRICAL FAULT-FINDING it was established that the section including the master
oscillator and the first divided chain contained the fault.
"DOWN UNDER" Leaving the heat lamp on so that the fault remained in
evidence, the correspondent then applied Electrolube
Two methods of tracing an intermittent fault, using freezer to each component in turn. The fault disappeared
Electrolube Freezer, have been described by a corres- when a mica capacitor in the master oscillator was
pondent to Electronics Australia, a leading Australian cooled by the spray. Fitment of a new capacitor provided
Technical Journal. In one, groups of components were a remedy.
heated and individually cooled in turn, whereas the re- The second case concerned a tape recorder which
verse procedure was employed in the second case. The would not record or play until it had been switched on
first involved a close circuit TV system. After about an for about 10-15 minutes.
hour's operation, the picture on the monitor began to "It occurred to me", wrote the correspondent, "that
expand and contract at approximately a 1 sec pulse rate. I might be able to identify the faulty component by heat-
The simple act of opening the side of the control-unit ing each component after mass cooling and note which
case was sufficient to cure the fault because of the loss one restored the performance most rapidly".
of heat. An infra-red lamp was therefore used to raise A group of components was cooled with the aerosol
the temperature of the equipment while the side was and the writer placed his finger on each in turn to apply
open. heat. A faulty transistor was discovered in this way to
The lamp was arranged to provide local heating of the correspondent's surprise as he did not suspect that
various sections of the equipment in turn. By this means a transistor could be temperature-conscious. A col-
league, however, suggested that the transistor contained
a broken bonding wire or faulty junction to either the
chip or pigtail, with only a microscopic gap.

mjuMSl

'GlXy-
"It's the latest in high frequency early
warning systems."
JANUARY 1970 341
 VERSATILE

A.C. MILLIVOLTMETER

by

G. W. SHORT

Intended for the more experienced constructor, this article describes a versa-
tile and inexpensive millivoltmeter design which incorporates many attractive
features. These include low battery consumption and the ability to operate
with any moving-coil meter having an f.s.d. within the limits SOO^A to 5mA.
Measuring ranges depend on the existing scale marking on the meter and
may be, typically, 5mV to 500V, The frequency response within 5% error
extends from 20Hz to 30kHz

Most millivoltmeters are designed around BATTERY DRAIN

meters of a specific sensitivity, usually I00;j,A
instruments. The transistor a.c. millivoltmeter The usual type of transistor millivoltmeter (see
described here is different. It has been deliberately Fig. 1) incorporates a Class A amplifier. This drives
designed to accept a wide range of indicating meters a rectifier meter, which is placed in a negative feed-
so that the constructor, by following some simple back path to improve the linearity.
rules, can adapt it to accommodate whatever meter The drawback with this type of circuit is that if
he happens to have, the only requirements being that a relatively insensitive meter has to be used the
it should be a d.c. instrument with a linear scale battery drain becomes too large. Suppose a 5mA
and a sensitivity of 500/xA to 5mA approx. The meter is to be used. If the input signal is a sine
amplifier part of the instrument may be used as a wave then, to drive the pointer to full scale deflec-
general-purpose audio amplifier with an output of tion, the input current to the meter must be 5mA
up to 2.5V r.m.s. The amplifier will deliver over r.m.s., or 7mA peak. The peak current is what
30mW to a 7012 speaker. matters, because it fixes the minimum current which
the second transistor must pass. In the present case,
this must be over 7mA d.c. or the peaks of the sine
SPECIFICATION wave will be clipped and the millivoltmeter will be
inaccurate. Unfortunately most millivoltmeters will
also be called upon to handle speech and music volt-
Input impedance: greater than 1MQ, ages. These have rather peaky waveforms, and to
Measuring ranges: 5mV - 500V approx., handle them without clipping calls for a much higher
full-scale. standing current, perhaps four times the meter full-
Standby consumption: 9V, 2.5mA approx. scale current. By the time the drain of the earlier
stages has been added the consumption will be about
Overload capacity: 240V, 50Hz on most five times the meter f.s.d. current, or some 25mA for
sensitive range. a 5mA meter, which imposes a severe drain on the
Frequency response: 2O-3O,O0OHz, ± 5 %. battery.
Indicating meter; 500/xA to 5mA, any SELF RECTIFYING OUTPUT STAGE
linear scale.
Amplifier section usable for other purposes. The battery drain can be reduced by using a
Class B output stage with a low standby current. A
Class B amplifier can deliver large peak currents
342 THE RADIO CONSTRUCTOR
on demand, which is just what is needed for dealing
with peaky waveforms. A possible method of using
it would be to substitute the Class B output stage
for the second stage of Fig. 1, driving the meter RADIO CONSTRUCTOR
from the Class B pair in much the same general
way. This, however, is needlessly complicated. An
alternative arrangement, shown in Fig. 2, saves the
rectifier by making use of the fact that the lower
output transistor of the complementary Class B pair FEBRUARY ISSUE
conducts only on alternate half cycles, and therefore
acts as a half-wave rectifier to drive the meter.
The only problem that now arises is that the meter
registers the standby current of the transistor even
when there is no input signal. The remedy is to bias
the transistor exactly to cut-off. In practice, however,
it is better to adjust the bias until the meter registers
a tiny current and then move the pointer back into
position by means of the zero adjuster screw.
This last approach gets rid of the problem of the
standing current, but it creates a new one. An
amplifier biased in this way suffers from crossover
distortion. Such distortion is not important in itself
in a measuring instrument, but it does spoil the
linearity. The answer is to apply heavy negative
feedback. Some negative feedback is introduced
automatically by the d.c. bias resistance RB, which
of course allows a.c. to pass from output to input ELECTRIC DRILL
as well, but the feedback given here is not enough.
An additional a.c.-only negative feedback resistance, SPEED CONTROLLER
RF, is therefore added. This simple-to-build device enables the speed of
an electric drill to be smoothly varied from near-
DESIGN STRATEGY zero up to maximum or set at any intermediate
point, taking advantage of the facilities offered by
Constructors familiar with the Fig. 2 type of a currently available triac. The speed controller
will prove tp be very useful to the home handyman
circuit will realise that it is not going to be much when using an electric drill for such jobs as sand-
use when very high sensitivity is required, because ing, buffing, polishing, sawing or when drilling
the gain is not very high. More amplifying stages large-sized holes.
are needed. At this point, however, it pays to forget
about the exact circuit configuration and think about COMPREHEIMSIVE
the essentials of the feedback arrangement.
These are shown in Fig. 3, and as many textbooks TIMING UNIT
will point out, the performance of such an amplifier A small battery powered unit utilising 7 tran-
sistors and 2 diodes in an interesting timing circuit
providing a good insight into pulse circuit tech-
niques. The circuit can easily be adapted to time
almost anything from eggs to the daughter's phone
calls! Operation is simple - only one toggle switch
►H-I—11 - and even this could be dispensed with!
EXPERIMENTER'S
POWER SUPPLY
.!< ' □ Offering a continuously variable output voltage
from zero to 25 volts, this power supply circuit
also includes an overload protection circuit en-
abling output current to be limited to any desired
r level up to 1 amp and an optional trigger circuit
VAV giving a visual indication when the overload circuit
Input begins to operate.
PLUS
% OTHER CONSTRUCTIONAL
I
PROJECTS
vvw m DATA SHEET 35
Fig. 1. Typical transistor a.c. millivoltmeter • SUPPORTING FEATURES
circuit. To avoid clipping the peaks of the
signal waveform, the collector current of the
second transistor must be several times the ON SALE 1st FEBRUARY
meter full-scale current
JANUARY 1970 343
 is governed very largely by the three ^resistances
shown, regardless of what is inside the "box", pro-
vided that the gain inside the box is very high. For
NEW HANWPAK practical purposes, the following statements can be
taken as true: . . _ De
Input impedance of whole circuit ^ 1,7-, 0 c
Voltage gain, Vout/Vin — Fv'Jrt*
PANEL SIGHS Current gain, lout/Iin — H r / KL
These equations tell us that, provided we leave
the feedback resistance RF alone, we can adjust the
IDEAL FOR WORKSHOP USE current gain (by changing RL) without affecting the
voltage gain. Alternatively, we can adjust the voltage
gain (by changing RS) without affecting the current
gain. This is just what is needed lor adapting
the circuit to a wide range of meter sensitivities and
scale markings.
m
RANGE SETTING
Different voltage ranges could be selected by
switching in different values of RS. In this respect,
RS acts in exactly the same way as a "multiplier"
in any ordinary voltmeter circuit. The idea is attrac-
tive if only a few relatively high voltage ranges are
required. With a reasonable amount of amplification,
• Easy to fix the overall sensitivity can be 1MQ per volt, which
is a lot higher than the sensitivity of even the most
• Permanent and durable sensitive non-electronic multi-range meters. But it is
still not high enough. For a range of 3mV f.s.d., for
• Stapled in booklet form example, RS would only be 3Kfl, which is much
• Designed to hang above workbench too small. , .„• „ . j
For this reason, the complete milhvoltmeter des-
• Pocket for loose cuttings cribed here uses a high input impedance buffer stage
• Professional finish in front of RS, and carries out its range selection
by means of a potential divider in front of the
buffer stage.
Sets Wording—WHITE
if Set 4 Wording—BLACK BUFFER AMPLIFIER
if Set 5 DIALS—clear background A mosfet makes an ideal buffer from the point of
if Set 6 DIALS—black background view of high input impedance. Fig. 4 shows a suit-
able circuit, incorporating an R.C.A. mosfet type
40468A, which is the successor to the well-known
The Perfect Transfer for 40468 (The 40468A is available from Amatromx,
Ltd., as indeed are all the other semiconductors
the Home Constructor employed in the instrument).
The mosfet is connected as a "source follower
Each Set contains Six Sheets with a "gain" of less than 1, and having good
stability, virtually infinite input resistance, and a
fairly low output resistance. The low output resist-
5'- per set ance enables fairly small values for RS to be used
if need be. The input potential divider shown has
(postage 4d.) six 10 to I steps and covers input ranges such as
5mV to 500V. All the resistances except R7 are
common standard values, and R7 can be made up
To Data Publications Ltd. from two 22ft resistors in parallel without much
57 Maida Vale, London, W.9 loss of accuracy.
Series resistance R1 and the diodes D1 and D2
Please supply. perform the very important job of protecting the
Set 3 Set 4 Set 5 Set 6 gate of the mosfet against high voltages. No damage
results from a sustained overload of 240V on the
1 enclose cheque/crossed postal order for most sensitive setting of the range switch. Note that
Name R1 must be able to dissipate 1W under these
conditions.
Address
COMPLETE CIRCUIT

BLOCK LETTERS PLEASE Those extra amplifying stages inside the "box" of
Fig. 3 appear in full in the complete circuit of big.
THE RADIO CONSTRUCTOR
344
 must drive a current through the load resistor R21
which is equal to twice the meter current. The first
step in adapting the circuit to a given meter is to
select a value for R21 which gives the required
current, working from the equation;
R21 = (Output voltage)/(Twice meter current)
= 500/Im
(where Im is the meter full-scale current in milli-
amps and R21 is in ohms). Thus a 1mA meter calls
for a value in R21 of 500.0. A 470.Q resistor may
i1-! in practice be used.
The meter current having been dealt with, we next
r—VM | have to select ranges which correspond to the exist-
ing meter scale markings. This is quite simple. The
most sensitive range should be between 3 and 30mV,
and we choose a figure that fits the scale markings!
Input If, for example, the meter is marked 0-500, the
iRF lowest range can be 0—5mV. Having done this, we
then use a value for RIO with as many kilohms as
the lowest range has millivolts; e.g. 5kO for 0-5mV.
Once again, it is quite in order to use a resistor of
standard value slightly below the calculated value;
e.g. 4.7kl2 instead of 5kQ.
The design is now complete as far as adapting the
Fig. 2. A linear class B amplifier draws only a circuit to the meter is concerned. There is still, how-
small idling current. It can also be made to do ever, a further important point to consider, and we
its own rectifying: a meter connected as shown shall next turn to this.
receives d.c. equal to half the mean a.c.
output current VOLTAGE LIMITATION
5. The whole of the main amplifier is direct-coupled In the writer's opinion, it is always better to be
in the simplest possible way, advantage being taken safe than sorry where high voltages are concerned.
of the ability of silicon transistors to operate with In the present case, the wattage rating of R2 imposes
collector to emitter voltages which are the same as an upper limit on the input voltage of 700V r.m.s.
their base to emitter voltages. Using a fairly close However, it is quite possible that an even lower
tolerance transistor (type BC168C) in the first stage, limit will be imposed by the voltage ratings of
the d.c. conditions for all stages can be set by means certain components. If, for instance, a co-axial input
of one preset resistor, R15. This is adjusted so that cable is employed, how much voltage can safely be
the standing voltage at the emitters of TR5 and applied between its inner and outer conductors?
TR6 is 4.7V above the negative supply rail. Setting Again, how much voltage will the input socket
the bias of the output pair is carried out by R18, stand? Or the insulation between adjacent tags on
which trims the voltage across the stabilising diodes the wafer switch? Or between the wafer and the
D2 and D3. These diodes can be almost any type of spindle?
silicon junction diodes. These are likely to be unknown quantities, and
For calibration purposes, the a.c. gain has to be when dealing with unknown quantities it is always
made adjustable, and this is allowed for by putting safest to be pessimistic about them. In consequence,
a potentiometer, R20, in the negative feedback it is recommended that the upper limit to the milli-
circuit. voltmeter ranges should not exceed 500V. On this
An extra transistor, TR7, is placed across the basis the ideal meter scaling is 0-5 (or 0-50, or
meter. Its job is to protect the movement from 0-500) because this gives six ranges starting from
switch-on surges and overloads. TR7 conducts when 5mV (which covers the signals likely to be obtained
the voltage across the meter coil exceeds about from audio sources such as pickups) and ending at
500mV. This is quite a lot more than the normal r
drop at full-scale deflection for most meters, and Rs f
if desired the drop across the movement can be WvV- -ww-
artificially increased to, say, 400mV at f.s.d. by
inserting a series resistor of appropriate value at
the point marked with a cross. TTie author used a
BC168 for TR7, but any other silicon p.n.p. type
will be satisfactory provided it has a low bottoming Current
voltage and is capable of taking about 100mA peak amplifier
collector current. out
'•'out Rl
ADAPTING TO THE METER
With a sine wave input the amplifier can deliver
a little over 2.5V r.m.s., but to allow for peakier Fig. 3. General picture of a current amplifier
waveforms it is advisable to design for full-scale with negative feedback, showing only the im-
deflection at an output of only IV r.m.s. This output portant resistances which define the gain and
input impedance
JANUARY 1970
345
 possible. A probe containing lOMfl series resistance
gives a voltage multiplication of 10 with this instru-
ment. No ordinary lOMfl resistor will stand 1,000V
R2 or so across its ends, so instead of one 10MO
IMn resistor use, say, ten IMfl resistors in series, these
r MOSFET being spaced out in line down the barrel of the
oo 3
lOOkn Drain probe. High voltages of the order we are considering
Gate Substrate can be very dangerous and the probe must, of
01 lOka course, be designed so that there is no possible risk
Source of shock to the user.
Ikn When such a probe is used, it is still quite in order
R To mam
Re S to have a six-position range switch. The wiring of
IOO D
i-- d2 3-3 amplifier the sixth step could be modified as shown in Fig.
6(a) so that the sixth position is now a "meter off"
position. (But remember that though the meter is
off the power supply is not, and an on-off switch
is still needed.)
Fig. 4. High input impedance buffer using a Alternatively, it is possible to dispense with the
mosfet. The diodes protect the gate from probe and modify the resistor values of the last
excessive voltages
two steps in such a way that the sixth position gives
a range multiplication factor, not of 10, but of 2
500V (which covers the mains and most transformer (Fig. 6(b)) or of 3 (Fig. 6(c)). By using this trick,
secondary voltages found in the power packs of a meter scaled 0-10 can be given a highest range
valve equipment). An instrument scaled 0-3, etc., of 0-200V or 0-300V instead of 0-1,000V. The only
which would allow 3mV to 300V to be covered, is disadvantage is that some mental arithmetic is need-
also very useful. A rating for R2 of i watt is ed when reading the meter on this range. When
sufficient in this case, while allowing a fair margin using the arrangements shown in Figs. 6(b) and (c)
of safety. the new resistor values are used instead of R6, R7,
If some other meter scaling is used it is best to and R8. The rather complicated network of Fig. 6(c)
incorporate only five ranges and to make some is employed merely to enable the right range to be
special provision for high voltages. The best arrange- obtained using standard resistors.
ment here is undoubtedly to make a special high- If a five-position range switch is used, one resistor
voltage probe. A probe with a tubular insulated of Ilia can replace R6, R7, and R8. In practice
body of thick plastic or s.r.b.p. can easily take a it may be easier to use two 220a resistors in parallel,
few thousand volts. The limiting factor is flashover though the value HOQ is also standard and obtain-
or "tracking" between connecting points, and the able from some suppliers. It belongs, however, to
remedy is to spread the voltage out as much as the "E24 Renard Series" in which successive resist-

s VM R15
/bc^\
(000) + 9V
D(o O)G 2-5mA
subs. 2N3794 ww
tr5
40468A Bciesc 2N3794 I =6
C2 z:d3 C7
TR :- 4
I 40468A TRp TR3 TR6
BCI68C BC168C 2N429I
subs. r
22

TR4
C3 bci68C C4 I =5 (A)Output
mmt1
Input (Rg)
TR? "21
r
20 l'
9
(Rp) 19
IS 44 1544

TR7- see text

Fig. 5. Complete millivoltmeter circuit. R10 and R21 are selected to suit the meter,
as described in the text
THE RADIO CONSTRUCTOR
346
 SUPPLY VOLTAGE

COMPONENTS The millivoltmeter is not unduly affected by vari-

ations in supply voltage, and no voltage regulation
Resistors is provided. The prototype, when adjusted for f s d
(All fixed resistors high stability, and i watt with a sine wave input and a 9V supply, showed
unless otherwise stated. R2 to R8 inclusive 5% an error of less than 3 % when the supply was 6V
or better. All other fixed resistors 10% or and less than 10% with 4.5V. The error at 4.5V
better. See text for details on potentiometers). was due to peak clipping and not to falling-off of
al
R1 lOOktl 1 watt ft.s.d. Jhcthe
j' By error is of cours
same' token, thee, error
smaller at lessifthan
is larger the
R2 IMQ ^ watt
R3 lOOkQ meter is used to measure peaky waveforms. In
R4 lOkfl general, the battery should be replaced when its
R5 IkO voltage has dropped to 6V.
R6 lOOQ (see text)
R7 22n (see text) CONSTRUCTION
R8 22n (see text)
R9 3.3kQ The instrument still has to be made, of course,
RIO See text and this is not a task to be undertaken in a slap-
Rll happy way, using odd bits and pieces from the junk
IMfi box. A prime consideration is to make sure, well
R12 1.5MQ
R13 lOkfl before beginning, that there is going to be plenty
R14 4.7kn of room for everything inside the cabinet and on
R15 5Mn potentiometer, preset the front panel. Last-minute alterations and the
R16 15kn cramming of components into odd corners are likely
R17 4.7kn to spoil the performance by introducing stray capaci-
R18 15kn potentiometer, preset tances and couplings which, even if they do not
R19 330Q cause instability, may still play havoc with the
R20 5kn potentiometer, preset accuracy. The following points should be given due
R21 See text attention.
R22 ion Screening. It is essential to house the millivolt-
meter in a metal box. In addition, the input con-
Capacitors nections and range switch should be put in a separ-
C1 O.I/iF, 250V wkg. ate screened compartment. This may seem over-
€2 I 25/j.F electrolytic, 10V wkg. elaborate for an audio instrument, but experience
C3 2.5/xF electrolytic, 64V wkg. with the prototype showed it to be essential. With-
C4 33()pF out screening, the meter registered a reading on all
C5 320/j.F electrolytic, 2.5V wkg. settings of the range switch when a 50kHz input
C6 125pF electrolytic, 10V wkg. at 15V was applied, even when TR1 was taken out
C7 250/xF electrolytic, 6V wkg. of circuit! This explained a rather striking rise in
the frequency response above about 20kHz. Screen-
Semiconductors ing the input connections eliminated the effect.
TR1 40468A Input wiring. The range resistors should be small
TR2 BC168C and wired directly between the tags on the range
TR3 BC168C switch. This keeps down the stray capacitance. Care-
TR4 BC168C ful attention to. the placement of the "earth" con-
TR5 2N3794 nection of the input circuit is essential. The im-
TR6 2N4291 portant thing is to return the bottom end of the
TR7 See text range attenuator (R7 and R8) to the earthy side of
D1 IS44 the input socket with a short, thick wire. If, as is
D2 IS44
D3 IS44, or any silicon junction diode
D4 IS44, or any silicon junction diode
Meter
Ml moving-coil meter, see text

1-pole 6-way wafer

S2 s.p.s.t., toggle or wafer
Miscellaneous
Coaxial sockets, terminals (as required)
9V battery

ances increase by 10% increments, whereas 2200

(and all other resistances in the diagrams) belongs
to the more popular E12 Series, with successive Nme lives? You'll need 'em if your's is like this!
increments of 20%.
JANUARY 1970
347
desirable, a coaxial input socket is used, then the resistor tolerance which can be accepted is 5%, and
"earth" connection can be made to a solder tag 1 % is worth having if the meter scale is reasonably
clamped to the metal front panel by one of the long. All resistors in the millivoltmeter should be
bolts that fix the socket. Fading this, make a con- high-stability. The tolerance of R1 is not important,
nection to the panel close to the socket. If an actual but it should be rated at 1W.
earth connection is used, it should go to the same The weakest link in the calibration circuit is R20.
point. All the negative wiring should be carried out Ideally, it should be wire-wound, but economy will
with thick wire, to reduce couplings between output usually dictate a carbon preset component. Use a
and input via the wiring. good make such as Egen type 467, and check the
Buffer stage. The mosfet should be connected last, calibration from time to time. (The Egen type 467
after all the other components, and especially after potentiometer and suitable potentiometers for R15
diodes D1 and D2 are connected into circuit. When and R18 are available from Amatronix, Ltd.)
soldering, remember to earth both the bit of the If an audio generator with a monitored output is
iron and the circuit. When handling the 40468A, available, check the frequency response. Taking the
always pick it up by the case, never by the leads. reading at 1kHz as a reference, the error should not
Since the gate presents a very high impedance, it exceed 5% between 20Hz and 30kHz, with the pos-
should be kept away from the input socket. It is sible exception of the second most sensitive range,
quite a good idea to mount the buffer on a little where the h.f. response may tail off at a lower fre-
circuit board or tagstrip supported behind the range quency because of the effect of stray capacitance
switch. The bolts which hold the switch wafer can across the potential divider output.
often be used to support the buffer circuit, and the It will be appreciated that the bandwidth of the
whole lot can then be put inside the screening com- amplifier between 3dB points is much larger than
partment for the input circuit. the "useful measuring bandwidth" of the instru-
Main amplifier. Use a logical layout, with the ment. An error of 3dB amounts to some 30%, which
input at one end and the output at the other, and is far too great in a measuring instrument. A wider
positioned so that the input lies near the buffer stage bandwidth can, however, be used when making com-

-os o5
05 39a
lOOn 56n<
39a
-06 2-2 kA
56n< 39a
22n^ ^22A

(a) Cb) CO

Fia 6 Modified "highest range" connections to SI. The circuit of (a) is used when only
five ranges are needed but the switch has six positions, whilst (b) gives a sixth range which
is twice the fifth range. The configuration in (c) gives a sixth range which is three times
the fifth range

whereupon the connection between these two is kept parative measurements at a fixed frequency outside
short. Any of the usual constructional techniques, the normal measuring band. Indications up to about
including pin-board, will do. 3 MHz are obtained.
Terminals. Co-axial connectors are best, and are If linearity is unimportant, the sensitivity can be
probably essential if hum pickup is to be avoided, greatly increased by short-circuiting the output ter-
but it is a good idea to provide wander-plug sockets minals. If this is done, the meter will register a
as well, g-eep the output terminals away from the reading with no input; the reading being due to
input. noise in the amplifier. (If the wiring layout is poor,
short-circuiting the output may cause the circuit to
CALIBRATION break into oscillation, in which case the meter reads
full scale)
Calibration can only be done with the aid of a The lower limit of the frequency response can be
known input voltage, which usually means monitor- pushed below 20Hz if desired, by increasing the
ing the input with another meter of known accuracy. value of C3. Note, however, that C3, if electrolytic,
The applied frequency should not be too high, and must have low leakage. This is because its leakage
should be well inside the range of accurate measure- current, if appreciable compared with the base cur-
ment for the other meter. (Many multimeters are rent of TR2 (which is less than 1 /j.A), will upset
only accurate up to 2kHz). In general, 50Hz will the biasing of the whole amplifier. The easiest way
be a good bet. If possible, choose a voltage such of ensuring low leakage is to use a capacitor with
that the extra meter reads near full scale but the a much higher voltage rating than the actual voltage
millivoltmeter reads about half scale. Adjust R20 across it in the circuit, which is above IV. This is
so that both instruments read the same. Only one why a 64V working capacitor is specified in the
adjustment is needed; the other ranges will be right parts list. (2.5/xF, 64V is one of the values in the
if the range resistances are correct. The worst range Milliard miniature range). B
THE RADIO CONSTRUCTOR
348
 There is 2-stage negative feedback
■■ ■■ of d.c., and the circuit is conse-
quently very stable. The large elec-
trolytic capacitor prevents negative
I Design for a Class A I feedback of the signal, though a
small amount of negative feedback
results from the top half of the
potentiometer being in the source
lead without a bypass capacitor.
Battery Amplifier This feedback is not large as the
■■
■■ ■■
■■ resistance involved is small in com-
■■
■■ ■■
■■ parison with the drain load.
■■
■■ a
Itir ■■
■■ A capacitor appears across the
■■ 11V
j ■■ input to the amplifier, this being
found necessary to avoid excessive
|| SIR DOUGLAS HALL, K.C.M.G., M.A.(Oxon) j| treble and consequent distortion.
■■
■■ Constructors may experiment with
— ■■ different values. The 0.001,aF value
Incorporating a field effect transistor, this comprehensive shown in Fig. 1 proved to be the
best compromise for various pick-up
design may be either constructed in the form of an a.f. cartridges tried with the prototype.
amplifier, or as an a.f, amplifier combined with a single
MODIFICATIONS
transistor medium wave tuning head. In the latter instance,
some of the components perform dual functions at a.f. and r.f. The circuit in Fig. 1 will func-
tion as an efficient amplifier as it
stands with almost all specimens of
IT DOES NOT SEEM THAT EITHER the 40468. It is set up by first setting
a gain of about 120 times. The the slider of the potentiometer close
commercial firms or designers of maximum dissipation of the 2N3702 to the end connected to the source
equipment for home constructors is 300mW at 250C and with about and then, with a meter in one of
cater very frequently for the listener 16 volts available from emitter to the battery leads, adjusting the
who wants to be able to hear collector (about 2 volts are lost
records, or local broadcast stations, potentiometer until a reading of
across the relevant part of the 25011 18mA is given.
using a battery amplifier which will preset potentiometer) a maximum
give really good quality at ordinary However, there are various modi-
current of 18mA is permissible. The fications which are worth while, and
room levels of volume. Most home output transformer chosen has a some of these are shown in Fig. 2.
listening is done at an average out- ratio of 9.2:1 and will therefore It will be noted that an a.f. choke
put level of less than 50mW, and cause a reflected load of about 25012 now appears between the drain of
for this purpose, a total of 150mW to appear across its primary when the f.e.t. and the positive supply
undistorted output allows for the connected to a 312 speaker. With a
peaks to be reproduced without dis- line. This component was omitted
typical 2N 3702 this will produce a in Fig. 1 because the great majority
tortion. load of 250 x 120 = 30kl2 at the of 40468's pass negligible leakage
The output of the amplifier to be base, and this is the drain load for current. But the f.e.t. type 40468
described is about 150mW, and this the f.e.t. which will ensure good has now been superseded by the
is at a very high level of quality amplification from that device. 40468A which, while offering cer-
which deserves to be fed into a Source bias for the f.e.t. appears tain improvements, is no longer re-
good speaker. The amplifier "draws across the lower half of the 25012 stricted for leakage current, and the
18mA from an 18 volt battery, and pre-set potentiometer. This poten- latter can, in rare examples, be as
provided fairly large cells are used— tiometer should be a standard 1 watt high as 100/iA. Let us consider the
the U2 size is recommended—it will component, and not a miniature. case where a high leakage f.e.t. is
prove economical to use, giving very
many hours service from a set of 12
of these cells.
4-18V
BASIC CIRCUIT 9-2:1 To-3n speaker
The circuit is shown, in its sim- o
plest form, in Fig. 1. The output
from a crystal pick-up, or a similar
high impedance source of signal, is 2N3702
applied to a high resistance volume
control, and thence to the gate of a 40468
field effect transistor. The drain of
the f.e.t. is directly coupled to the 2Mni
base of a medium power silicon log '
p.n.p. transistor, the drain current
of the f.e.t. being the base current
of this output transistor. At the time Input ■ O-OOI
1
of writing there is still a dearth of 250n
silicon p.n.p.-transistors which will
maintain their amplification factor Wi rewound
When more than about 1 or 2mA 640pF
pass through them, but the inexpen- 6 V wkg
sive 2N3702 will maintain an hFE I
of from 60 to 300 in the conditions
applying in this circuit. Typical Fig. 1. Basic amplifier circuit, illustrating the combination of
specimens may be expected to show f.e.t. and output transistor
JANUARY 1970
349
used in combination with a 2N3702 will have stamped into it "20kS2 a transistor holder first, and then
offering a current amplification of l.Okfi". There are inferior imitations insert the f.e.t. afterwards.
300. If this output transistor passes which have "LT44" applied by what A tone control has been added in
18mA, a base current of 18mA appears to be a rubber stamp. These Fig. 2, this consisting of a SOOpF
divided by 300, i.e. 60/iA, will be imitations are not suitable for the
required. But such a current is lower present requirement, as both the 1 'i
than the f.e.t. leakage current alone, d.c. resistance and the inductance of
so the circuit will not function. the winding will be found to be far ■ components i
At first sight it might appear that too low.)
a resistor of about lkt2 between the In both Figs. 1 and 2, the sub- 1 Resistors
drain of the f.e.t. and the positive strate of the f.e.t. is left unconnected. 1 (All fixed values i watt 10%)
supply line would put matters right, Turning to other points in the cir- R1 3.3Mn
as the f.e.t. would then pass a total cuit of Fig. 2, the non-earthy input R2 330kS2
of about lOOfiA whilst the output lead is now taken to a 330kI2 resistor R3 39kn
transistor was passing a base cur- which is isolated from the volume VR1 2.5kn or 5kS2, preset
rent of 60/jA. The leakage current control by a 0.0 IpF capacitor. The potentiometer, minia-
of the f.e.t. then becomes a small capacitor should be rated at 250 ture •
proportion of the total current volts working. Also, two silicon VR2 2Mt2 potentiometer, log,
passing through it. However, the diodes (not germanium) are con- with switch S3
IkSl resistor would reduce the drain nected back to back between the VR3 25012, preset potentio-
load of the f.e.t. and amplification slider and the earthy end of the meter, wirewound, 1
would fall drastically. A better volume control. The 0.01/iF capaci- watt
scheme is to use the primary of an tor protects the gate of the f.e.t.
Eagle LT44 interstage transformer from any direct voltage which might 1 Capacitors
instead of the resistor. The primary accidentally find its way there, and C1 4/iF electrolytic, 6V
of this transformer offers a d.c. re- the diodes limit alternating voltage wkg.
sistance of about Ikfi, and presents on the gate to about 0.5 volt so that C2 640/jF electrolytic, 6V
the a.f. signal with a high inductive accidental applications of, say, mains wkg.
load, whereupon all is well. The hum, are rendered harmless. What- C3 0.0 I^F
transformer primary inductance is ever other refinements are added it C4 0.00 l^F
damped by the output impedance of is particularly recommended that 05 0.01/xF, 250V wkg.
the f.e.t. and the input impedance the diodes be included in the cir- 06 SOpF
of the 2N3702, and no noticeable cuit, or the f.e.t. may die a 07 0.02^F
deterioration of quality accompanies mysterious death one day. It is also 08 1,000/iF electrolytic,
its inclusion in the circuit. The green very wise, incidentally, to use a 25V wkg.
lead and outer red lead are used transistor holder for the f.e.t. Al- VC1 SOOpF variable air-
here, the central red lead and the though an iron can be used for spaced or solid dielec-
black and white leads being left wiring it up with all due precautions, tric (see text)
disconnected. (Make sure that the it is only too easy for a_ lethal dose VC2 SOOpF variable solid
LT44 is an Eagle component. This of a.c. to be applied to the gate. It is dielectric
will be sold in a box and its clamp better to make all the connections to 1 Inductors
LI, 2 (see text)
T1 Output transformer type
Drain Source T/T7 (Radiospares)
T2 Interstage transformer
(o^o) type LT44, primary
connected only (Eagle)
+18V
Substrate Gate 2N37oa 1 Semiconductors
4O468,4O460A Lead-outs TR1 BCI68C
Lead-outs TR2 40468 or 40468A
TR3 2N3702 or TIS61
D1 OA81
D2,3 silicon diodes
A.F (e.g. IS44)
choke 2N3702 02 Switches 1
330 SI 2-pole 2-way rotary
S2 2-pole 3-way rotary (or
—-fWM- 3-pole 3-way (see text)
c
8 S3 s.p.s.t., combined with
1 VR2
Sockets,
X etc.
OOI SOpF PS1, 2 phono sockets
pF 40468A 1 transistor holder (for TR2)
3 battery holders for 4 U2 cells
3n speaker each
j SOOpF mounted with
amplifier 1 Battery
12 U2, cells
□ 1 Speaker
3S2 speaker
I 1 Miscellaneous
Sin. x iin. ferrite rod
Control knobs, as required
Fig. 2. A more comprehensive circuit, developed from Fig. 1. Cabinet, etc.
Unmarked components have the same values as in the basic circuit
350 THE RADIO CONSTRUCTOR
 +I8V
fCoo^ A
bee
BCI68C rjt-
Lead-outs

trs v Dp
2N3702
TR, A
.t£
BCI68C s
6 2b
A b
-IrHI
Cb
TR2
c
3 ^VR 4O460A
||-^WV
/?VC 2 D,
3
R|
VW—(
C| a >
C4
I n

Fig. 3. Adding a medium wave tuning head. All components in this diagram are specified
in the Components List

variable capacitor. Enmeshing the itself is concerned, it will probably flex" design is very suitable here,
vanes of this component causes a be found easier to obtain a 3-pole offering in this case an output im-
very smooth loss of the highest 3-way component rather than a 2- pedance of about 330kS2.*
audio frequencies. The range of con- pole 3-way switch. In this case the All the components to the right
trol available is improved at inter- unwanted pole is ignored. of the new switch, SI in Fig. 3,
mediate settings of the volume con- It will be seen that the earthy have already appeared in Fig. 2, and
trol by the presence of the 50pF end of the volume control is now the wiring of this part remains un-
capacitor across the non-earthy half taken to the negative supply line via changed, although some of the com-
of the volume control circuit, this the secondary of the output trans- ponents now have a radio frequency
giving a degree of treble lift when former. This provides a useful addi- function to perform in addition to
the variable capacitor is set to min- tional degree of negative feedback their duties at audio frequencies.
imum. Finally, the 39kt2 resistor which considerably improves quality. The radio signal appears across LI
prevents any possibility of spurious At first sight the amount of feed- which is wound on a length of ferrite
oscillation by the f.e.t. back might appear excessive, but it rod, and is tuned by VC1. The
So much for modifications at the should be remembered that the out- whole of this tuned circuit forms the
input end. At the output end a 2- put transistor functions as an emit- base input of TR1 which, al radio
pole 3-way switch is employed to ter follower, which means that al- frequencies, is a common collector
make provision for an internal or though it is an efficient power amplifier with a high input impe-
external speaker, and to take the amplifier it gives no voltage gain. dance. The r.f. output, much ampli-
output direct from the f.e.t. when it So the additional feedback intro- fied as to current, appears across
is wished to use the amplifier to re- duced is, in effect, over the f.e.t, diode D1 in the emitter circuit. The
cord on to tape. As shown in Fig. stage alone. It may be necessary to diode next functions as an audio
2, the switch is at position B, and try both ways of connecting the frequency load and TR1 amplifies
an internal speaker, built together leads on the secondary of the output again, this time as a common base
with the amplifier, is in use. If the transformer. If connected wrong way audio amplifier. A combination of
switch is turned to position A the round, positive feedback will be comparatively high direct voltage
output is taken to a phono socket, given with consequent oscillation. from the battery and a type of
into which a lead from an external transistor which maintains good
speaker may be plugged. If the TUNING HEAD amplification when passing only
switch is turned to C, the output is SCfiiA means that a very high output
taken to this same socket direct from Fig. 3 shows how, if desired, a *The author's "Spontaflex" circuit was
the f.e.t. and a tape recorder may simple medium wave tuning head originally introduced in the June 1964
be used. The internal speaker re- may be added. As the amplifier has issue. The latest and most up-to-date ver-
mains connected and may be used a high input impedance it is useful sion was described in "Modifying the
for the output of the tuner to be 'Spontaflex* Circuit for Silicon Tran-
as a monitor though volume will be sistors", published in the May 1969 issue.
reduced owing to the additional load similarly at high impedance. The —Editor.
across the f.e.t. So far as the switch author's single transistor "Sponta- (Continued on page 359)
JANUARY 1970 351
 HIGH-SEL

T.R.F. R

F. G. RAYER, A

Our contributor recently turned his i

receiver and decided to re-examine ij
design and construction of a simple t
taken to ensure that efficient regenen
the receiver is described in this article
of the tests he carried out on his desk
between its performance a

Few circuits can provide such good results obtainable from a t.r.f. set depend almost entirely on
with so little complication as an efficiently oper- regeneration. F. E. Terman, in discussing regenerative
ated t.r.f. receiver. The writer recently returned detectors, points out that "Although regeneration re-
to the use of a small t.r.f. circuit for compact equip- presents an inexpensive means of increasing the radio-
ment required for /A operation. The performance frequency amplification, it increases the selectivity
obtained, together with subsequent tests, show that excessively ...."*
the t.r.f. receiver is in fact an excellent project when Nobody tuning a crowded amateur or other short
the complicated circuitry of a superhet is to be wave band these days is likely to feel that selectivity
avoided. can be excessive, and it is worth recalling that the
This article will describe the t.r.f. receiver used. adjunct to some expensive communication receivers
First, however, the writer would like to give details is a Q-Multiplier, this being a circuit which takes ad-
of the tests he carried out, if only to demonstrate to vantage of regeneration to increase selectivity.
those who automatically advocate the superhet that
the t.r.f. receiver can offer a comparable short wave *F. E. Terman, "Radio Engineers* Handbook" (First Edition, page
performance. And, of course, the t.r.f. receiver is 574), Mc-Graw Hill.
much easier to construct and bring up to full working
order.
REGENERATION
Before making comparisons between t.r.f. and
superhet receivers, it must be noted that the results

Amplitude
Superhet " 1
<1

T.R.F.
Off tune O Off tune
Sig. gen. Frequency
Fig. 1. The method employed for comparison Fig. 2. Illustrating the sharpening of selectivity
of t.r.f. and superhet receiver performance which results from the use of regeneration
352 THE RADIO CONSTRUCTOR
LECTIVITY

Cover feature
IECEIVER MAAAAM

by
Fig. 2 indicates the effects of regeneration with the
t.r.f. receiver. With no regeneration, tuned circuit
A.I.E.R.E., G30GR selectivity is broad, as in curve A. Introducing regen-
eration increases selectivity, and the response is
i sharpened, as in curve B. With critical regeneration,
a high and very sharp curve is obtained, as shown by
attention from the superhet short wave C.
tts t.r.f. counterpart. The result was the In tests made in conditions favourable to the t.r.f.
t.r.f. circuit in which particular care was receiver (that is, with weak and relatively weak
signals over the frequency range 15 to 3 MHz) curve
ration was obtained. The construction of C was on the average indistinguishable from that
le, and the author begins by giving details obtained with a superhet having two i.f. transformers
ign, these tests giving direct comparison incorporating four 465kHz tuned circuits. The super-
het also had one r.f. stage but this contributed
and that of a superhet receiver virtually nothing to adjacent channel selectivity, being
provided to reduce second channel interference. This
interference arises with a superhet from signals
spaced by twice the intermediate frequency from
wanted signals, and can be a troublesome effect. It
One way of comparing t.r.f. and superhet receiver is of course wholly absent with a t.r.f. receiver.
performance is shown in Fig. 1. Each receiver has a At frequencies higher than 15MHz or so good
speaker, with muting switch if wanted. Input is first results were still obtainable with the t.r.f. set, but
applied from the signal generator. The method is to more care in setting up was required. At lower fre-
tune in the signal generator output accurately with quencies, very strong signals may be present, as from
both receivers, then check for individual off-tune medium wave stations, and these tend to swamp a
response by retuning the generator. The superhet t.r.f. receiver. Strong local signals on any frequency
a.g.c. is switched out or temporarily disabled for this tend to swamp weaker adjacent signals, but this effect
test. can be reduced by suitable means.
R
io
47kn 22kn 2-7kn 470n
w vvw mt- wv
I C4 I =6 i Cg
VR 4pF T32 8pF
ookn JLpP
25kn C3 T I
220
V TC lOOOpF
oOpF OOSjjF Speaker
IOOO OOI
6BR7 PF PF
VR 2b
SO pF
19 C-5 1 1
Mn
bVZ^'
I Ma ECL82
TC2 IMn C8
* vc, / 25pF
oOpF 365** 2-2ka
P
VC37. Avc2
330a

AC mains
Fig. 3. Complete circuit of the t.r.f. receiver
JANUARY 1970
353
 dust cores, some of the ranges listed in the Table
TABLE I differ slightly from the nominal range figures quoted
in Denco literature.
Frequency Ranges VC1 is for coarse tuning, or band-setting. For
(Coils employed are Denco Miniature Dual-Purpose, short wave tuning and critical attention to a narrow
Green.) band of frequencies, such as in an amateur band,
the small variable capacitor VC2 is employed, this
Range Coverage giving fine tuning. VC2 should operate smoothly.
Its exact value is to some extent a matter of choice,
1 150-500kHz because a valve which gives adequate coverage on
lower frequencies is too large for easy band-spreading
2 500-1,600kHz of higher frequencies. If operated without a slow-
3 1.5-5MHz motion ball drive a value of about 5 to lOpF is
4 5-18MHz suggested. A value of some 15 to 20pF will be
satisfactory if VC2 is fitted with a ball drive or if
5 15-36MHz the main listening interest is in the low frequency
bands. A recommended component is the Jackson
type C804 variable capacitor.
Trimmer TCI is practically essential in the circuit,
and it reduces aerial loading with all but short aerials.
SENSITIVITY Since one ofthe prime objects in the receiver design
is the ability to obtain smooth, easily controlled
Sensitivity was checked using a laboratory type regeneration, there are three variable components
signal generator as input, as shown in Fig. 1, the which affect performance in this respect. The first of
t.r.f. receiver being that to be described here. This these is VC3, the reaction control proper. This has a
receiver employs Denco plug-in coils, the ranges relatively small capacitance and it is paralleled by
covered being listed in Table I. To compare the results trimmer TC2. The result is that a wide angle of rota-
given for differing L/C ratios in the tuning circuit, tion in VC3 is possible for a relatively small change in
tests were carried out at 16MHz, which is both at the overall capacitance, and its adjustment is consequently
low frequency end of Range 5 and at the high fre- eased. The third variable component is VR1, which
quency end of Range 4. Table II gives the results, varies the screen-grid potential of VI and enables
and it will be seen that readings were also taken at a this valve to be in the optimum condition for smooth
frequency near the highest tunable, SOMFlz, and at regeneration. With long aerials, adjustment of trim-
9MHz. The input figures quoted in Table 11 are for mer TCI will assist in obtaining best regeneration
the weakest signal capable of being copied with the performance.
t.r.f. receiver, using its loudspeaker, in conditions of The 6BR7 specified for VI is particularly suitable
insignificant background noise. They vary somewhat, as a detector, being a high gain type of rigid mechan-
depending on absolute optimum adjustment of regen- ical design. VR2 is an audio gain control, and the
eration. If there is local ambient background noise, following triode and pentode sections of the ECL82
headphones would give better results. provide audio amplification. The h.t. supply is
Practical results with an aerial appear to confirm obtained from a half-wave rectifier circuit with ade-
the tests. Stronger distant signals, such as 28MHz quate smoothing, particularly to the early stages.
U.S. amateurs, and those on 21 and 14MHz, could
be received in great number with a modest aerial MECHANICAL CONSTRUCTION
when propagation was suitable. The best Dx was
Australia. In fact, with the arrangement of Fig. 1, Construction must be rigid, so a 4-sided chassis is
it was quite difficult to find a signal with the superhet used, with 3in. brackets to give additional support
which could not also be brought in with the t.r.f. to the panel. The brackets were actually a 4 by 3in.
set. The signals lost with the t.r.f. receiver were very Home Radio "Universal Chassis" side cut from oppo-
weak transmissions, and weak or other signals too site corners at 45 degrees. The chassis is drilled for
near a strong transmission. Over the 30 to 1.8MHz components as in Fig. 4.
range, the t.r.f. set will receive literally hundreds of
amateur and other short wave signals.
The t.r.f. receiver employed in the tests will now
be described. It should be mentioned that equally TABLE H
good results could have been obtained with any T.R.F. Sensitivity
similar type of t.r.f. receiver in which particular
attention has been paid to obtaining efficient regen-
eration. Input Signal Voltage
Frequency
CIRCUIT NOTES a.m. c.w.
A few points should next be mentioned regarding 30MHz 4/xV 2fiV
the receiver circuit, which is given in Fig. 3. Com-
mercially made plug-in coils are used for simplicity 16MHz (Range 5) 10/xV 4/jlV
in band changing and wiring. As may be seen from 16MHz (Range 4) 5fiV 2/J,V
Table I, five coils cover all short wave bands as well SjuV 3/j.V
as medium and long waves, should these last two 9 MHz
ranges be required. Due partly to slightly different
tuning and stray capacitances, and the adjustment of
354 THE RADIO CONSTRUCTOR
 Chassis, 6xIOx2
mc
Aerial r -n Pin 8,Li
socket
3V 230V

To D|
lV2 23/4
Primary
*2 /
mc
0

S To C|
vc
VC|

a.
v
\
Panel, s'x fo'' Cursor 2'/8 Pointer

Fig. 4. Top view of the chassis, giving the more important dimensions

VC1 is bolted to the chassis, and its spindle must taken to chassis as in Fig. 5. Blue is neutral, going
match up with the ball drive. Spacing washers are to "B" on T2 (see Fig. 3). Brown is live, running
essential between chassis and this capacitor. A bolt from the plug fuse to SI, which in turn connects
secures the ball drive lug to the panel. If a drive is to "A" on T2.
to be used with VC2, set this capacitor back suffi- It is important to note that the tag positioning
ciently, and mount it on a strong bracket. for T2 shown in Fig. 4 is typical only, and that
Fig. 5 shows the underside. Holes for VR1, VR2, individual transformers may have different tag posi-
VC3, the mains lead grommet, and to provide clear- tions or may have wire lead-outs. The mains trans-
ance for the speaker and aerial/earth sockets are former should be connected up so that it agrees with
most readily made with a fin. chassis punch, before the circuit diagram of Fig. 3. Suitable components
chassis assembly. (The type of chassis listed consists for both T2 and T1 are available from Home Radio,
actually of a 6 x lOin. flat plate, with separate 6 x 2in. and the appropriate Catalogue Numbers are given
and 10 x 2in. flanged sides which are bolted on.) in the Components List.
WIRING NOTES
Heater and a.f. leads run against the chassis. R.F.
leads and components are clear of the chassis. The
wires from T2 run down through a grommet (See
Fig. 4). The rectifier, R4, C4, RIO and several other
items and leads are anchored on a 9-way tagstrip,
as shown in Fig. 5. A small tagstrip anchors the
positive leads C6 and C9. On top, TCI is soldered
to the tags of a similar strip, this being bolted above
the Aerial-Earth sockets.
If difficulty is experienced in obtaining a 9-way tag-
strip having earthed tags in the positions shown in
Fig. 5, a suitable alternative is a 9-way tagstrip with
the two end tags earthed, this being available from
Henry's Radio, Ltd.
The mains lead is 3-core flex, taken to a 3-pin plug
with 3 amp fuse. With the new mains lead coding,
use green and yellow striped for earth, this being Looking down on the top of the chassis
JANUARY 1970 355
 For chassis wiring, use insulated wire or sleeving
where necessary. The connecting wire in the r.f. cir- TABLE III
cuit around the coil and the variable capacitors
should be 20 s.w.g. or thicker. All r.f. leads are short Frequency/Dial Readings
and direct. Wiring details should be clear from Figs.
4 and 5. In these diagrams, "MC" indicates earthing Range 1. Range 4^
tags which are bolted securely to chassis. The tag 150 kHz 100 5.0 MHz 100
of VR1 which connects to chassis is that correspond- 160 93 5.5 83
ing to the maximum anti-clockwise position of the 170 78 6.0 71
slider. 180 70 6.5 61
200 58 7.0 52
EXTERNAL CONNECTIONS 220 49 7.5 46
240 40 8.0 41
The receiver should not be operated without a 260 35 9.0 33
speaker or suitable alternative load connected across 300 27 10 27
the output, as high a.f. voltages may then appear in 350 21 11 22
the primary circuit of T1. For weak signals or personal 400 17 12 19
short wave reception, phones may be connected in 500 10 13 16
parallel with the speaker. If it is intended to occasion- 14 14
ally use phones on their own, a 100 1 watt resistor 15 11
should be permanently connected across the speaker Range 2. 16 9
socket terminals to function as a load. This resistor 17 7
will not cause any significant reduction in speaker 500 kHz 100 18 5
output. Low impedance phones will give greatest 550 83
sensitivity but remember to keep the a.f. gain control 600 71
at a low setting when searching for signals, because 650 61 Range 5.
of the high power available from V2(b). 700 53
Since the h.t. voltage depends on the mains voltage 750 46 15 MHz 52
and several other factors, a voltage check can be 800 42 16 45
made across C9 if wished. The voltage here should 900 33 17 41
be about 180 to 200 volts, and RIO can, if necessary, 1,000 27 18 37
be changed in value to obtain this figure. 1,100 22 20 31
The aerial will probably depend upon what is avail- 1,200 20 22 26
able—such as an end-connected wire for all bands, or 1,300 17 24 22
perhaps an aerial made for one band only. Aerial 1,400 14 26 19
tuners of the usual type may be connected between 1,500 12 28 17
aerial and receiver. 1,600 9 30 14
The receiver will, of course, have an earth con- 32 12
nection via the mains socket, but it is desirable to 34 10
augment this with a more efficient (for r.f.) earth con- Range 3. 36 8
nection run over the shortest route possible. A good 1.5 MHz 100
earth connection can give a considerable improve- 1.6 88
ment with weak low frequency band signals. 1.7 78
1.8 70
BAND COVERAGE AND TUNING 1.9 63
2.0 57
As already mentioned, five plug-in coils are avail- 2.5 37
able, the actual frequency coverage obtained being 3.0 27
listed in Table I. 3.5 20
3.8 16
4.0 14
4.5 10
5.0 6
p mT-*jsfii
v ifflL -JV I Coverage at the low frequency end of each range
w
Jh MS depends considerably on the position of the adjust-
1 1 1
able cores. The full swing of VC1 is not used for
- Jp il Range 5. Ranges 1 and 2 are the usual long wave
> ^ / / and medium wave sectors, while Range 3 is very use-
ful for certain amateur and other transmissions. A
very large number of short wave transmissions fall
into Range 4. Range 5 is primarily for hi. bands
capable of long distance propagation, but these are
much subjected to seasonal and other variations.
" * As a guide. Table III gives frequencies and dial
Rear upper view. The two panel-strengthening readings obtained with the prototype with VC2 at
brackets are clearly visible. half-capacitance. VC1 is fully closed when its cursor
356 THE RADIO CONSTRUCTOR
 3-core
nams lead
Speaker
em
Co
To sec.
Q To TC|
C6

primary

C5
?IO ToX CP
K

R3 mc mc
ToC, ^7- mc
C3
Cft TC
vc.
Mains ' Slider
lead L" r
VR2 vRj
VC3

Fig. 5. Lower view of the chassis. The three valveholders should be mounted
with the pin orientation shown here

is at 100 on the scale. Slight differences at the high eration sufficiently to stop this whistle. The most sen-
frequency end of the ranges may occur, due mainly sitive condition for a.m. reception is obtained when
to varying values chosen for VC2. VC3 is very slightly opened from the setting which
Bandspreading with the aid of VC2 does not in- causes oscillation and the appearance of heterodynes.
crease selectivity or the actual separation between If VR1 is advanced too far, VC3 will need to be
stations, but it does help by causing a very narrow near minimum capacitance to stop oscillation, and its
band of frequencies to be covered with full 180 adjustment will be critical and difficult. Should TC2
degree rotation. insert too much capacitance, oscillation is likely even
VC1 is used for general tuning, to find a band, with VC3 at minimum. But should TC2 insert too
or to set the cursor to a previously logged spot. little capacitance, the benefit it gives in easing adjust-
During this time VC2 may be left half-closed (50 on ment of VC3 is largely lost. The optimum setting
its scale). Signals at nearby higher and lower fre- for this trimmer will soon be found after a little
quencies can then be tuned in with VC2. experience.
Regeneration is influenced by the position of the
REGENERATION coil cores. If these are set up for the range coverages
given in Table III they will also be in suitable
Correct regeneration is the point upon which positions for good regeneration. The cores should be
virtually the whole of the receiver selectivity and sen- locked, with 6 BA nuts passed over the protruding
sitivity depends. VR1 has a considerable effect on threaded stems, to retain calibration.
regeneration, due to its changing the screen-grid With certain aerials an abrupt loss of regeneration
voltage of VI. Closing VC3 increases regenerative may arise at some frequencies, this being due to
feedback, as does increasing the value of TC2. VC3 resonances in the aerial system. Should this occur,
and VR1 are never merely turned to the maximum TCI is adjusted to clear the effect.
clockwise position, as is possible with an audio or There should be no great difficulty in setting up
r.f. gain control. TC2 and VR1 so that VC3 provides an extremely
Initially set TC2 and VC3 to about half their smooth control over regeneration. Once adjusted,
maximum capacitance. Rotate VR1 from its mini- TC2 can be left. VR1, on the other hand, will have
mum (anti-clockwise) position until a heterodyne to be readjusted when making considerable changes
whistle is heard when VC1 is tuned through a trans- in frequency, or after changing coils.
mission. Opening VC3 slightly should reduce regen- For c.w. reception, it is necessary to maintain VI
JANUARY 1970
357
 COMPONENTS
Resistors T1 Output transformer, 40; 1 (Home
(All fixed values 4 watt 10% unless otherwise Radio Cat. No. T043)
stated) T2 Mains transformer, secondaries 230V
R1 1MQ 45mA and 6.3 V 1.5A (Home Radio
R2 47kQ 1 watt Cat. No. TM26A)
R3 68ka Valves
R4 22kQ VI 6BR7
R5 2.2kQ V2 ECL82
R6 220kQ
R7 IMa Rectifier
R8 2.7k 1 watt D1 250V (1,000 p.i.v.) silicon rectifier,
R9 330Q 1 watt BY100 or equivalent
R10 470n 2 watt (see text) Switch
VR1 25kf2 potentiometer, linear wire- SI s.p.s.t., part of VR2
wound, 1 watt (Home Radio Cat. Valveholders, Sockets
No. VR25) 2 B9A valveholders with skirts (for VI
VR2 500k Q potentiometer, log, with and V2)
switch SI 1 B9A valveholder without skirt (for LI)
1 Socket strip, "Ae/E" (Home Radio
Capacitors Cat. No. Z101A)
C1 50pF silver-mica 1 Socket strip, "LS" (Home Radio Cat.
C2 0.05,xF paper or plastic foil, No. Z101C)
250V wkg
C3 0.001/iF paper or plastic foil, Miscellaneous
350V wkg. 1 chassis, 6 x 10 x 2in. (Home Radio Cat.
C4 4/j.F electrolytic, 350V wkg. No. CU26)
C5 0.001 /xF paper or plastic foil, 1 plate (for panel), 6 x lOin. (Home
350V wkg. Radio Cat. No. CU180)
C6 32/xF electrolytic, 350V wkg. 2 brackets, made from single flanged
C7 0.01/xF paper or plastic foil, chassis side, 3 x 4in. (Home Radio
350V wkg. Cat. No. CU144)
C8 25pF electrolytic, 25V wkg. 1 epicyclic ball drive, 6:1, type 4511/F
C9 8fiF electrolytic, 350V wkg. (Jackson Bros.)*
VC1 365pF variable, Jackson type 01 * If desired, a second ball drive may be em-
(Home Radio Cat. No. VC1A) ployed with VC2 (see text).
VC2 (see text)
VC3 50pF variable, Jackson type C804 2 knobs 1 7/16in. dia. (Home Radio Cat.
TCI 60pF mica trimmer No. KN63)
TC2 60pF mica trimmer 3 knobs 1 l/8in. dia. (Home Radio Cat. No.
KN64)
Inductors 2 cursors or pointers (home constructed)
LI Denco Miniature Dual-Purpose 2 3-way tagstrips, centre tag earthed
(Valve) coil. Green, ranges as required 1 9-way tagstrip, 2 tags earthed (see text)
(see Table I) Tuning scales, from Data Panel-Signs,
L2 2.6mH r.f. choke, type RFC5 (Denco) Set No. 5.

in the oscillating condition and slightly off-set the similar process, the tuning being adjusted to the miss-
tuning to produce a heterodyne. S.S.B. transmissions ing carrier frequency.
can be received above certain signal levels by a |

300
£

-40
V n

358 THE RADIO CONSTRUCTOR

 is no variable tone control since
VC2 has become a reaction control,
DESIGN FOR A CLASS A BATTERY but C6 still plays its part by pre-
venting a slight change in tuning or
AMPLIFIER reaction as VR2 is adjusted when its
(Continued from page 351) slider is at or near the maximum
volume position. If C6 is removed
from circuit this effect can be rather
annoying and is due to small internal
load can be used with considerable capacitances within VR2 which are
LI and L2 are wound, over a easily swamped by C6.
voltage amplification. It will be seen paper sleeve, on a ferrite rod Sin.
that, with SI in position B, the long with a diameter of ^in. LI con-
330kS2 and resistor R2, used by the AUTHOR'S PROTOTYPE
sists of 70 turns of 32 s.w.g.
amplifier on its own as part of the enamelled wire close-wound, whilst The author's prototype, which has
tone control arrangements, now be- L2 has 9 turns of the same gauge been through all the stages described
comes the collector load for TR1. wire, close-wound and spaced by in this article, and many others
TR1 will not function as a com- about iin. from LI. The windings are which were abandoned for one rea-
mon collector radio frequency am- phased as shown in the circuit dia- son or another, now takes the form
plifier unless its collector is at earth gram, the same ends of both con- of Fig. 3. It is built into a wooden
potential for radio frequencies. With necting to TR1 base and C3 respec- case which houses an Sin. round
VC2 fully open, therefore, only a tively.
very small proportion of the signal loudspeaker, and the 12 U2 cells.
It will be seen that bias for the These are kept in three plastic con-
across the tuned circut will find it- base of TR1 is taken from its col- tainers, for four cells each, made
self across the diode, such propor- lector circuit. This provides a by Eagle Products. An external
tion as does appear being due to measure of d.c. feedback and is
to the minimum capacitance of VC2. speaker is frequently used. Results
preferable to the use of a separate from records are extremely good,
But as the vanes of VC2 are engaged potentiometer network. (The BC and the amplifier is fully loaded
TR1 then starts to function correct- 168C specified for TR1, together by an input of the order of 250mV,
ly. L2 is a small coupling coil with the 40468A f.e.t., is available the exact sensitivity depending on
which allows feedback to occur from Amatronix Ltd.) the characteristics of the transistors
between the emitter and base cir- It is assumed that the correct used. Local stations give splendid re-
cuits, and VR1 is set so that VC2 connections for the secondary wind- sults on either internal or external
functions as a reaction control. To ing of T1 have already been made speaker and 20 or 30 other stations
set VR1, the vanes of VC1 and VC2 to ensure that negative, as against can be received at reduced volume
should both be fully enmeshed and positive feedback, takes place in the after dark.
VRI next adjusted so that oscillation amplifier part of the circuit. How-
just starts, denoted by a hiss, or a It has been found that during
ever, even when this has been done the hours of daylight a number of
whistle if a station is present at the it may be found that a different stations which are scarcely loud
low frequency end of the band. It form of instability takes place when
will then be found that with the enough for normal listening will
switched to radio. The instability produce an amply sufficient signal
vanes of VC1 fully open oscillation may only be noticeable when VC2 for recording on tape. At the
will start with the vanes of VC2 is adjusted to the point of oscilla-
about half way in. This setting will author's home in South Devon this
tion and, if it does occur, it indi- applies particularly to Radio 3 on
remain fairly constant for about cates unwanted coupling between T1 194 metres, which produces a weak
the first half of the movement of and the ferrite rod assembly. It but interference-free signal during
VC1 but, as the vanes of this capaci- can be cured by reversing the con-
tor are further enmeshed, it will be daylight hours.
nections to both the windings of Tl. Layout may follow the construc-
found necessary to further increase Both sets of connections have to be
the value of VC2 to obtain oscilla- tor's whim provided that input and
changed over because reversal of output circuits are sensibly separ-
tion. This applies when a solid the connections to one winding only
dielectric capacitor is employed for ated. It is especially important to
would introduce positive feedback keep Tl well away from the ferrite
VC1. If an air-spaced component is in the amplifier.
used, the setting of VC2 will remain rod assembly, and to arrange for the
C6 has been mentioned with re- axis of the coils of Tl to be at right
rather more constant throughout the spect to its function as part of the
tuning range. angles to the axis of LI and L2.
tone control network. On radio there

CAN ANYONE HELPt

w
^ * •Mxsfs
R3673.—A. Howard, 2 Castle Close, Reffley Estate, Communications Receiver Type G12.—B. Lawrence,
Kings Lynn, Norfolk—circuit, manual, modification 32 Bridle Path, Beddington, Croydon, CRO 4SB—
details or any other information. circuit diagram, manual, or any other information.
Radio Altimeter APN1.—K. K. Lloyd, P.O. Box Loan or purchase.
1164, Lusaka, Zambia—circuit or manual.

JANUARY 1970
359
 CONSTANT CURRENT GENERATOR

FOR OHMMETER CONVERSION

by
J. C. MAYCOCK

The constant current generator circuit described here enables resistance to be

measured directly with the aid of a high impedance voltmeter. If the voltmeter
has sufficient sensitivity, very low resistance readings are feasible

There must be many constructors who have The circuit of the constant current generator is
built themselves high input impedance voltmeters shown in Fig. 1 with Rx connected across the output
and who lack the means to measure resistance. terminals.
This article describes a simple unit which can be
employed with a high resistance voltmeter to enable CIRCUIT OPERATION
resistance measurements to be carried out. It requires
no modifications to the existing voltmeter. Also, it The unknown resistor Rx forms the collector load
causes only a low current to be passed through the of silicon transistor TR2, this operating from a 15
resistance being measured and is therefore suitable volt supply in what is effectively a grounded base
for all but the most delicate of components. configuration. A large change in collector voltage
The basis of the technique consists of feeding a produces a relatively small change in collector current,
known current through the unknown resistor, Rx. and therefore the current through Rx remains con-
The potential developed across Rx is then a measure stant for a very wide variation in its value.
of its resistance, and can be read off from the scale The collector current of TR2 is controlled by the
of a high impedance voltmeter. emitter current. The latter is adjusted to give load

2 r
•— r WW2
Range
switch
VR 3 R3
I \^/
t-—mt ww
R
| JVR4 4

ZD, ' TR,

0C2O3 TR
SZ56A 0C203
15V D.C.

(0^0^ dot
0C203
Lead-outs
Fig. 1. Circuit of the constant current generator. Rx is the
resistor whose value is to be measured
360 THE RADIO CONSTRUCTOR
T
Constant ^
I5V D.C. current Voltmeter COMPONENTS
generator ^
_
Resistors
Fig. 2. Showing correct connections for the (All fixed values i watt 5 %)
two units required R1 33012 high-stability
R2 3.3kQ high-stability
currents in Rx of 10mA, 1mA, 100/j.A and 10/jA by R3 33kQ high-stability
means of range switch SI and preset controls VR1, R4 330kO high-stability
VR2, VR3 and VR4 respectively. R5 ikn
To make the emitter-collector current independent VR1 100n wirewound potentiometer
of fluctuations in the supply, the base voltage is stabil- VR2 IkO wirewound potentiometer
ised by zener diode ZD1. A 5.6 volt zener diode VR3 lOkQ wirewound potentiometer
rated at 1 watt was used. Should a stabilised power VR4 lOOkO wirewound potentiometer
supply be available, then the zener diode may be
replaced by a 56012 resistor. The maximum current Semiconductors
drawn from the 15 volt supply is approximately 20mA. TR1 OC203
The potentiometers VR1 to VR4 should all be wire- TR2 OC203
wound and, sufficient to say, of good quality. Colvern ZD1 SZ56A (Henry's Radio, Ltd.)
1 watt types were used in the original. Switch
SI single pole 4-way, rotary
VOLTMETER
The voltmeter used with this circuit should be
either a valve voltmeter or the transistor equivalent. reading would be 2Mn. A range switch offering a
A digital voltmeter is ideally suited and was in fact fifth position would also, of course, be required.
used with the original circuit. A 12 volt supply may be employed instead of the
15 volt supply, if desired. In this case, the values of
RANGE OF THE UNIT the ranging resistors should be changed to multiples
of 22, commencing with 2200 for R1 and ending
The voltage appearing across the output terminals with 220kfi for R4. R5 should then be 82012 and
under open circuit conditions is about 10 volts. the zener diode a 4.7 volt component (or a 4700
The maximum effective voltage for which the resistor if a stabilised supply is used).
present circuit is intended is about 2 volts, since a N,
fall-off in accuracy exists above this voltage. Range CALIBRATION
calculations may be based on this. Thus, 2 volts and
lOfiA give a maximum measurable resistance on Calibration is best carried out using standard
Range 4 of 200kf2. The accompanying Table gives resistors of known values and adjusting the potentio-
details of the maximum resistance reading offered, meters on each range until the voltage indicated is
at 2 volts, by each range. Since the resistance values the required multiple or sub-multiple of the resistance.
are multiples of voltage readings, the value of any
resistance being measured may be quickly judged
from the corresponding voltmeter indication. For Constant ±
example, a voltage reading of 1.5 on Range 3 corres- 15V D.C. current Voltmeter
ponds to a resistance in Rx of 15kf). The minimum generator
useful resistance reading for each range depends upon \
the lowest voltage the voltmeter can reliably measure.
Should a range higher than 200kf2 be required, a
further range offering 1 /xA could be added, although Fig. 3. Showing errors that may occur through
this has not been tried out in practice by the author. incorrect connection of leads
The values of the preset potentiometer and range
resistor in the emitter circuit of TR2 would be 1MQ TRANSISTORS
and 3.3Mf2 respectively, and the maximum resistance
Any silicon transistors in the series OC200, OC201,
OC202 and BZY11 may be used with success in the
TABLE circuit, though differences in stability may be found.
The current stability of the circuit is about 0.05%
Maximum Resistance and thus the accuracy is similar.
Range Load Current Reading (for 2 Volts
across Rx) USING THE UNIT
1 10mA 2000 On the low ohms range trouble can be experienced
2 with lead resistance unless the voltage sensing output
1mA 2kn and current input are both connected directly to
3 lOOpt A 20k£2 resistor Rx, as in Fig. 2. Fig. 3 shows an incorrect
4 10/7, A 200kO method of connection which causes the resistances
of the leads "a" and "b" to be added in series with Rx.
JANUARY 1970 361
 f=
2*VLC
UNDERSTANDING

METER SHUNTS

by W. G, Motley

IN LAST MONTH'S ISSUE WE EXAMINED THE MECH- appear across the shunt. The current flowing in the
anism of the pointer measuring instrument which shunt is 99mA so, from R = E, the shunt resistance
is used far more frequently than any other type T
in electronic work, this being the moving-coil meter. must be 1 volt divided by 99 amp, which works
We concluded by introducing the concept of the
shunt, this being a resistor whose function is to cause 10 L000
the meter to indicate a higher current than that which out as 100 12.
it indicates on its own. 99
We shall now deal with the subject of shunts in This simple example shows us that the shunt re-
greater detail. K sistance required is exactly one ninety-ninth part of
the meter resistance. It also demonstrates that there
SHUNT RESISTORS is no necessity to know the voltage dropped across
the meter at full-scale deflection, since the same volt-
In the introductory example we discussed last age is also dropped across the meter and shunt com-
month, we had a 0-lmA moving-coil meter which we bined. To find the value of any shunt, all we need to
wished to give a full-scale deflection of 100mA, and know is how many times greater than the meter f.s.d.
we connected a shunt across it as shown in Fig. 1. value is the current it has to pass when the meter in-
(In this and succeeding diagrams, the arrows indicate dicates full-scale deflection. Fig. 2(c) shows a 0-lmA
the flow of "conventional" current, i.e. from positive meter having a shunt whose function is to cause the
to negative.) As we saw last month, we give the shunt meter to read f.s.d. when 10mA flows through the
a value which causes 99mA to flow through it when combination of shunt and meter. Obviously, the shunt
100mA flows through the combination of shunt and has to pass 9mA, whereupon it follows that its re-
meter. The remaining 1mA must then flow through sistance needs to be one-ninth of the resistance of the
the meter, with the result that it gives an indication meter. If the meter resistance in Fig. 2(c) is 10012
of this value when the actual current flowing through then the shunt resistance required is 100 Q. Should
the combination of meter and shunt resistor is ~
100mA. The same relationship holds true for meter the meter resistance be 7512, the shunt resistance re-
scale indications lower than 1mA. If these are multi- quired is_75 n, and so on.
plied by 100 they will show the actual current flow- ~9
ing through the combination of meter and shunt. It is desirable to employ a common equation to
To find the value of resistance needed in the shunt cover all simple shunt resistance calculations and this
it is necessary, first of all, to know the resistance of may be conveniently based on the "scale multiplying
the meter (i.e. the resistance offered at the meter ter- factor" which results from the addition of the shunt.
minals due to its coil and the connections to that
coil). Let us say that the resistance of the meter in
our example is 1000. We may next consider the volt-
age dropped across the meter when, on its own, it is lOOmA'
indicating 1mA. From the equation associated with
Ohm's Law, E = IR (where E is e.m.f. in volts, I is
current in amps and R is resistance in ohms), this ImA 99mA
voltage must be 1 x 100, or 1 volt, and it is
1,000 10 O-lmA
indicated in Fig. 2(a). In Fig. 2(b) we add the shunt 0 : Shunt
and increase the current flow to 100mA whereupon
the current flowing through the meter is once more
1mA. Again, J_volt will be dropped across the meter
10
and it is obvious, by inspection, that J_ volt must also
Fig. 1. Connecting a shunt across a 0—1mA
10 meter to cause it to read 100mA f.s.d.
JANUARY 1970 363
In the two examples we have considered the scale where Rs is shunt resistance, Rm is meter resistance
multiplying factors have been 100 (1mA to 100mA) and N is scale multiplying factor.
and 10 (1mA to 10mA). The actual current flowing The writer has introduced this equation by way of
through the shunt for f.s.d. is then the current in the several intermediate steps because it has been his ex-
meter multiplied by the scale multiplying factor minus perience that beginners tend to be confused by the
one. Thus, a scale multiplying factor of 100 means appearance of the "minus one" on the right hand
that the current in the shunt has to be 100 minus 1, side. It is, of course, necessary to include the "minus
or 99, times meter current, this agreeing with what we one" in order to allow for the current flowing in the
have already seen. meter. To reiterate, the equation tells us that for, say,
The general equation for calculating shunt re- a scale multiplying factor of 10, the shunt resistance
sistance is: is the meter resistance divided by 10-1, or 9. Note
Rs = Rm that there is no necessity to include in the equation
• N-i the actual currents passed by the meter or the shunt
as these are automatically catered for by working in
terms of the "scale multiplying factor".
ImA An interesting case arises when it is desired to have
a scale multiplying factor of 2, as shown in Fig. 2(d).
This will occur if we want, say, a 0-lmA meter to
lOOn ( \ VlOvolt read 2mA f.s.d. (as is indicated in the diagram), a
0-50mA meter to read 100mA f.s.d., or any other
J meter to read twice its scale value. The term N in
the general equation for shunt resistance now be-
comes 2, whereupon we have
Rs = Rm
C<0
ITT
giving us the result that Rs is equal to Rm. This is
lOOmA to be expected since, for a scale multiplying factor of
2, equal currents must flow in the meter and in the
99mA ] shunt.
lOOn /IO volt
0 MULTIPLE SHUNTS
A very large number of measurements in electronic
work are carried out with the aid of an instrument
(b) which is usually referred to as a "testmeter" or a
"multi-testmeter". This instrument employs a single
basic moving-coil meter and, by means of suitable
lOmA switches and/or sockets, can be set up to give an f.s.d.
ImA 9mA 1 indication for a wide range of different currents and
voltages. Testmeters of this nature also have one or
O-lmA^V^ more ranges capable of indicating resistance values.
: Shunt So far as the current measuring function of a test-
meter is concerned, it would appear that a number of
different current ranges could be obtained by the
simple expedient of connecting different values of
shunt resistance across the single basic meter. A sys-
Cc) tem of this nature is in fact used, but the circuit em-
ployed has to incorporate features which are designed
2mA to protect the basic meter against accidental overload
and to reduce the risk of errors in readings on the
ImA ImAf higher current ranges.
O-lmA^y^ At first sight, it would seem that the switching
: Shunt circuit required is that shown in Fig. 3. In this dia-
gram, the basic meter is coupled all the time to the
terminals of the testmeter, and the range selector
switch simply brings into use the shunt required for
the particular range selected. At one switch position
(d) there is no shunt at all. This would correspond to the
lowest current range, and the readings indicated
Fig. 2(a). Indicating the voltage dropped would be those given by the basic meter on its own.
across the meter of Fig. 1 at f.s.d. when the In practice, the circuit of Fig. 3 represents ex-
meter has a resistance of 100n tremely poor design approach, and should not be em-
(b). The same voltage appears across the ployed in a working testmeter. The reason for this is
shunt resistor when the overall current flow that it is possible for excessively high currents to pass
is 100mA through the basic meter due to faulty operation of
(c). Adding a shunt to make a 0—1mA meter the range switch. To take an example, let us assume
read 10mA f.s.d. that the testmeter is switched to a high current range
(d). In this instance, the shunt causes the and that a correspondingly heavy current is being
meter to indicate twice its normal f.s.d. value passed through the instrument. Very nearly all of this
364 THE RADIO CONSTRUCTOR
 current should flow through the shunt selected by the
switch. If, however, the switch contacts happen to Range
become momentarily open-circuit or should momen- selector
tarily present a relatively high resistance, virtually switch
all the current will flow through the basic meter
itself whereupon the latter can be seriously damaged. Testmeter
The possibility of a momentary open-circuit or high terminals
resistance contact is likely to occur in any switch, this M Basic meter
being particularly true when the switch is being R3
turned from one range position to the next. A further
disadvantage is that the inevitable resistance offered
by the switch contacts appears in series with the
shunt. Shunts for high current ranges may have re- (a)
sistances considerably lower than 1ft, and the con-
sequent risk of inaccuracies due to even extremely
low values of switch contact resistance becomes very
high.
These disadvantages can be overcome in testmeter Testmeter
current range switching circuit design by following termma s
the principle that currents greater than the f.s.d. value
of the basic meter must always flow directly through \) Basic meter
the appropriate shunt, and that the meter should then
be connected to that shunt. This principle is followed UU
in the simple circuit shown in Fig. 4(a). Here, there
are three current ranges and the current to be Shunts
measured flows through Rl, R2 and R3 in series. The
highest current range is selected by position 3 of the S,,S2- range selector switch
switch, whereupon the appropriate shunt is R3. On
position 2 of the switch, the shunt is provided by R3 Cb)
and R2 in series. Position 1 of the switch selects the Fig. 4(a). A circuit which, although not prac-
lowest current range, the shunt being given by R3, ticable for reasons explained in the text,
R2 and Rl in series. It will be seen that only the demonstrates the preferred method of switching
meter current flows through the switch contacts, with shunts in a testmeter
the result that errors in meter indication due to switch (b). A range switching circuit which is suit-
contact resistance become correspondingly reduced. able for use in a testmeter. S1 and S2 are
In no instance, also, can open-circuit or high re- ganged
sistance switch contacts allow excessive current to
flow accidentally through the meter. Whilst Fig. 4(a) pass through the shunt, the meter being then con-
ably demonstrates the correct approach for switching nected to that shunt. Only the relatively low meter
shunt resistors, the circuit has the disadvantage that, current flows through S2. The current being measured
due to the relatively high resistances of Rl and, in passes through SI, but the accuracy of the instrument
turn, R2, excessively large voltages will be dropped will not be impaired if this switch should exhibit a
across the chain of resistors when measuring the small contact resistance. Such a contact resistance
larger currents. merely results in a slight increase in the voltage
An alternative switching circuit is illustrated in dropped across the terminals of the testmeter as a
Fig. 4(b), in which the two switches shown are whole. The circuit has the slight further advantage
ganged. The current to be measured, if higher than of enabling the test terminals to be applied direct
the f.s.d. value of the basic meter, passes first through to the basic meter on the lowest current range (at
SI and then flows through the shunt selected. The switch position 1 in the diagram) with no shunt in
meter is then coupled by S2 to the same shunt. The circuit at all. The only disadvantage of the circuit is
circuit meets all the requirements for correct design that it requires a range switch having two ganged
procedure. The current to be measured must always banks of contacts.
UNIVERSAL SHUNT CIRCUIT
A current range selection circuit which embodies
nearly all the advantages of that shown in Fig. 4(b),
+ and which has the added advantage that a single-gang
Testmeter Range selector
Basic meter switch switch with one bank of contacts can be used, is
terminals shown in Fig. 5(a). This is known as the universal
©
shunt circuit. It is shown with four switch positions
in Fig. 5(a), but the principle may be readily adapted
for fewer or more positions.*
The lowest current to be measured is given by the
switch in position 1, whereupon the basic meter is
shunted by Rl, R2 and R4 in series. The next higher
Shunts
Fig. 3. This method of current range selection * As is explained later in this article, the basic meter will have an
external resistor in series with one of its terminals, whereupon the
in a testmeter may appear attractive at first effective meter resistance presented to the universal shunt circuit is
sight but in practice it represents a poor design equal to the actual meter resistance plus this external resistance.
approach and its use is not advised For the present discussion it can be assumed that the meter re-
sistance is that provided by the meter on its own.
JANUARY 1970 365
 current range is provided by position 2, whereupon
the shunt is given by R2, R3 and R4 in series, with
R1 appearing in series with the meter. The resulting 0-2mA
circuit is shown in Fig. 5(b), and it should be noted
that, so far as the shunt is concerned, the effective Range 2f04mA -1mA
meter resistance is the actual meter resistance plus selector Rs G0lOOn
.
Rl. The values of R2, R3 and R4 are such as to take switch
this fact into account. On the next higher current r
2 V2Rs
range, selected by moving the switch to position 3,
R3 and R4 provide the shunt whilst the effective
meter resistance is the actual meter resistance plus
Rl and R2. On the highest current range, given at R| -SOn
switch position 4, R4 is the shunt and the effective R2- SOn
meter resistance is the actual resistance plus Rl, R2
and R3. Fig. 6. A very simple universal shunt circuit
Calculations for resistor values in the universal
shunt circuit are complicated by the fact that, in is automatically added in series with the meter. Thus,
switching to a higher current range, further resistance the values of resistors in the shunt and series sec-
tions for each switch position tend to be interdepen-
dent. Fortunately, calculations for universal shunt re-,
sistor values may be considerably eased by following
the general procedure illustrated in Fig. 5(c), which
assumes that each switch position increases the
current f.s.d. reading by 10 times. It is first of all
Testmeter necessary to determine the lowest current range to
terminals Range Basic meter be used and calculate the total value required for Rl,
selector R2, R3 and R4 in series. This value is then referred
switch to as Rs. Switch position 2 multiplies the f.s.d. read-
ing by 10. For the circuit to function correctly, the
sum of R2, R3 and R4 must be equal to _1_ Rs.
10
The process continues to R4, which needs, in this
example, to be 1 Rs. Since Rs is known the
(a)
1,000
R
I value of R4 may then be determined. The value of
-W- R3 then becomes 1 Rs minus the value of R4,
loo
the value of R2 becomes _1_ Rs minus R3 and R4
Testmeter 1(1
terminals Basic meter in series, and the value of Rl becomes Rs minus R2,
0 R3 and R4 in series. It will be found that calculations
along these lines can be applied to any number of
resistors employed in a universal shunt chain.
The scale multiplying factor, as the switch moves
from one range to the next, need not be 10, and any
CO other factor can be used, as is illustrated in the simple
numerical example shown in Fig. 6. This depicts a
universal shunt circuit where position 2 of the switch
doubles the f.s.d. reading of the meter. For simplicity
it is assumed that the meter has an f.s.d. of 1mA and
a resistance of 100(2. It is required that the lowest
current range, given on switch position 1, will be
0-2mA. The value of Rs is then equal to meter re-
xlO sistance, i.e. 100(2. On position 2 the f.s.d. value is
doubled whereupon the value of R2 needs to be one-
x IOO half of Rs. This is 50(2. Since Rs is 100(2, Rl must
xl.OOO TOR also be 50(2.
toor On switch position 1 in Fig. 6, therefore, we have
a 1mA meter of 100(2 resistance with a shunt (Rl
R4S VlpOORs plus R2) of 100(2 across it. The f.s.d. reading is thus
-1 L_J ! 2mA. On position 2 we have a 50(2 shunt (R2) and an
effective meter resistance (the actual meter resistance
(C) plus Rl) of 150(2. The result is that three times more
current must pass through R2 than through the meter
Fig. 5(a). An example of the universal shunt circuit and Rl. If 4mA is applied, 3mA will flow through R2
(b). The circuit conditions existing when the and the remaining 1mA through the meter and Rl,
range selector switch is in position 2 whereupon the f.s.d. of the instrument becomes 4mA,
(c). Illustrating the procedure required for as is desired.
finding shunt values The universal shunt circuit is incorporated in vir-
366 THE RADIO CONSTRUCTOR
 Sockets
4 SWAMP RESISTOR
Range I The shunt resistors employed in testmeters are
made up with resistive materials (usually one of the
Range 2 resistance wire alloys) having a very low tempera-
ture coefficient of resistance. The coil of the moving-
:r2 coil meter has, on the other hand, to be wound with
Range 3 t>— Basic meter copper wire, and copper wire has the relatively high
temperature coefficient of 0.0039 per degree Centi-
1 grade. As a result, the resistance of the meter coil
Range 4 increases with ambient temperature and, possibly,
with any heating effect due to the current which
Common passes through it. Changes in coil resistance due to
negative temperature variations can be surprisingly high. For
Fig. 7. The simplicity of the universal shunt instance, an increase in wire temperature of 13°C
circuit makes it possible to dispense with the results in an increase in winding resistance of 13 x
range selector switch and use sockets instead. 0.0039, which is equal to 0.05, or 5%. Variations
This is a version of Fig. 5(a) with sockets such as this can cause serious inaccuracies in test-
tually all conventional commercially manufactured meter readings if the basic meter is connected directly
testmeters. Its advantages have already been referred to the current shunts.
to. It has the disadvantage that, since resistance is
inserted in series with the meter when higher current
ranges are selected, the voltage drop across the instru-
ment as a whole is somewhat greater than in a circuit
similar to that of Fig. 4(b). If readers care to work Swamp
the figures out, they will find that the resistance resistor
offered by the circuit of Fig. 6 with the switch in Testmeter J
position 2 is 37.512 whereas, if a basic meter of the terminals
same type had been made to read 4mA f.s.d. by the
provision of a single shunt connected directly across Basic meter
its terminals, the resistance offered to the external
circuit would be 2512. Thus, the universal shunt :r4
circuit causes rather greater power to be absorbed
from the circuit into which it is connected than does
that of Fig. 4(b). A slight disadvantage is that no Fig. 8. To reduce the effect of errors due to
range is available which connects directly to the variations in meter- resistance with changes in
meter, thereby preventing the minimum possible coil temperature, a testmeter incorporates a
f.s.d. reading from being available at the testmeter swamp resistor in series with the basic meter.
Here, the swamp resistor has been added to
terminals. In practice,, this disadvantage is not of the circuit of Fig. 5(a)
great consequence, and it is usual to make Rs such
that the minimum current range selected by the This problem is overcome in a testmeter by insert-
switch is only some 20 to 35% greater than the actual ing a swamp resistor in series with the basic meter, as
f.s.d. rating of the basic meter. shown in Fig. 8. The swamp resistor has a value
The simplicity of the universal shunt circuit makes which is some 3 to 6 times the resistance of the meter
it possible to dispense with the range selection switch coil and employs resistance wire having a negligibly
altogether. Instead, a number of sockets may be pro- low temperature coefficient. Changes in the resistance
vided, as illustrated in Fig. 7. The test leads of the offered by the meter and swamp resistor in series are
instrument are then simply plugged into the sockets then reduced to an acceptably low value, and shunt
corresponding to the particular range required. resistance values are calculated on the assumption

FERRANTI LTD. ENLARGE SEMI-

COMPONENTS
CONDUCTOR ASSEMBLY PLANT
Where the above appears in this magazine, readers
The Electronics Department of Ferranti Ltd. plans can rest assured that each and every item quoted is
to double the size of its semiconductor assembly currently available from most component suppliers.
plant at Ormsgill, Barrow-in-Fumess, Lanes. This ex- Where an item is supplied from an individual source,
pansion, which will double the size of the present the address is stated.
26,000 sq. ft. factory, will provide a production We take great care to ensure that all parts quoted
capacity of more than 20 million semiconductor com- in the various lists are not only obtainable by readers
ponents per annum. The present production capacity at the time of publication but also for some time
at Barrow is nine million components per annum. ahead. Current catalogues are consulted, letters
The first batch of equipment for the new extension written and/or telephone calls made to suppliers con-
will be commissioned in July, and the enlarged plant firming the present and future stock positions during
is scheduled to be fully operational by the Autumn. the pre-publication processing of all articles.
Ferranti's plans for the new extension will provide From time to time however, and unknown to us,
a further employment boost for the Barrow area. manufacturers change specifications of a particular
The present workforce of some 250 workers will be product or, unfortunately from the constructors point
increased to over 700 employees by the end of this of view, cease producing a particular item - this is a
year. fact of life over which we have no control!
JANUARY 1970
367
that the meter resistance is equal to its actual re-
sistance plus the swamp resistance. The swamp re-
sistor is so called because it "swamps out" the varia- f CURRENT SCENE 1
tions in the resistance of the meter coil which result *
from changes in temperature.
Since the only requirement of the swamp resistance
is that its value should be at least several times the TOP BAND DX
resistance of the meter coil, it is possible to give it
a value which results in the total swamp plus meter If you are a sleepy head don't bother to read this
resistance becoming a convenient round number of column!
ohms. This can result in the shunt resistor values If you can clamber - somewhat sleepily maybe - out
being expressed in similarly convenient round num- of that warm, snug and comfortable bed at the un-
bers. A further point follows from the fact that, due earthly hour of 0530 clock time on a Sunday morning,
to variations in thickness within tolerance of the and be ready to enter the shack at 0600BST (0500
GMT) you will be in time for the Top Band Trans-
copper wire employed, it is difficult to produce meter Atlantic Tests!
coil windings to a precise value of resistance, this
being especially true in the case of sensitive meters and 28th of held
The Tests so far were on 30th November, 14th
December. Tests still to be held are on
having a very large number of turns on their wind- 11th January, 1st and 15th of February between 0500
ings. It is possible to take up variations in individual and 0730 (0600 to 0830 BST or clock time).
meter coil resistance by pairing up each meter with a During alternate five minute periods, call "CQ DX
swamp resistor whose value allows the total resistance Test", the W's and VE's commencing at 0500, unless
to have the desired round number of ohms. The pro- in QSO, keep to these periods. Eastern W/VE stations
cess of setting up swamp resistance to give the operate between 1,800 and 1,820kHz; western W/VE's
requisite total resistance is carried out at the factory between 1,975 and 2,000kHz. European stations will
where the testmeter is produced. use 1,823 to 1,830 kHz or 1,851 to 1,861 kHz.
A swamp resistor is also required with the circuit It is not necessary however to confine trans-Atlantic
of Fig. 4(b), and it may be inserted directly in series Dx contacts within the date limits of these tests. Activity
with either terminal of the meter. Once again, shunt "across the pond" has been noted since last October.
resistor values are calculated on the assumption that During the last two weeks of that month, the following
Dx stations had been heard busily making contacts on
meter resistance is the actual meter resistance plus this side of the Atlantic - K1BPW (0525, 1,804kHz);
that of the swamp. W1BB/1 (0515, 1,803kHz) and W1EXI (0545,
A fact which is not generally realised is that, since 1,803kHz). The Bermudan VP9GW (0530, 1,802kHz)
a testmeter incorporates a swamp resistor for its has also been active.
current ranges, the voltage drop across the testmeter By the time these notes appear in print many more
terminals for any current reading is significantly signals will be coming through from across the Atlantic
greater than would be given by a standard moving- with the season moving to its climax. Last year a
coil meter without a shunt which indicates the current veritable spate of Dx was apparent from the W eastern
coastal region and reported in our feature QSX. This
directly. The voltage drop in a testmeter is increased year it is hoped that more stations will join in the tests
further by the fact that the universal shunt circuit is and attempt to span the Atlantic with a successful QSO.
almost always employed. In practice the voltage drop For short wave listeners the tests are a trial of
in a good quality commercially manufactured test- Dx-capability both of the listener and the station equip-
meter is not sufficiently high to cause undue dis- ment, Reception of a W on Top Band represents quite
turbance of operating conditions in any normal a feat of Dx - much more so than some of the trans-
circuit into which the testmeter is inserted, but the missions logged on the 14Mhz band.
effect can be troublesome on occasion and its presence So, if you can make it for opening time - of the tests
needs to be borne in mind. we hasten to add - why not join in the fun and games?
NEXT MONTH
The writer had intended to introduce the use of
the moving-coil meter for the measurement of voltage
in this article, but lack of space now prevents this.
In consequence, we shall turn to the question of
moving-coil voltmeters in next month's issue. gj

GUNN EFFECT - A MULLARD

PAMPHLET
Readers may care to know that an article on Gunn
effect, devices and their applications, originally pub-
lished in Mullard Technical Communications, is now
available as a pamphlet. 160
It contains an explanation of the Gunn effect, des-
cribes the construction of the device and gives advice on
the design of oscillator cavities for Gunn devices. Fac- QSL card from W1BB/1 received for a period
tors affecting stability are discussed and information is reception report sent last year. Of six such
given on noise, frequency locking and pulsed operation. reports sent, each produced a reply in the form
Requests for the pamphlet should be made on com- of a card, three had accompanying appreciative
pany-headed notepaper to I.E.D./Valves Sales, Mullard letters and one a bulletin of Top Band activity
Limited, Mullard House, Torrington Place, London, and news
W.C.I. ■
368 THE RADIO CONSTRUCTOR
This month we find Smithy the Serviceman, aided as always oo
by his able assistant Dick, starting the New Year in a Work-
shop which contains no equipment for repair whatsoever! o oo
n
But Smithy is not one to waste time, and he takes advantage
of the situation to discuss with Dick the latest hints received
from readers

'APPY NEW YEAR, SMITHY!'■ "You didn't," queried Smithy your

"And the same," replied slyly, "get another false beard stuck
"H Smithy affably, "to you on you, did you?"
my boy." Dick winced, then hurried quickly
Dick glanced warily at the Work- over towards the motley array of
shop clock, the hands of which in- utensils ranged alongside the Work-
dicated that normal starting time shop sink. work
had occurred at least ninety minutes "I think," he remarked, hastily
earlier. changing the subject, "that it would
"Sorry I'm late," he offered. be an excellent idea if I put the
"Not to worry, Dick, I've only kettle on for tea."
just got in myself," admitted Smithy. "A very sensible suggestion," res-
"As I told you before we packed up ponded Smithy warmly. "Also, see-
last night, there was no need for ing that we've got a nice quiet
you to come in early today." New Year's Day in front of us,
The pair looked at the 'For Re- how would it be if we had another
pair' racks. They were completely session with readers' hints?"
bare. "That would be fine," said Dick
"Thank goodness," commented eagerly, as he returned to his stool.
Dick, "that there are no sets to be "Good," remarked Smithy, open- built. When he considered the situa-
fixed. All I want to do today is to ing a drawer in his bench and ex- tion, however, he found that wood
take things nice and easy." tracting a sheaf of letters. "Well, I was rather too cumbersome for the
He stretched himself luxuriantly won't waste any more time. We'll housing, whilst Perspex didn't quite
then removed his raincoat and hung have- a look at the first one right seem to fit the bill. And so he used
it up on the hook behind the door. away." copper-clad laminate instead."
The Serviceman opened one of "Copper-clad laminate?"
READERS' HINTS the letters and surveyed its contents "That's right. In other words,
closely. ordinary plain printed circuit board
"Did you," asked Smithy, "have a "Now here's a jolly good hint," before the circuit pattern is etched
good New Year's Eve celebration he continued, "and it's just what on it. This material proved ideal for
last night?" we need to start the ball rolling. the purpose, and all that had to be
"I had quite a pleasant little do," Our correspondent states that he re- done was to cut the front, sides,
replied Dick. "Rather quiet, but cently had to make up a housing top and bottom to size, and secure
otherwise very nice." for a small electronic gadget he'd them together with the copper on
the inside to form a small cabinet."
(Fig. 1).
Copper clad laminate "How did he join the bits to-
gether?"
Smithy consulted the letter once
more.
"He soldered internally at the
corners," he replied. "Our corres-
pondent says it's best to spot solder
the sections together, rather than to
run the solder all along each corner.
The case which results is attractive

TV

Copper surface
on inside Spot solder joints
Fig. 1. An unusual but effective method of assembling a small
equipment case
JANUARY 1970
369
in appearance. In addition, the "I'll say it is," said Dick. "Dash Dutifully Dick picked up his
copper surface on the inside pro- it all, why on earth didn't I think stool, f irried it over to Smithy's
vides screening. Another advantage of that myself?" bench and settled himself on it com-
is that small components can be "The solutions to most problems," fortably. He then looked at the
mounted to the inside walls by pronounced Smithy, "are usually circuit accompanying the letter
soldering them direct to the copper obvious once you've been told what Smithy was holding. (Fig. 2).
surface." they are. A final point which is "As you can see," remarked
"That," remarked Dick, "is what made in this letter, incidentally, is Smithy, "the idea is not at all com-
I call a proper copper-bottomed that the idea of using rubber bands plicated. At the extension control
hint!" to hold components in place can be point you have an on-off switch in
"It's certainly a novel idea," just as useful when you're working series with a 250kO log pot, and
chuckled Smithy, as he picked up with really large boards as it is with these are coupled to the set by twin
another letter. "Now, this letter des- small ones." lighting flex. The volume control in
cribes quite a simple sort of tip, Smithy picked a further letter the receiver is set up to provide
but it's still one which can save a from the pile on his bench. adequate volume level for quiet
lot of time and trouble. As you "Here," he announced after a speech when the switch at the ex-
know, the process of soldering com- moment's perusal, "is a hint which tension point is open. If the sound
ponents to printed circuit boards enables you to fit an extension programme changes over to loud
and to Veroboard may seem easy volume control to a TV set. I'd music the remote switch is then
in principle, but in practice you can better add quickly that the scheme closed, whereupon the extension
sometimes run into the odd minor doesn't apply to all TV receivers volume control can be adjusted to
difficulty." and that, since it involves a direct allow the music to be reproduced
"That's true enough," interrupted connection to the chassis of the at a comfortable level. In most
HT + cases, after the extension volume
control has been initially adjusted to
the desired level, it will only be
necessary to operate the switch.
This is just the thing, I might add,
I IN to have positioned on the arm of
II your favourite armchair."
"Won't there be trouble," asked
Dick, "due to hum pick-up on the
Receiver leads to the extension point?"
volume I A.F output "There wasn't any trouble on this
^ control valve score in our correspondent's set-up,"
A.F voltage
amplifier 500kn replied Smithy. "And he simply
uses about fourteen feet of ordinary
unscreened twin lighting flex. How-
ever, you would very definitely get
trouble with hum if the extension
n volume control circuit were intro-
duced any earlier in the receiver
i a.f. amplifier stages than at the grid
JL Receiver of the output valve. Even when the
chassis extension circuit appears in the out-
put grid circuit it will still be neces-
On-Off switch sary to connect the TV set to the
mains such that its chassis is com-
x mon with the neutral and not the
live side. It would also be pre-
ferable to keep the wire to the ex-
c250kn Twin lighting tension volume control spaced well
log flex away from any other mains wiring
in the room. Actually the circuit
is a wee bit experimental in this
respect because, in houses which are
Remote controls in wired up with p.v.c. cable, it may
insulated case be impossible to completely eradicate
Fig. 2. A simple extension volume control for use with television hum pick-up on the extension lead.
receivers whose existing volume control appears in the grid Apart from the question of picking
circuit of the a.f. output valve. All insulation in the extension up hum, it is extremely important
circuit must be adequate for mains voltage to pay attention to the safety fac-
tor, The extension volume control
components and wiring are all at
Dick. "What happens every now and set, all the circuit points in the ex- mains potential, and the remote
again is that, when you turn the tension wiring require insulation potentiometer and switch must be
board over to solder the lead-outs which is fully suitable for mains completely enclosed in an insulated
to the copper, one or more of the voltages." case which ensures that none of the
components drops out or moves out "What sort of TV set," asked connection points can be touched.
of its proper position." Dick, "can the volume control be Also, the wire to the remote control
"Exactly," agreed Smithy. "Well, used with?" must have insulation suitable for
the answer to this problem is to get "It can be employed," replied mains voltages. If the remote volume
in a stock of rubber bands of various Smithy, "with sets in which the control connects to the receiver by
sizes and just slip one or two of slider of the internal volume con- way of a socket fitted to the back of
these over the board to keep the trol couples to the control grid of its cabinet the socket must offer insu-
more insecure components in place the a.f. output valve. Bring your lation suitable for mains voltage. It
whilst you're soldering them into stool over here and have a look at could, in practice, be a socket in-
circuit. Simple, isn't it?" the circuit." tended for mains use, and a suitable
370 THE RADIO CONSTRUCTOR
type would be one of those small "It's original function," said
2-amp 2-way types." Smithy, "is to provide a means of Four Local Supplier
positioning drawer partitions. But
CORRUGATED STRIP the use to which our correspondent LONDON
puts it is rather different. He fits
Smithy turned back to his bench it inside equipment housings to en-
and picked up the next letter. As he able printed circuit boards or Vero- Established 1910
removed the contents from its enve- boards to be supported at the ends
lope, a length of corrugated strip along the slots between the corru- H. L. SMITH
slipped out and fell to the floor. gations."
Dick picked this up and examined Smithy pulled his notepad towards & CO. LTD.
it with interest. Fig. 3(a)). him, and sketched out the applica-
"Hello," he remarked. "What's tion. (Fig. 3(b)). Comprehensive stocks of
this, then?" "Gosh," said Dick, impressed.
components by
all leading makers.
287-9 EDGWARE ROAD
LONDON W.2
Tel: 01-723 5891

THE
n n n n n n n ry /t n n n n n n n n MODERN BOOK CO.
Largest selection of English &
Adhesive surface covered
by backing strip American radio and technical
books in the country.
(a)
Equipment case Printed circuit board 19-21 PRAED STREET,
LONDON, W.2
Tel: PADdington 4185/2926

ST. HELEN'S RADIO

Hi-Fi Equipment
Screen
Tape Recorders
Radio Receivers
nnrmnnnnn innnn Television
SPECIALISTS IN RADIO
3 strips of Divide-it & ELECTRONIC TEXTBOOKS
(b) ST. HELEN'S GARDENS
LONDON W.10
Fig. 3(a). The appearance of 'Divide-It' strip Tel: 01-969 3657
(b). Employing three strips of 'Divide-It' to position a printed
circuit board and a screen in an instrument case HAMPSHIRE

"It's a strip of special material," "That's a neat idea."

replied Smithy, as he studied the "It is, isn't it?" commented BOURNEMOUTH
letter, "which is known as 'Divide- Smithy. "Incidentally, I must say
It', and which you can buy for a that it's rather surprising to find LARGEST HI-FI AND
few shillings at most Do-It-Yourself how many uses there are in elec- RADIO COMPONENT
stores. One side is flat, and has an tronic work for simple domestic STORE IN THE SOUTH
adhesive surface which is revealed gadgets such as this."
after pulling off a backing strip. The At that instant, the Workshop
other side is deeply corrugated and, kettle gave voice to an ear-splitting
as you can see if you dig at it with whistle, and further activity ceased FORRESTERS NATIONAL
your thumb nail, the corrugations whilst Dick set about the all-impor- RADIO SUPPLIES LTD.
are quite strong physically. The tant task of preparing the tea. Even- 70-72 Holdenhurst Road
material of which the strip is made tually, he placed Smithy's disgraceful Bournemouth
offers good insulation along its sur- tin mug, brim-full with the precious
face." beverage, alongside the Serviceman. Telephone 25232
"What's the strip intended for?" "Ah," said that gentleman as, after
JANUARY 1970
 letters on his bench. "Well now, let's
see what we've got here."
Smithy selected a letter and read
it with what was patently a con-
Spin drier tinually increasing interest.
"This one," he chuckled even-
tually, "is a real dilly! Tell me, Dick,
do you remember the old business
of using OC7Ts as phototran-
sistors?"
Clearly, the contents of this last
OC7I letter had caused Smithy to dismiss
from his mind, at least temporarily,
the displeasure which had followed
Dick's ribald observations on his
wartime activities.
Plasticine "Using OC7Ts as phototran-
sistors?" repeated Dick. "Why, cer-
Fig. 4. A suggested method of displacing the opaque filling tainly I do. The early OC7Ts had
in a glass-cased 0C71. The transistor can then function as a glass cases covered with opaque
phototransistor paint to keep the light out. The
thing to do was to scrape the paint
"You defended this country?" away, whereupon you had a first-
judicious blowing over the surface class phototransistor!"
of the mug, he drank deeply of its queried Dick incredulously. "From "You've got it," said Smithy. "In
contents, "that's the stuff to give the what they tell me, all you ever did fact, I've converted quite a few of
troops!" was to rise to the rank of acting those old OC7Ts in that way my-
Smithy noisily smacked his lips unpaid lance-corporal in the Pioneer self."
in appreciation. Despite the many Corps." "After the early period, though,"
years he had known the Serviceman, Smithy scowled at his assistant. cut in Dick, "the design of the
Dick had still retained his sense of "We ex-Service types," he re- OC71 was changed. It still had the
wonder at the vast quantities of tea marked loftily, "don't talk about our opaque paint on the outside of the
which Smithy was capable of ab- wartime careers." glass housing but, in addition, the
sorbing. "Much." filling inside was made opaque, too."
"I've never," he remarked conver- "Let me tell you," growled "That's right," confirmed Smithy.
sationally, "known anyone who Smithy furiously, "that I did my "The result was that, even if you
drinks as much tea as you do." fair turn of service with all the did scrape off the paint, you still
Smithy quaffed even more mightily rest. Brothers-in-arms, that's what couldn't use the OC71 as a photo-
at his mug, then placed it back with we were.' ' transistor. Eventually, the OC71
a clatter on his bench. Dick turned and addressed the changed over to a metal case, but
"There's nothing like it," he res- empty Workshop. there must still be plenty of glass
ponded cheerfully. "In fact, I rely "We shall next," he announced ones with the opaque filling knock-
on a really good intake of tea each derisively, "have a vocal rendition ing around."
day to keep my kidneys fully flushed of 'Comrades' with an accompani- "All this," remarked Dick, "sounds
out!" ment on the spoons!" intriguing. Go on, Smithy."
Dick shuddered. It was obvious that this last re- Smithy returned to the letter.
"There are times," he complained, mark on Dick's part was the cause "What our correspondent has suc-
"when you're really horrible, Smithy. of deep offence to the Serviceman. ceeded in doing," he resumed, "has
Shall we carry on to the next letter?" "I have no wish," he commented been to shift the opaque filling
"Half a jiff," responded Smithy, icily, "to pursue this subject any inside a glass OC71 away from the
taking a further gargantuan draught further. I shall, instead, turn to the internal transistor junction by taking
from his mug. "Let's get this little next letter." advantage of centrifugal force! The
lot filled up again first." "As you wish," said Dick oblig- scheme consists of pushing the
Disdainfully, Dick picked up ingly. transistor leads into a half-inch cube
Smithy's empty mug and carried it Smithy gazed suspiciously at his of Plasticine, and of then fixing the
over to the battered Workshop tea- assistant, but that worthy's face Plasticine to the inside wall of a
pot. He returned with the mug, carried no expression other than washing machine spin drier (Fig. 4).
charged once more with the fluid that of interest in the next hint He found that, after some two to
so vital for the smooth operation of to be discussed. three minutes spinning around in
Smithy's renal functions, and placed "Hmm," grunted Smithy, turning the drier, the opaque filling of the
it carefully on the Serviceman's back and foraging amongst the OC71 was all forced to the bottom
bench. Smithy snatched it up
greedily and again imbibed deeply.
"Ye gods," snorted Dick in dis-
gust. "You're nothing but a tea-
junkie, that's all you are. I bet
you've got a dirty great hypodermic
syringe full of the stuff back home,
kept handy for the times when
you're really in need of a fix."
"Nonsense," replied Smithy indig-
nantly. "And if you must know, it so Cabinet bottom Polythene seal
happens that I acquired my taste
for tea during those perilous war Woodscrew fitted
years when I was defending this in recess
country, so that the future genera-
tion - including you, mate - could
be brought up in a state of peace Fig. 5. Using the polythene seal from a flagon cap as a
and tranquility." non-scratch cabinet foot
THE RADIO CONSTRUCTOR
372
 of the transistor, leaving the junc- Smithy detached a sheet bearing
tion free and clear. The OC71 then the diagrams from the remainder of Four Local Supplier
functioned as a perfectly good the letter and laid it down on his
photo transistor." bench (Figs. 6(a) and (b)). KENT
"The idea behind this hint," he
CABINET FEET went on, "is to provide temporary EMSAC BASIC ANTENNA SYSTEMS
screening for a wire or cable if it CNl Nuvlstor 2 metre converter.
"That idea," grinned Dick, "is is suspected of picking up or radiat- IF 28MHz. Absolutely complete with
very definitely one for the book!" ing hum or interference, or of caus- separate mains power supply, £11.10.0.
"It certainly represents a jolly Converter only, £8.10.0. P&P 4/6.
ing instability whilst in the un- GDI A G5RV type multi-band dipole.
crafty approach," replied Smithy, screened condition. What you do is 102 ft. top. 100 ft. feeder. Complete
as he took up a further letter. to fit a sheath of aluminium foil and ready to go! 68/-.
There was silence as the Service- or aluminium paper around the GV1 General purpose vertical, useful
man read its contents. range 1.8MHz to 28MHz. Adjustable 7
wire, and it is possible to apply ft. to 23 ft. 92/-.
"Here's another simple one," he this temporary screening over quite TU2 Receiver antenna tuning unit.
announced, after a moment. "Our long lengths. Aluminium coated Amateur and SW broadcast bands 1.5
correspondent advises that the poly- paper is to be preferred because it to 30 MHz. 90/-.
thene seals inside the caps of Send SAE for details. Open weekends.
is tougher, but aluminium kitchen G31AR Carriage extra.
flagons should not be thrown away, foil is nearly as good and is more ELECTRONIC & MECHANICAL
as they make perfect improvised readily obtainable since most grocers SUB-ASSEMBLY CO. LTD.
non-scratch feet for small cabinets. sell it. It's available in widths of Highfieid House, West Kingsdown, Kent.
Tel: West Kingsdown 2344.
SURREY
Portable Transistor Amplifier
plus Dynamic
Microphone
A self-contained
fully portable 'i shahaoaA
mini p.a. system.
Many uses —
nTTT ideal for Parties,
Fold Press Wrap or as a Baby
Squeeze Earth wire Alarm, Intercom —-j/-
foil edges doubled foil tight whipped round Telephone or Record Player, Amplifier,
together foil by hand end of foil etc. Attractive rexine covered cabinet,
size 12 x 9 x 4in., with powerful
round 7 x 4in. speaker and four transistor one
watt power amplifier plus dynamic
microphone. Uses PP9 battery. Brand
new in Makers' carton with full makers'
Ca) guarantee. World famous make QH/ Post
Full List //- Only "lJ/"Free
RADIO
237 COMPONENT
Whitehorse SPECIALISTS
Road, West Croydon,
Surrey. Phone 01-684-1665
SUSSEX
20 approx.
E. JEFFRIES
For your new television set,
tape recorder, transistor radio,
and hi-fi equipment.
PHILIPS, ULTRA, INVICTA,
DANSETTE, MASTERADIO. PERPIO,
12 approx. MARCONI, PHILCO, FIDELITY.
6A Albert Parade,
0 Victoria Drive,
EASTBOURNE SUSSEX
(b) EIRE
Fig. 6(a). Successive steps in wrapping a 2 inch strip of
aluminium foil or aluminium coated paper around a wire to PEATS for PARTS
provide temporary screening ELECTRONIC COMPONENTS
(b). How the 2 inch strip is obtained
RADIO & TELEVISION
There's a sketch in the letter too, about 10 or 20 inches, the 20 inch For the convenience of IRISH
and it shows one of these seals width being better because it works enthusiasts we supply:
being used in this manner." out cheaper." The Radio Constructor,
Smithy showed Dick the sketch "How do you apply the foil to Data Books and
(Fig. 5) then preceded to the follow- the cable?"
ing letter. Panel-Signs Transfers
"You first of all cut it," replied Also a postal service
"This is quite a long hint," he Smithy, "so that you have a long
remarked. "Also, there are several strip two inches wide. You then Wm. B. PEAT & Co. Ltd.
diagrams
a
which have to be looked lay the two inch strip along the 28 PARNELL STREET
I you're going to understand cable to be screened and cut it to DUBLIN 1
the general scheme of things." the length required, allowing your-
JANUARY 1970
373
 formers of this nature. You first of
ail screw the former onto the stud-
ding so that the lower starting point
of the winding is level with the slot
in the post. You then anchor the
end of the coil wire by wrapping it
Top view several times round the left hand
P nail, after which you lay it in the
( slot and pass it round the former.
P
You next wind the coil on the
former, making a close-wound or
( > spaced winding as desired. When
P all the turns have been put on, the
wire is taken over to the right hand
nail and anchored there by wrapping
/ / | it round several times. You now
/ Slot in 1 have the coil wound on the former
2 nail Post post OBA studding 2'nail Wooden base in its correct position, and with the
wire under slight tension. The final
process is to dope the coil. When
the dope is dry you have a nice
neat coil which is all ready to be
connected up."
"That's a jolly useful gadget,"
said Dick thoughtfully. "It would
be just the job for making up
10.7MHz i.f. transformers and things
Side view like that, where you use those black
Aladdin-type formers with the
square bases."
Smithy cast a mildly surprised
glance at his assistant.
"That's a very percipient remark,"
Fig. 7. The construction of a simple coil winding jig he commented, "because I0.7MHz
transformers are one of the main
classes of winding which our corres-
self a little extra just to be on the from the end. Curiously enough, pondent uses the jig for."
safe side. The strip is then folded in blunt scissors are to be preferred for "What's the other hint in the
two along its length and the cable this operation because very sharp letter?"
is either placed in the fold, or the scissors cause the cut edges to stick "It's also to do with coil wind-
fold placed over the cable, accord- together, whereupon the two inch ing," replied Smithy. "And again
ing to which is more convenient. strip is liable to tear when it's un- there are some diagrams to show
You then wrap the foil around the folded." you. (Figs. 8(a) and (b)). This
cable, holding it in place at one end "That's certainly an original second idea is designed to assist in
by wrapping some thin bare tinned idea," commented Dick. "How much the winding of close-wound coils
copper wire around it for several wire can you cover with one roll of on plain formers. Before winding
turns. The other end is similarly the paper or foil?" the coil, a doubled loop of fine
held in place with the aid of thin "A roll of 20 inch foil," said string is placed on the former. The
bare copper wire. After a little Smithy, consulting the letter again, wire at the start of the coil is passed
practice it is possible to make quite "should be sufficient to screen rather through the loop at one end, after
a neat job along the whole length more than 100 feet of wire." which the coil is wound on top of
of the wire and, despite the tem- "Blimey," said Dick. "A roll of the fine string. The wire at the finish
porary nature of the job, the screen- the foil should last for ages, then. of the coil is then passed through
ing which results can be looked I'll have to have a dig round our the other loop of the string. The
upon as being of a semi-permanent kitchen when nobody's looking and two free ends of the string are next
nature. Of course, the main inten- see if I can find any!" carefully pulled tight, whereupon the
tion behind the scheme is the pro- But Smithy was already deeply wires at both the start and finish of
vision of temporary screening in engrossed in the next letter. the coil are held securely under the
order that causes of pick-up or "There are two hints here," he two loops of string. The coil is then
radiation may be isolated. After announced, "and they both have to doped and the excess lengths of
this, the temporarily screened wire do with coil winding. The first des- string cut off."
may then be replaced by proper cribes a coil winding jig for single-
screened cable as necessary." layer coils. Here's the diagram for HEALING THE RIFT
"How," asked Dick, "do you cut it."
out the two inch wide strip?" Smithy showed Dick the sketch With a gesture of finality, Smithy
"You start," replied Smithy, "by in the letter. (Fig. 7). returned the last letter to the surface
unrolling the paper or foil and then "As you can see," he continued, of his bench.
folding it up in the same way as "all the bits and pieces are mounted "And that," he remarked, "is the
they fold lengths of cloth in the on a block of wood. There are two end of this present session with
draper's. The first fold is made two-inch nails, a piece of O BA readers' hints."
after one foot, and then you simply studding, and an upright post which "We certainly had a good selec-
continue to wrap the succeeding can be made of any convenient tion this time," remarked Dick.
foil or paper around the initial one material and which has a horizontal "They were, as always, excellent,"
foot length. The end result is a flat slot in its surface. Many coil for- agreed Smithy. "I always find that
pack measuring roughly 20 inches mers of the type intended to take going through these hints represents
by slightly in excess of 12 inches. threaded iron dust cores can be a very pleasant task."
You obtain a two inch strip by cut- screwed onto O BA studding, and He picked up his mug and drained
ting across the pack two inches in the jig is specifically intended for it of its contents.
374 THE RADIO CONSTRUCTOR
 Coil former Fine string
Stella Nine Range Cases
Manuafactured in Black, Grey. Lagoon or Blue Stelvetite
and finished in Plastic-coated Steel. Morocco Finish with
Aluminium end plates. Rubber feet are attached and there
is a removable back plate. There is also a removable front
panel in 18 s.w.g. Alloy.
Now all Aluminium surfaces are coated with a stripable
plastic for protection during manufacture and transit. All
edges are polished.
LIST OF PRICES AND SIZES
which are made to fit Standard Alloy Chassis
Width Depth 4" Height 6" Height 7J" Height
£ s. d. £ s. d. £ s. d.
BJ" 32" 12 6 15 0 18 0
6i" 4c" 13 6 18. 0 10 0
(a) 8i" 3f 15 0 1 0 0 1 1 0
8i" el 110 16 6 1 11 3
lOJ" 7g" 1 8 6 1 15 6 1 18 9
12i" - 3f 110 16 6 1110
12i" 52" 180 1 14 0 1 17 0
Pull string ends 121" 82,, 1 16 0 2 3 0 2 7 3
to finally secure 14i" 32" 150 1 11 6 1 14 0
141" 9* 2 3 0 2 15 9 2 18 6
coil wire 161" 62" 1 18 6 2 6 3 2 11 6
161" 10?" 2 10 0 3 5 0 3 11 9
Cases—Post 4s. 6d. per Order.
Discounts available on quantities.
CHASSIS in Aluminium, Standard Sizes, with
Gusset Plates
Sizes to fit Cases All 2i" Walls
s. d. s. d. s. d.
6"x3" 5 6 10"x 7" 8 6 14"x 3" 7 3
6"x 4" 5 9 12"x 3" 6 9 14"x 9" 14 6
8"x3" 6 6 12"x 5" 7 6 16"x 6" 10 9
8"x6" 7 9 12"x 8" 10 9 16"x 10" 16 0
Chassis—Post 3s. Od. per Order.
Discounts available on quantities.
Coil wire ends passed
through string loops E. R. IMICHOLLS
MANUFACTURER OF ELECTRONIC INSTRUMENT
(b) CASES
(DEPT. R.C.) 46 L0WFIELD ROAD
ST0CKP0RT-CHESHIRE
Fig. 8. How to secure a close-wound coil on a plain former. Tsl:0ei-480 2179
In (a) a double loop of fine string Is laid on the surface of the
former. The coil Is then wound over the string (b), after which
the string ends are pulled tight to secure the coil wires
NEW VHP KIT
Receives Television Sound, Ambu-
"Would you," asked Dick help- we younger generation can't really lances, Aircraft, Radio 2, 3 and 4
fully, "like some more tea?" appreciate what did go on during on VHF etc.
This novel little set will give you end-
Smithy suddenly recalled his pre- the war." less hours of pleasure and can be built
vious altercation with his assistant. "Probably not," replied Smithy. in one evening. The Kit comes with
"Not," he remarked irritably, "if "In any case there's one point on easy to follow instructions and circuit.
Powered by 9v Battery. Complete with
you're going to make silly comments which I must definitely correct you. built-in Jack Plug Socket for use with
about my habits and, in particular, My eventual rank during the war Earphones or Amplifier.
about what I did during the war." was acting unpaid corporal." ONLY 57/-. P & P FREE UK ONLY
"I'm sorry, Smithy," said Dick "Was it really?" said Dick, assum- Postal Orders, Cheques to
contritely. "Actually, I was indulging ing the expression of the suitably Dept. R.C.3.,
in a little bit of leg-pulling just now. impressed. "Gosh, that was pretty Galleon Trading Co.,
After all, it's only fair that I should high up." 298A Lodge Lane,
do so. You take the mickey out of "I had to fight for that rank," Romford, Essex
me quite often, you know." said Smithy darkly.
"Do I?" responded Smithy, molli- "I bet you did," said Dick.
fied. "Oh, very well then! I must "If this," continued Dick happily,
say that a spot of tea wouldn't be "is a sample of what the New Year
at all a bad idea. All this talking is going to be like, I reckon that
has made me thirsty again." I'm really going to enjoy living in FANTASTIC SALE
Whilst Dick busied himself with 1970!" 1,000 of the following Amplifier Kits
the preparation of a further mug to clear - List Sale
of tea, Smithy carefully returned the EDITOR'S NOTE Price Price P.S.
sheaf of letters to the drawer under 5 watt amplifier £3.15.0 £1.15.0 9y
his bench. Dick came back and The hints published in this 7 watt amplifier £6.15.0 £2.5.0 9v
placed the third mug of tea for that episode of 'In Your Workshop' were 10 watt amplifier £8.19.0 £2.15.0 9v
morning on Smithy's bench. submitted, in the order in which If two are required deduct the follow-
"An excellent brew," remarked they appear, by S. Webb, G. M. ing: 2x5 watt 5/-, 2x7 watt 7/6,
the Serviceman, sipping apprecia- Watson, A. Luty, J. P. Carlile, D. 2x10 watt 10/-. Please add 2/6 P&P.
tively. "I must commend you, Dick, Playle, D. A. Bishop, T. E. Millsom M. CLEAVES AND BROTHER
on your ability to make really good and J. Rankin. Further hints for
tea." this feature are welcomed, and pay- 28 WENNY ESTATE
Dick beamed. ment is made for all that are pub- CHATTERIS
"Going back," he commented, "to lished. CAMBS
our earlier conversation, I suppose
JANUARY 1970 375
 HI March. Mysterious development Ph.D., together with readiness to
at Defence Ministry surplus sale - sign the Officials Secrets Act.
the entire stock of 10,000,000 tran- July. N.A.S.A. announces pre-
sistors type CVU1057 is purchased parations for forthcoming American
for £5,000 by bearded individual in Moon flight to carry out geological
dark glasses and hippy outfit. Minis- investigations. It is proposed to sur-
try of Technology announces that, vey for strata faults by passing elec-
due to the uncontrollable multiplicity tric currents between remote points
m of transistors currently available, it of the Moon's surface. Transistors
is making available a computer ser- (Soho) Ltd. set up small working
vice to other Ministries to enable plant in a cellar in Greek St., Lon-
them to locate specific types of tran- don, WI. The production equipment
sistor after feeding in details of the consists of one high-speed transistor
characteristics required. The com- re-coding machine operated by two
puter can also advise on equivalents middle-aged ladies during the periods
and near-equivalents. Unmanned when they are not required upstairs.
Soviet Moon-shot capsule returns to 1,116 new transistor types released.
Earth with samples of the Moon's August. Retailers shocked by plans
surface. 724 new transistor types of Post Office Corporation to trans-
released. fer to them the administration of
April. Working model of TV re- TV and radio licence issues.
By Recorder ceiver incorporating new integrated Rumours leaking from the U.S.S.R.
circuit shown to Trade Press. The persistently refer to a potential new
only components external to the Russian unmanned Moon-shot to in-
integrated circuit are 60 fixed re- vestigate magnetic phenomena. A
JANUARY 1970? sistors, 40 capacitors, 22 preset poten- senior Soviet physicist is quoted as
Yes, it's true-the calendar can- tiometers, 7 i.f. transformers, the saying that the low temperature on
not lie - and another New Year tube, the scanning yoke, the line and the dark side of the Moon enables
is upon us again. This is the field output transformers and a u.h.f. superconducting electro-magnets of
period when journalists are expected tuner. Cooling of the integrated cir- exceptional power to be set up
to take a surreptitious shufti into the cuit is achieved by means of a motor- without the necessity for complex
future and, conscientious as ever. driven blower rated at j h.p. Mys- cooling apparatus. Meanwhile, the
Old Moore Recorder is now in pro- terious bearded individual in dark high ambient temperature on Earth
cess of setting the regeneration on glasses and hippy outfit registers new resulting from an unusually hot
his crystal ball so that it is just company under the name of Tran- Summer necessitates modifications to
below the oscillation point. sistors (Soho) Ltd. Retailers shocked the TV receiver with the integrated
Let us, therefore, examine what is by 100% Budget increase in S.E.T. circuit. Engineers increase the blower
to happen to us in the electronic for service engineers. motor power to i h.p.
world during this year. May. Under pressure from radio September. Defence Ministry pur-
retailers. Board of Trade announces chases, for £2,500,000, a quantity
PREDICTIONS FOR 1970 that certification of trade-in prices of 10,000,000 transistors type
may also be carried out by retailers' 2N152765A from Transistors (Soho)
January. There is an encouraging engineering staff - details of qualifi- Ltd. Post Office Corporation, as first
start to the year as a spokesman for cations required to be announced step towards metrication, changes
the Ministry of Technology states later. Ministry of Technology states maximum weight for First Class 5d.
that 1970 will be marked by out- that, due to a disappointing lack of postal packages from 4oz. to 2 gr.
standing new developments in British response on the part of other Minis- N.A.S.A. announces that next
integrated circuit production. At the tries, the new computer service for American Moon-shot will take place
same time, the manufacture and in- advising on transistor types is to be in early December. 1,251 new tran-
troduction of new types of discrete made available also to bona fide sistor types released. After an ex-
semiconductor components continues companies in the public sector. tensive Trade Press advertising cam-
to increase unabated and, during the Defence Ministry invites tenders for paign stating that each set "consists
month, 476 new transistor types are supply of new semiconductor devices, of little more than a single inte-
released. Defence Ministry sets up notably 10.000,000 of transistor grated circuit", a pilot quantity of
monster surplus sale for excess stocks type 2N152765A. American research 10,000 TV receivers incorporating
of transistors - notably 10,000,000 scientists state that recently re- the new integrated circuit is offered
of type CV111057. Latest American covered Moon-dust contains high to the retail trade.
Moon-shot vehicle returns with fur- percentage of silicon with n-type October. Total pilot quantity of
ther samples of Moon-dust. characteristics. 910 new transistor 10,000 TV receivers is purchased, at
February. Board of Trade shocks types released. a price only slightly greater than
radio retailers by stating that used June. Front page 6-coIumn story that of the integrated circuits, by
radio and TV trade-in prices must, in Pravda reveals that superior Transistors (Soho) Ltd., after threats
in future, be certified as representing U.S.S.R. unmanned space vehicle on the part of the purchaser to in-
true market value by specially ap- has retrieved material from Moon's voke the Trade Description Act con-
pointed B.O.T. Inspectors. Leading surface containing large quantities cerning the phrase "little more than".
semiconductor manufacturer intro- of silicon having p-type charac- Post Office Corporation engineers fit
duces sensational new integrated cir- teristics. Post Office Corporation en- rectangular plugs to public tele-
cuit, claiming that this replaces vir- gineers fit rectangular plugs to the phone coin box slots to prevent in-
tually all the circuitry of a standard slots of public telephone coin boxes sertion of shilling pieces. Reports
monochrome television receiver. An to prevent insertion of sixpenny filtering out of the U.S.S.R. state
interesting feature of the new inte- pieces. Transistors (Soho) Ltd. con- that next Russian Moon probe firing
grated circuit is its dual row of con- sults Ministry of Technology com- is scheduled for mid-December.
necting terminals, 150 on each side. puter with respect to devices type November. 1,523 new transistor
Post Office Corporation promises CV111057 and 2N152765A. Radio types released. Transistors (Soho)
further improvements in telephone retailers shocked by Board of Trade Ltd. dispose of transistor re-coding
and postal services, these being pronouncement that minimum staff machine and purchase two screw-
directed at attracting increased cus- qualifications for trade-in value drivers, two pairs of pliers and two
tom and revenue. certification are B.Sc., M.Sc., or pairs of side-cutters. Excited French
376 THE RADIO CONSTRUCTOR
scientists studying American Moon- in Greek St., London, W.l. Mass circuiting of the latter to cause
dust samples state that the n-type panic occurs throughout the world adequate damping.
material coalesces into a homo- as Moon suddenly bursts into self- The following modification clears
geneous n-type mass when subjected sustained oscillation, emitting electro- this problem, and the components
to magnetic and electric fields posi- magnetic radiation over the entire referred to are all shown in Fig. 1
tioned at right angles to each other. spectrum from 10kHz to the ultra- on page 594 of the May 1968 issue.
December. Extra police called in violet. Under cover of the resultant First, remove L3. Then, connect the
to control vast crowds of radio communications black-out, a mys- arm of SI (a) to the junction of VC1
home-constructors in Lisle St., Lon- terious bearded individual wearing and Cl. Add a 680pF silver mica
don, W.C.2,. where, using tem- dark glasses and a hippy outfit slips capacitor between the base of TR1
porarily rented premises, Transistors off contentedly to the Bahamas. and the top contact of SI (a) (this
(Soho) Ltd. successfully dispose of is the contact to which the switch is
600,000 fixed resistors, 400,000 THE "SPONTAFLEX" S.A.4 turned in Fig. 1). These alterations
capacitors, 220,000 preset potentio- will result in a much better control
meters, 70,000 i.f. transformers, and Well, let's hope that the New of volume and they still allow the
10,000 each of cathode ray tubes, Year won't be quite as bad as that ! whole of the long wave band to be
scanning yokes, line and field out- And my humble apologies to any tuned. If an upper wavelength of a
put transformers, u.h.f. tuners and company whose name even remote- little over 1,500 metres is enough, a
i h.p. blower motors. Americans set ly resembles that of the fictitious 500pF capacitor may be used instead
up Moon surface electrical probes "Transistors (Soho) Ltd."—all the of the 680pF component. There will
along a line of Moon's latitude. Post above is, of course, just a gag. be some drop in sensitivity owing
Office Corporation announces im- And now let me turn to a matter to the lowering of the inductance/
mediate withdrawal of all coin box which has been raised by Sir Doug- capacitance ratio, but if the Radio 2
telephones due to lack of use by las Hall, the popularity of whose signal is strong enough to make the
public. Isvestia announces that the receiver designs continues unabated. modification worth-while, there will
unmanned Russian space capsule is One of the most ingenious of these still be plenty in hand. This occurs
now set up to produce an intense is the "Spontaflex" S.A.4 Transistor even in South Devon, where Sir
magnetic field along line of Moon's Portable, which was described in Douglas Hall has checked the oper-
longitude. Radio retailers shocked our May, 1968 issue. Although that ation of the altered circuit.
by new Home Office regulations for issue appeared nearly two years ago,
mains lead coding, the new colours readers are still embarking on the BRAND NEW YEAR
being red for earth, black for neutral construction of the S.A.4 receiver.
and green for live. Fortuitously, A minor difficulty experienced by And so I have now come to the
Americans and Russians switch on a few constructors has risen in dis- end of my contribution for this
their apparatus on the Moon's sur- tricts where a very strong signal brand new year of ours. Whereupon,
face simultaneously. Militant Cleans- from Radio 2 on 1,500 metres is there's nothing better I can do, in
ing Department operatives jostle received, the effect being that this finishing, than wish readers all the
G.L.C. officials whilst demanding cannot be sufficiently reduced in vol- very best of luck together with an
bonus pay for clearing 10,000 tele- ume. The trouble arises from the exceptionally Prosperous and Happy
vision chassis, each bare except for fact that the long wave coil is too New Year.
one integrated circuit, from a cellar far from the reaction coil for short-

NEW ITT-COMPONENTS
PUBLICATION LAST LOOK ROUND
ITT Components Group Europe
has just published the first of a set
of components summaries which • ALL CHANGE
heralds a new style for presenting Are you confused by the decimal coinage relationship with the
a concise run-down on all the ITT present currency?
component ranges. If so, we feature next month a Decimal Coinage Calculator—a
The first publication is the simple calculating device enabling direct conversion from current
Passive Components Summary (ref. pennies to New Pence by way of a potentiometer circuit and a meter.
6000/301). This 116-page short-form We certainly ring the changes!
catalogue covers such products as:
Quartz Crystals and Filters, Mag- • BIRDS OF A FEATHER
netic Materials, Transformers, Ever tried tape recording birds?
Capacitors, Resistors (fixed and If not, you will find that one on tape is worth two in the bush!
variable), Infra-red filters, etc. The Try it and see with the advice given in Tape Recording Birdsong—
also to be included in our next issue.
books are available in English,
French and German languages.
• 349
Copies are available free on
application to ITT Components On that page in this issue you will have noted another article by
our popular contributor Sir Douglas Hall, K.G.M.G., M.A. Always
Group Europe, Publicity Depart- a source of great interest, his circuits are nothing if not ingenious!
ment, Edinburgh Way, Harlow, In the next issue we shall be publishing another of his interesting
Essex. Please use company letter- articles entitled Developing the "Spontaflex" Short Wave Receiver-
head and state job title. make sure you get your copy!
JANUARY 1970 377
 «300.5£ ,—
'o> ^al io o >0 O O'O^HVO'—I
voo»oosTf \0'-HVD»-<VO
go m r- vo — oo
^ on
o
o •-H OO fN r-
T— r4 M — ^ xf Tj- »o\ovor-r~- oo
00 ^
• S T3
0) c®»
W -2 §
UJ ^31
"cd "o TO rj- vo
vo r- o o o o o o 00o (N
o vo
o o o o o
D
00 Oh r4
Tt o oo
v-j r^-
o
oo on
on m
^ r- tj-
--i wo
vo r-
o -3"
oo rq O Tj-
wo ON
vo m
00^
c/)o .—c n. X ^ ^ ri rf ri ri ri ri
3-
c- r-l
<
^•3 2
> Oo^
■"cn £So ■*-*
d<u o cj n 00 Tf "^- wo vo vd o oo rf
oo ONOOOO OOO'
o LU vo —-c^vof^o r-
> <-.D %4) d ra CN wo
ro ra
xf- On wo vo r- oo On cdN O —H rv)
(D Ooo >o o-
d on
<
"'•a 21,1
CC ■2-S
d 3 oo
LU o o o o o o o wo
o oo o o o o o ooo
0)wT.^'
: d wo O cS
^ tN o o
wo ro WO
ro O
Tt wo O
Tl" wo WO \D wo
vo O r-
wo oo
o oo
WO o
O-N wo o
ON O
> dM
5 O)a) 4) PJh
<
OJS3 13> 4»4)
O4) 4)S3 -*-'
•g ^is O OO
O r-;
O VO
O WO
^ o'dP ON co ra ^ p o o p -H —J
o i-^ r4 d- wo NO r-^ OO ON cso
— r-of vd
ro wo ro fN 00
^ ON
O
ot/T'd
d ^g.so
^ wo NO t-
13 o
>.S
C3 BBtj
sDH «C «tH
•■> £ ai
< O<» rd0d >2r CO P
NO rt
00 p —;
d^ o ^sI p O OO NO d OO NO d" CO
00 wo OO d- On
UJ
r^- sj ^ rf wo
(N OO
(N
o 0. cHj:
o -*-'^ dd
tc 'd ^
Of) '§.s
CO iji (A t-i
8 Ss ^ p
Tt CO p
Tj- P
NO 00 O
as CO tJ" no t oo
rgo D. rf rt- wo 00 ON r4 Tf
oo O S3
o
v-
S ^0.0
SOD 1—I
^ WlOX)
££G
o Od d£
S3 T-nr-fCOTfWO NO C- 00 ON O o CO
C^f o Tt
o WO
o o
NO o OO
f- o On
o Oo
d^
d<u ^c S300
H fc-S
w od o>
Q
U
DC