WIT111133 'RAH

ELECTRONICS. RADIO, TELEVISION

Managing Editor: HUGH S. POCOCK, m.LE.E.

Editor: H. F. SMITH JUNE 1956
Assistant Editor: F. L. DEVEREUX, B.SC.

In This Issue Editorial Comment 251

Simple Wobbulator .. B. T. Gilling 252

World of Wireless .. 254
Unconventional F.M. Receiver .. M. G. Scroggie 258
Long Range on V.H.F. .. 262

Characteristics of Fixed Resistors G. W. A. Dummer 263

More About Noise .. " Cathode Ray" 266.

Physical Society's Exhibition .. 271

Test and Measuring Gear 275

New Valves and Semi-conductors 279
Letters to the Editor .. .. 281

Cascode A.F. Amplifier .. .. L. B. Hedge 283

Wide -Band Television Aerials M. G. O'Leary 288
Manufacturers' Products .. 292
Aerial Cross -Over Network .. L. S. King 293
Abnormal V.H.F. Propagation A. H. Hooper 295
Books Received .. 298
VOLUME 62 NO. 6
Short-wave Conditions .. 298
PRICE: TWO SHILLINGS
Live and Recorded Music 299
Random Radiations .. " Diallist " 300
FORTY -SIXTH YEAR
OF PUBLICATION Unbiased .. " Free Grid" 302

Street, London, S.E.1.

PUBLISHED MONTHLY (4th Tuesday of preceding month) by ILIFFE & SONS LTD., Dorset House, Stamford
and Overseas, E1 12s. 6d.
Telephone: Waterloo 3333 (60 lines). Telegrams: " Ethaworld, Sadist, London." Annual Subscription: Home
Midland 7191.
U.S.A. and Canada $5.00. BRANCH OFFICES: BIRMINGHAM: King Edward House, New Street, 2. Telephone:
Telephone: Central 1265.
COVENTRY: 8 -10, Corporation Street. Telephone: Coventry 5210. GLASGOW: 28B, Renfield Street, C.2.
MANCHESTER: 260, Deansgate, 8. Telephone: Blackfriars 4412. OVERSEAS OFFICES: U.S.A.: Ill, Walnut 4- 5631.
Broadway, New
York, 6, N.Y. Telephone: Digby 9 -1197. CANADA: 74, College Street, Toronto, 2, Ontario. Telephone:

www.americanradiohistory.com
 82 ADVERTISEMENT WIRELESS WORLD JUNE, 1956

\\\\\\\\\\\\\\\O\\\\O\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\'\\\\\\\\\\\\\\\

TRANSISTORS IN AUDIO AMPLIFIERS (Part Two)

Although in principle a large number of circuits can be from (a )2RL /r';n, the a.c. current gain being a' and the
obtained by combining grounded emitter, grounded base or voltage gain a RL /r'ln. This expression assumes that RL
grounded collector configurations with transformer or R -C is very much smaller than r'22 and emit.
coupling, in practice transistor audio amplifiers tend to
follow a simple pattern. A typical circuit can be considered
Here, ay'in, etc. are Small- Signal parameters given in
published data and computed for the working point
to have grounded emitter stages in cascade, with R -C employed. As the load on an R -C coupled stage is formed
coupling, and with d.c. stabilisation provided by the potential by its collector resistance in parallel with the input resistance
divider and emitter resistor method. of the following stage, the power and voltage gain for each
The maximum power gain available with perfect matching stage can be calculated by working backwards through
(and transformer coupling) when the effective load resistance the cascade.
in the collector circuit RL = r'22. r'out and the effective Class AB push -pull operation in which the bias corres-
ponds very nearly to that for true Class B operation is a
source resistance R5= /CH .r';n natural choice for the output stage when a transistor
is amplifier is to be designed as a power amplifier, that is, to
z give the highest output power permitted by the collector
r'2z dissipation pcmax, without objectionable distortion. The
(v;'vT; .
quiescent power consumption is very small and the efficiency
is high. The Mullard 0072 is intended for this mode of
R -C coupling is preferred generally to transformer coupling operation. An actual circuit is shown in the diagram, the
for low cost and phase shift and good response, but the output power being 200mW for 10% total harmonic dis-
power gain of each stage then arises solely from the in- tortion for an input of about 6mV at CI or 500mV at
herently high current gain of the grounded emitter stage, and R1. Negative feedback is applied over the driver and out-
the higher gain which would be available by impedance put stages by R13, which is matched to the loudspeaker. A
matching with the transformer is not achieved. small amount of bias is provided to the 0072's by the
The factors entering into the design of an R -C coupled potential divider R11 -R12, which is effective in reducing the
transistor cascade are not difficult to appreciate; many of high crossover distortion inherent in a true Class B tran-
them are similar to those encountered when working with sistor output stage.
valves. The collector voltage and
current are limited by d.c. ratings Speaker T2
Vcmax and Icmax, and by a.c. Load RO Turns
ratings vc(pk)max and ictpklmax. (n) (kn) Ratio
For high gain and output power 3.0
7.5
39
56
7.3,1
4.6 :1
the battery voltage should be 10.0 68 4.0:1
high, but a lower voltage and 15.0 82 3.3 :1

hence smaller current drain is

more economical. The high value
of collector load resistance R1

required for maximum gain can-

not be obtained with R -C coup-
ling, aS there is no advantage in
making the collector load very
much greater than the effective
parallel input impedance of the
next stage. In addition, the load
resistance and collector current
determine the voltage available t Sec text
across the transistor, which is 2795A All resistors should be t 5°/e Io values at 25 °C ambient
also reduced by the emitter resistance included for stabilising. The value of Rl l must be chosen from the range 6.8, 6.2,
The collector current should therefore be small so that a large 5.6, 5.1, 4.7, and 4.3k S2 so as to adjust the total quiescent
collector load resistance can be used; on the other hand a large current in the output stage to 1.3mA+10% at 20 °C or
collector current swamps the variation in collector leakage l.6mA +10% at 25 °C. The operating ranges with speech
current I'c(ol with temperature. and music are 15 °C to 45 °C ambient temperature and 4.5V
After allowing for these various conflicting claims, the to 2.7V (or even 2.0V, depending on the distortion tolerated
number of stages is chosen to give the required overall gain by the listener) with a Leclanché type battery.
when feedback is applied. Since the signal swing in the early Suitable transformers can be obtained from R. F. Gilson
stages is small, the d.c. working point can be chosen for low Ltd. The phase splitter transformer is type W0780 and the
current drain (and noise), provided they have potential output transformer W0781. The secondary resistance must
divider and emitter resistor d.c. stabilisation. The power gain be specified as 3.7552, 7.5t , or 1552 when ordering the
in the grounded emitter R -C coupled stage can be calculated output transformer.

Reprints of this series of advertisements WILL NOT he available

T.S.D. DATA and PUBLICATIONS SECTION, MULLARD LTD., CENTURY HOUSE, SHAFTESBURY AVE., LONDON,W,C.2
iem356

www.americanradiohistory.com
 JUNE
1 9 5 6

Wireless World vol. 62 No. 6

Combined Audio Show?

WITHIN the space of six weeks, from April 13 to are too many of them, both on sound and (in the
May 26, two shows covering the same subject and form of film) on television. Anyway, we cannot
run on broadly similar lines were held in London. resist the temptation to quote from a leader in The
These were the Audio Fair and the annual exhibi- Times (May 21). The newspaper's comments
tion of the British Sound Recording Association. were linked to a recent lecture on the RCA system
That meant a duplication of effort on the part of of television recording on magnetic tape, but
most of the exhibitors (to say nothing of many could be applied with equal or even greater force
visitors) which can hardly be allowed to carry over to the use of films.
into future years. That was acknowledged by " One of the possibilities is that technical pres-
Norman Leevers, the retiring president of sure may cause television to lose some of its
B.S.R.A., who, at the Association's annual dinner, spontaneity. There is nothing, except a deter-
speculated on possible ways of arriving at a more mined will to do so, to stop the making of a tele-
satisfactory arrangement. vision programme becoming something like that
Mr. Leevers put forward two alternative pro- of a film. This could alter the whole artistic nature
posals : that the B.S.R.A. show should be moved and scope of television. Again, there may come a
to the autumn, or that it should be combined with strong tendency for the tail to wag the dog-the
the Audio Fair. content of the programme being influenced by the
To us, the idea of a combined exhibition seems fact that it is a recording. Some of these fears are
to be the better. Demonstrations under reason- theoretical. That does not mean to say they will
ably good conditions are an essential part of a not materialize."
sound reproduction show, and, practically speak-
ing, the only suitable venue is a large hotel. In
London, hotel accommodation is easier to obtain
Research and Measurement
THE number of new instruments and techniques
in the spring than in the autumn. And, anyway,
it is doubtful whether the majority of manufac- shown at the Physical Society's exhibition, re-
turers would continue to support two shows in the ported elsewhere in this issue, is still on the
year. increase, and several new trends were noticeable.
An effective basis for collaboration between the In particular, there was fairly widespread use of
organizers of the Audio Fair and the B.S.R.A.
should not be too difficult to work out. Facilities
transistors in instruments, both as h.t. generators
and, in some cases -e.g., in physiological work
where their small size and self-contained nature is
-
might be provided for the B.S.R.A. to stage non-
commercial exhibits and demonstrations of various an especial advantage. There was also a new type
aspects of sound reproduction. Then the Asso- of transistor with quicker response than that of
normal junction types. This device is perhaps
ciation's annual convention might be arranged to
coincide with the Fair. Such things as these symptomatic of the general development of junc-
would, we feel sure, prove to be attractive to many tion transistors capable of working at higher fre-
visitors. Collaboration between the two bodies quencies than at present. The same tendency was
on these lines should be beneficial to both of them. evident in oscilloscope tubes with ability to record
frequencies of the order of 1,000 Mc /s; these are
specially adapted to the recording of very fast
Recorded Programmes transients. On the same trend, super- high -fre-
THOUGH the programme side of broadcasting is quency valves capable of being tuned over quite a
no real business of Wireless World, we sometimes wide range were shown.
feel impelled to comment on it when fundamental A particularly commendable and much appre-
issues likely to affect the growth of the service are ciated section of the exhibition was the demon-
concerned. One such issue is that of transmissions stration of the principles of colour television,
from recordings; sometimes it seems to us there arranged by the Physical Society Colour Group.
251
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 Simple Wobbulator
Electro- Mechanical Modulation (6 Mc s) in a

Single -valve Circuit By B. T. GILLING

Modulator unit with Any miniature medium-mu triode is suitable for

ceramic top plate the valve, in fact one of the acorn triodes in the alti-
removed, showing meter is an admirable choice. The frequency-
semi -circular silver- modulated output is taken from the anode of this
ed fixed electrodes valve and its amplitude is controlled by a simple
and domed and per- attenuator.
forated diaphragm.
Alternating current for driving the moving coil is
obtained from the heater supply. A series variable
resistor controls its amplitude, hence the frequency
deviation, and a fixed resistor prevents over -drive.
The sweep voltage for the X-plates of the oscillo-
scope is also taken from the heater supply.
THE wobbulator to be described reaches just about
the limit of simplicity, consisting as it does of a
single-valve oscillator; yet it is most satisfactory in Return Trace Suppression
use, and gives on Band I a linear sweep of six mega- In an ideal case the forward and return traces
cycles. would coincide exactly, but owing to phase shifts in
In the American radio altimeters now available the amplifiers this does not happen, and a trace as
on the surplus market there is a component which in Fig. 2(a) is always obtained. There
makes the above possible. The altimeters have the ways of removing the second trace and the are several
type numbers RT-40 /APN, AN /APN -1, adopted in the present case is method
AN /ARN -1, AYB -1, AYD and there may be others. voltage from a phase shift networkto across
take a blanking
The component is often described in advertisements supply and apply it to the grid of the tube. the heater
This not
as a " magnetic sounder " but it would be more only blacks out the unwanted half of the trace but
appropriately called an electro-magnetic frequency brightens the centre portion of the wanted half
modulator. where its brilliance decreases owing to the increased
A typical example is shown, partially dismantled, speed of the spot. A single trace of
in the accompanying photograph. It consists of a throughout its length is obtained, as even brightness
in Fig. 2(b).
magnet and coil, similar to that of a moving-coil At full amplitude of the moving coil a sweep of
loudspeaker, to which is attached a slightly domed 10 Mc /s in Band I is easily obtained. This is illus-
aluminium diaphragm of about two inches dia- trated in Fig. 2(c). With this amplitude, however,
meter, freely suspended and perforated all over to frequency linearity suffers, but for a sweep of 6 Mc /s
prevent air loading. Mounted in front of this dia- it is very good, as will be seen from Fig. 3.
phragm on a ceramic cover are two metal plates
and these with the diaphragm form a two -gang +H.T.
capacitor. The capacity swing of each section is Fig. I. Circuit diagram
from 10 to 50 pF. It will be obvious that if this of single-valve wobbu-
capacitor is connected across the coil of an oscillator 50 VOLTS
lator.
and the moving coil energized by an alternating
current, the oscillator will be frequency modulated.
The circuit of the complete wobbulator is given in 50 pF
Fig. 1. Ample frequency
deviation on Band I could be
obtained by using only one 20011
O FROM SIGNAL
GENERATOR

section of the f.m. capacitor, MODULATOR 1 5 p

but since it was also desired UNIT Ikft
to sweep the i.f. band around O F.M. OUTPUT
10 Mc /s the two sections were IoOpF
connected in parallel to give a
maximum swing of 20 -100 pF.
This means that one side of
the capacitor is earthed, mak- O EARTH
ing it necessary to use an oft
electron -coupled type of oscil-
lator. The 10-pF variable 10011 WIRE -WOUND

capacitor in parallel with the HEATER

coil is brought out to a con-

trol on the panel to enable the O GRID
middle frequency to be set. C.R.T.
OX AMPLIFIER}
252 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 due to
Fig. 2. Three forms of the some response from a
partly aligned television receiver. (a) Displaced return half,
6 -Mc /s sweep. (c) Same response with 10 -Mc /s sweep.
phase shift. 6 -Mc /s sweep. (b) Return half suppressed.

The trace obtained is of little use unless some 59

means are provided of ascertaining the frequency

of all points along it, so a marked " blip " is applied 58
by adding the output from a signal generator to the
output of the wobbulator through a small capacitor.
With a normal wide -range Y- amplifier this will 57

appear as a thickening of the trace on each side of

the marked frequency. To get a sharp blip the fre- 56
quency range of the amplifier must be severely
limited and the simplest way of doing this is to con-
nect a 0.001 -11F capacitor across its input. The 55

resultant blip is seen in the three oscillograms in

Fig. 2. As the frequency of the signal generator is 54
varied the blip will travel along the trace marking
the spot frequency at any given point.
This particular wobbulator was designed for 53
Channel 3, Band I, and the i.f. band from 9 to
14 Mc /s with switched coils. Both coils are wound
on half -inch formers and are as follows. Channel
520 2 3 4
3 :9 turns, tapped at 3 turns, spaced to occupy á SPOT DISPLACEMENT (INCHES)
inch. I.F. :18 turns, tapped at 6 turns, 34 s.w.g. Fig. 3. Linearity of trace on Channel 3.
close -wound. Adjustments to make the coils reson-
ate at about the middle frequency can be made with
either iron dust or brass cores. re-
frequency modulation is far simpler and moresatis-
liable than the reactor valve and has proved so
Alternative Modulator factory that it is used extensively, especially in
America, for commercial wide -range beat frequency
The heart of this wobbulator is the unit from theit wobbulators.
altimeter, but if this component is not obtainable
is possible to make a very satisfactory one from a
three -inch moving coil loudspeaker. There are two B.S.R.A. Convention and Exhibition
ways of doing this, both of which entail the cement-
ing of a thin aluminium disc to the diaphragm. In ONCE again the annual exhibition* and convention of
held
the first case a metal plate is fixed immediately in the British Sound Recording Association was made
front of the disc as closely spaced as practicable to in the Waldorf Hotel and the increased space
form the variable capacitor. In the other the actual available was fully booked. Of the 42 firms exhibiting,
oscillator coil is wound in pancake form and mounted 27 gave continuous demonstrations in separate rooms in
the hotel, instead of sharing the use at specified times
closely to the disc. As the disc moves the changing of a communal listening room. Conditions were more
eddy- currents alter the inductance of the coil. This comfortable and there was much less fluctuation in pre-
the
latter method is not as satisfactory, as it prevents density of visitors in the main exhibition than on
other coils being switched in to provide alternative vious occasions.
ranges. Much interest was shown in the competition by mem-
The only disadvantage of the generator in the bers of the Association for the best amateur- constructed
form shown is that it requires a coil for each channel equipment. The winner of the President's Trophy was
swept. At present, when there are only two, or at V. L'Estrange for a versatile lightweight magnetic re-
most three, channels operating in any area, this is corder with three tape speeds. The Wireless World for a
prize was awarded by the judges to G. A. Jeary
of no real significance. To cover future develop- synchronizing system for a film projector and magnetic
ment and the possibility of the introduction of tape recorder and the Committee Prize went to J. M.
colour television, which will call for the examination Beukers for a beautifully finished pre -tuned f.m. re-
of a band almost in the audio region, this unit can ceiver and gramophone pre -amplifier.
be looked on as the basis of a more elaborate wobbu- At the annual dinner Norman Leevers, the retiring
lator of the beat frequency pattern. In this it would president introduced his successor, J. F. Doust (M.S.S.
form the frequency -modulated oscillator which beats Recording Company) who has had a long association
with a tunable oscillator to give a resultant swept with technical developments in disc and tape recording.
output over a very wide range of middle frequencies. * Sound reproducing equipment shown at this
and other recent
This electro-mechanical method of obtaining exhibitions will be reviewed in the July issue. -EDITOR.
253
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 WORLD OF WIRELESS
Organizational, Personal and
Industrial Notes and News
National Radio Show
OF the eighty or so British radio equipment manu-
facturers included in the initial list of exhibitors at
the National Radio Show (Earls Court, August 22nd
to September 1st), about 50 per cent are makers of
domestic sound and television receivers-the re-
mainder being component and accessory manufac-
turers. With the addition of the "user" exhibitors,
publishers, banks and wholesalers, the total number
of stand -holders is 111.
It is anticipated that some 400 television receivers
will be in operation at the Show where both Band
I and Band III programmes will be distributed.
As in past years, the main part of the ground floor
will be devoted to manufacturers' stands, On the
first floor will be the displays of the B.B.C. and, for
the first time, the I.T.A. and the London programme
contractors. There will again be a " careers and
electronics " display on this floor.
A brochure for prospective overseas visitors has Because the final stages of the passages of the British Railways
been circulated and copies may be obtained at the vessels on the Heysham -Belfast service have to be navigated
various offices of the British Information Service. stern first, the radar installation for the three latest vessels has
been specially adapted by Kelvin Hughes. As will be seen in
Southern England TV this photograph of the experimental installation at the K. & H.
research station on Southend pier, the masts will be fitted with
THE permanent aerial on the new 500ft mast at the deflector plates to eliminate spurious echoes. The vessels,
B.B.C. television station at Rowridge, Isle of Wight, which incidentally are fitted with bow rudders, will have
is being brought into service on June 11th. This a second radar display unit on the after navigating bridge.
will increase the e.r.p. to more than three times that
provided by the temporary aerial system on the Birthday Honours
200ft tower which has been in use since the station
opened in November, 1954. As the new aerial is Sir Gordon Radley, C.B.E., who has been director
directional the vision e.r.p. varies from 1 to 32 kW. general of the Post Office since 1955 and was pre-
Provision is made for the mast to be used for v.h.f. viously engineer -in- chief, is appointed a Knight
broadcasting also. Commander of the Order of the Bath.
The temporary transmitter on Truleigh Hill, near John Anderson, C.B.E., chief scientist at the
Brighton, will continue in service for the time being Admiralty Signal and Radar Establishment, Cosham,
although it was originally intended only as a Hants, is appointed a Companion of the Order of the
temporary measure until Rowridge was operating on Bath.
full power. L. H. Bedford, O.B.E., M.A., B.Sc.(Eng.), chief
engineer, Guided Weapons Division of the English
Institution of Electronics Engineers? Electric Company, which he joined in 1947, is pro-
moted to Commander of the Order of the British
ALTHOUGH originally formed in 1946 as some- Empire. He was for many years director of research
thing of an " old comrades' association " the Radar at Cossors and was one of the first two industrial
Association has recently taken on the aspect of a engineers to be taken into the confidence of the
technical institution. As such it has arranged a Government on radar. He evolved the " Bedford "
number of technical meetings during the past year at attachment for early gun- laying radar.
which lectures on such topics as underwater T. Constantine, chairman of Bonochord Limited,
television, guided ,missiles, radio astronomy and is appointed a C.B.E. for public and political
colour television have been given. It is now planned services.
to start a students' section for the benefit of those Appointments as Officers of the Order of the
studying electronics. British Empire are conferred on A. B. Howe, M.Sc.,
Sir Robert Renwick, the president, speaking at the A.R.C.S., assistant head of the research department
10th anniversary dinner of the Association, referred of the B.B.C., which he joined in 1924, and H. G.
to the need for the electronics industry to have its Sturgeon, director and chief engineer of Ultra
own technical institution. " I believe," he said, " that Electric.
the Radar Association is destined to assume this J. Treadgold, B.E.M., lately principal station radio
rôle." He suggested that the scope of the Association officer, Admiralty Civilian Shore Wireless Service,
should be extended and that its name should be and E. E. Frewin, chief technical superintendent
changed to cover the whole field of electronic of the Gold Coast broadcasting department, are
engineering. appointed M.B.E.
254 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
PERSONALITIES H. S. Jewitt has left Decca Radar, where he was
senior engineer in charge of the receiver design group
Professor H. E. M. Barlow, B.Sc.(Eng.), Ph.D., at the research laboratory, Tolworth, Surrey, and
M.I.E.E., who has been a member of the academic emigrated to the U.S.A. to join the Raytheon Manu-
staff of the Faculty of Engineering, University College, facturing Company, of Boston, Mass. Before joining
London, since 1925 and is now Pender Professor of Decca in 1952 he was working on pulse circuitry at
Electrical Engineering at the college, has been elected Ferranti's after graduating as B.Sc.(Eng.) from Queen
a Fellow cf the American I.R.E. The citation reads : Mary College, London University, in 1949. It will be
" For contributions to engineering education, telecom- recalled that Mr. Jewitt won one of the R.I.C. technical
munication, and high- frequency techniques." At the writing premiums for his articles on i.f. amplifiers pub-
beginning of the war Professor Barlow, who is 57, was lished in Wireless World in 1954. He is 33.
at T.R.E. and subsequently became superintendent of R. T. Lakin, A.M.I.E.E., A.M.Brit.I.R.E., chief
the radio department at R.A.E., Farnborough. He has research technologist with the Whiteley Electrical Radio
served on a number of Government boards and coun- Company, and H. W. Read, London manager, have been
cils including the Radio Research Board, to which he appointed to the board of directors. Both of them have
was appointed in 1948. been with the company for more than twenty years. In
For his contributions to the development of air 1953 Mr. Lakin was appointed M.B.E. for his scientific
traffic control systems, J. Fenwick, a senior signals contributions during the war.
officer in the Ministry of Transport and Civil Aviation, K. M. McKee, B.Sc., A.M.I.E.E., formerly a senior
has been awarded the British Silver Medal of the Royal engineer for five years at the E.M.I. Research Labora-
Aeronautical Society. He is the senior telecommunica- tories, has joined the Geo. Tucker Eyelet Company,
tions officer in charge of the Southern Air Traffic Con- where he is in charge of the technical and commercial
trol Centre adjacent to London Airport and has been development of a range of automatic component
responsible for the design of the radar simulator which assembly machines for use in printed circuitry.
is used by the Ministry for the rapid training of control
officers in radar control. Mr. Fenwick has also designed L. S. King, contributor of the article in this issue
and engineered the whole of the radar display systems on the Band I /III crossover network, has been with
fo: the new Southern Air Traffic Control Centre. Standard Telephones and Cables for many years and
for over thirty years has been part -time lecturer at the
C. P. Fogg has been appointed head of the ground West Ham Municipal College and the Northampton
radar department at the Radar Research Establishment Polytechnic, London, E.C.1.
of the Ministry of Supply, at Malvern, where, for the A. H. Hooper, who contributes an article in this
past five years, he has been superintendent of basic issue on the graphical derivation of radio refractive
techniques. He joined the staff at the Bawdsey Research index, is a meteorologist. Among his other interests
Station in 1937 and in 1939 was made leader of a group is experimental amateur radio and recent private investi-
responsible for research and development of radio re- gations have included the feasibility of routine assess-
ceivers. He went to Malvern in 1945 and was at one ments of conditions for v.h.f. propagation through the
time in charge of the three divisions working on trans- troposphere.
mitters, aerial test gear and receivers and display. He
is 42.
OBITUARY
W. M. York, who has been in charge of Ekco pub- Stanley Whitehead, M.A., D.Sc., M.I.E.E., F.Inst.P.,
licity for the past twenty -four years and has been an director of the British Electrical and Allied Industries
executive director of the company since 1951, has been Research Association (E.R.A.) for the past ten years,
appointed commercial director with a seat on the Board. died on May 5th, aged 54. Dr. Whitehead, who had
He recently visited the Ekco organization in India. been a member of E.R.A. since 1925, was for some years
M. M. Macqueen, manager of the radio and tele- chairman of the International Special Committee on
vision department of the G.E.C. since 1930, has Radio Interference (C.I.S.P.R.) and was a member of
accepted the chairmanship of the British Radio Equip- the advisory committees appointed by the P.M.G. to
ment Manufacturers' Association for the third succes- investigate radio interference from ignition systems
sive year at the express wish of the members of the and small motors. Since 1954 he had been joint
council. He was also chairman in 1949 and 1950. Mr. honorary secretary of the Parliamentary and Scientific
Macqueen, who joined the G.E.C. in 1923, has Committee.
represented the industry on a number of committees Maurice G. Hammett, M.I.E.E., chief engineer of
including the Radio Rearmament Advisory Committee the electronics division of Murphy Radio since Septem-
set up by the Government in 1951 for liaison between ber, 1953, died early in May at the age of 48. For
the Ministry of Supply and the radio industry. some time before joining Murphy's he was with the
Ekco organization as chief engineer of the electronics
division at Malmesbury. He started his engineering
career with E.M.I.

WHAT THEY SAY

" Tall oaks .".-" Looking round this room I
am struck by the thought that not much more than
fifty years ago the entire manpower engaged in the
radio industry numbered fewer than those representa-
tives of it present here to -day. Indeed, it would be true
to say that when my father [Guglielmo Marconi] first
came to England he was the radio industry of the
time "- Marchese Giulio Marconi at the Radio In-
dustries Club's 25th anniversary luncheon.
Incognito.-"Hitherto it has been the practice to
classify this new electronics industry as an off -shoot of
the long-established electrical industry, and as such
it possesses no separate identity and no official classi-
fication. Statistically the electronics industry does not
M. M. MACQUEEN R. T. LAKIN exist despite the fact that it employs some 250,000

WIRELESS WORLD, JUNE 1956 255

www.americanradiohistory.com
 people with an annual turnover exceeding £250 million N.W. Germany. -With the reorganization of the
and exports valued at over £60 million." -Sir Robert broadcasting service in what was the British zone of
Renwick (president) at the 10th anniversary dinner of occupation in Germany, the Nordwestdeutscher Rund-
the Radar Association. funk has been disbanded. There are now two sound
Obsolescent L. & M.W.? -"
The B.B.C. is rapidly
carrying through a complete conversion from long and
broadcasting organizations in the zone: the Nord -
deutscher Rundfunk (N.D.R.) with its headquarters in
medium waves to v.h.f." -Sir Ian Jacob, addressing the Hamburg and the Westdeutscher Rundfunk (W.D.R.)
C.C.I.R. delegation. centred on Cologne. There is, however, a co- ordinated
television service for the zone, in which there are twelve
television stations, operated by the Nord - and West -
IN BRIEF deutscher Rundfunkverband (N.W.R.V.).
Receiving Licences. -The overall increase in sound The Radio Industry Council is planning to hold a
and television licences in the United Kingdom during Scottish Radio and Television Exhibition at the Kelvin
the year ended April 30th was 278,533 bringing the Hall, Glasgow, in mid -May next year. Although since
total to 14,295,980. During the year television licences the war there have been two Glasgow radio exhibitions
increased by 1,231,453 to 5,812,178 and sound licences organized by the Scottish Radio Retailers' Association,
decreased by 952,920 to 8,483,802. this will be the first manufacturers' show in the city
The B.B.C. has selected the site for its Cumberland since 1935.
television station which it is planned to bring into Northern Electronics Show.-The 1 1 th annual elec-
operation by the end of next year. It will be built at tronics exhibition, organized by the Northern Division
Sandale which is 1,200 feet above sea level and some of the Institution of Electronics, will be held at the
14 miles south -west of Carlisle. The station will also College of Technology, Manchester, from July 12th
' be used for v.h.f. sound broadcasting when the system
to 18th (excluding Sunday 15th). It will open on the
is extended to that part of the country. first day at 2 p.m. and on subsequent days at 10 a.m.
B.B.C.-I.T.A. Co- siting.-Replying to a general ques- and will close daily at 10 p.m. except on Saturday when
tion in the House of Commons on the sharing of masts it will be 6 p.m. During the course of the exhibition,
by the B.B.C. and I.T.A., the Postmaster General which will include research and manufacturing sections,
announced that this will definitely be done in Cumber- some fifty lectures will be given and sixteen films
land. shown. Admission tickets will be obtainable free from
I.T.A. Test Transmissions. -The Belling-Lee mobile exhibitors or from W. Birtwistle, 78 Shaw Road, Thorn-
television transmitter, which has done yeoman service ham, Rochdale, Lancs., from whom lecture programmes
in radiating test transmissions from London, Lichfield are obtainable. Separate tickets are being issued for
and Winter Hill during the building of the I.T.A. each lecture.
stations, has now been acquired by the Authority. In
July it will be on Emley Moor, the site for the York-
No 1956 Amateur Show. -It has been decided by
the Radio Society of Great Britain not to hold an
shire transmitter, for test transmissions with an e.r.p. amateur exhibition in London this year. The attendance
of 1 kW. at last year's exhibition was down by nearly 20 per cent
R.I. Club Jubilee. -The 25th annual report of the on that of the previous year and the Society incurred
Radio Industries Club (London) records a record mem- a loss of about £70. The proposal that this year's
bership of 885. At the anniversary luncheon on May exhibition should be held in the provinces- Manchester,
29th Sir Harold Bishop (B.B.C. director of engineering) Birmingham or Bristol -has not materialized.
handed over the presidency of the Club to Eric K. Cole
who introduced the trade name Ekco in 1922 when he E.M.I. College of Electronics has been adopted as
produced-at the rate of five or six a week -a two -
the title of the department of E.M.I. Institutes which
provides full-time day courses in radio and electronic
valve battery receiver.
engineering. Over 200 students are now attending the
B.R.E.M.A. Council.-Member -firms whose repre- full -time courses at the college which offers eighteen
sentatives will constitute the council of the British Radio scholarships for science sixth -formers to undertake the
Equipment Manufacturers' Association for the ensuing four -year course in electronic engineering and two
year are : Balcombe (E. K. Balcombe); Bush (G. scholarships for the three -year telecommunications
Darnley Smith); Cole (G. W. Godfrey); Cossor (J. S. course for which the entrance standard is G.C.E.
Clark); Ferguson (S. T. Holmes); Ferranti (E. Grundy); (ordinary level). Full particulars of the scholarships
G.E.C. (M. M. Macqueen, chairman); Gramophone Co. and the courses, which begin in the autumn, are obtain-
(F. W. Perks, vice -chairman); Kolster- Brandes (P. H. able from the College, 10 Pembridge Square, London
Spagnoletti); Philips (A. L. Sutherland); Pilot (H. L. W.2.
Levy) and Ultra (E. E. Rosen).
The annual contests for A summer school on communication theory, modula-
radio controlled models, tion and noise is being held at Birmingham University
organized by the Inter- from July 1st to 13th. It is intended for engineers in
national Radio Controlled the electrical communications and allied fields who have
Models Society, will be taken their degrees without much theoretical work in
held in the Midlands on communications and for those who feel the need for a
August 5th, 6th and 7th. " refresher " course. The fee is £8, excluding accom-
The contest for aircraft modation. Particulars are obtainable from the Director
will be held on the 5th of Extra -Mural Studies, The University, Edmund Street,
at the aerodrome, Welles- Birmingham, 3.
bourne Mountford, near
Stratford -on -Avon, War-
wickshire, and that for
boats at Bournville, Bir- BUSINESS NOTES
mingham, on the two G.E.C.-B.T.M. Collaboration.-The General Electric
following days. Further Company and the British Tabulating Machine Company
particulars a n d entry are to collaborate in the design of " competitively priced
form s are obtainable computers and like equipment to cover the field in
from H. Croucher, 27 which such devices can be employed, and particularly
St. John's Road, Spark - J. F. DOUST the sphere of office machinery and automation." They
hill, Birmingham, 11. (See page 253) are jointly to form a new company for this purpose.
256 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 RCA Photophone, Limited, which was established in Amplivox celebrates 21 years of hearing -aid manù-
this country in 1929 by the Radio Corporation of facture this year. The Company, of which A. Edwin
America primarily for the introduction of the Photo - Stevens is governing director, also manufactures micro-
phone system of sound recording for pictures, has phones and headphones.
changed its name to RCA Great Britain, Limited. The In the refurbished demonstration centre recently
company is now the sole U.K. distributor for all opened by Philips, at Century House, Shaftesbury
products of the parent organization. Incidentally, the Avenue, London, W.C.2, a " hi -fi " demonstration room
fiftieth anniversary of the formation of the Radio Cor- is provided in which elaborate switching arrangements
poration of America was celebrated at the end of permit various combinations of loudspeakers and ampli-
February. fiers to be compared.
Ekco car radio, Model CR152 /K, has been approved Birmingham Sound Reproducers are extending their
by the Rootes Group as standard or alternative optional factory space in Londonderry, Northern Ireland, to give
equipment for their new Hillman " Minx " special a total of 180,000 sq. ft. B.S.R., who make the Monarch
saloon, de luxe saloon and convertible cars. record changer and other gramophone units, have their
An industrial television camera has been installed by head office and another factory at Old Hill, Staffs.
Pye as part of the new wind- tunnel of the Aircraft A.B. Metal Products are extending their Glamorgan
Research Association at Bedford. Direct observation factory to give an additional 50,000 square feet of
through the usual plate -glass window is not possible manufacturing space.
in this tunnel because of the special method of lining
the walls. Savage Transformers Limited, makers of the well -
known " Massicore " transformers, have added an office
Included in the Marconi equipment being installed block to their factory at Devizes, Wilts. This has
in the recently built Manchester Venture and Manchester released factory space for production.
Vanguard, which will trade between Manchester and
the Great Lakes, is a new ship -to -shore radio -telephone Gate Electronics, Limited, have moved into their
set. Known as the " Seaway," it was designed by the new factory in Tudor Grove, London, E.9. (Tel.:
Canadian Marconi Company to meet the requirements Amherst 8484.)
of the Canada /U.S. Great Lakes Treaty. Electrical Instrument Repair Service, of 329, Kilburn
The new factory at Uxbridge opened by Alfred Imhof, Lane, London, W.9, have changed their name to E.I.R.
Limited, instrument case makers, provides over 25,000 Instruments Limited.
square feet of production space. The other factories New offices and showrooms in Lancashire House,
at Islington and Thornton Heath will remain in use 9 South Street, Manchester 2, have been opened by
but the drawing offices and development section are Marconi Instruments.
now accommodated at Uxbridge.
This year marks the coming -of-age of Painton and OVERSEAS TRADE
Company, of Kingsthorpe, Northampton, manufacturers
of a wide variety of components and accessories, includ- Poland. -On their stand at the International Fair in
ing attenuators and faders. Poznan (June 17th to July 1st) Kelvin and Hughes will
The distribution of Goodmans loudspeakers and be showing echo sounders, industrial equipment,
acoustical resistance units in Scotland is now under- miniature motors and also sound recording and repro-
taken solely by Land, Speight and Company, of 2 ducing equipment manufactured by their associates,
Fitzroy Place, Sauchiehall Street, Glasgow, C.3. Simplex -Ampro, Limited.
Export enquiries for the Sinfonia and President tape
recorders and amplifying equipment manufactured by
Phillips and Bonson, Limited, should now be addressed
to Barnett Shipping and Export Company, St. Magnus
House, 25 Monument Street, London, E.C.3.
Italy. -An order for £20,000 worth of equipment,
including decade oscillators, telegraph distortion
measuring gear and transistor test equipment, has been
received by Winston Electronics, Limited, from the
Italian Government.
Australia. -A three -camera O.B. vehicle has been
supplied by Pye to the Australian Broadcasting Com-
mission.
U.S.A. -William B. Allen, of 1601 Orleans Ave.,
New Orleans, Louisiana, wishes to get in touch with
a United Kingdom manufacturer making a low- priced
transceiver operating on 465 Mc /s. It should operate
on a 6- or 12 -volt car battery and have a range of ten
miles or more.
Yugoslavia. -Pye telecommunication and television
equipment and scientific instruments manufactured by
their associates, W. G. Pye and Unicam, are to be
shown at the Scientific Exhibition at Ljubljana, from
August 4th to 12th.
U.S.A. -Olson Radio Warehouse Co., 275 East Mar-
ket Street, Akron, Ohio, are interested in importing
radio components, accessories, valves, etc., from the
A by-product of the Rechfon GR!74a.m. f.m. radio-telephone United Kingdom.
equipment is that it gives the marine user the opportunity Honduras. -A. Idìaquez y Cia., of Tegucigalpa,
to overcome interference should other vessels nearby be D.C., are interested in importing, to sell in their own
operating on a.m. on the same frequency. The combined stores in Tegucigalpa and San Pedro Sula, a small
a.m. /f.m. 20-channel radio-telephone is shown installed in battery -operated domestic receiver, preferably in a
a deep -sea trawler. wooden cabinet.

WIRELESS WORLD, JUNE 1956 257

L

www.americanradiohistory.com
 Unconventional F.M. Receiver
Designed for Local Station Reception Using Pulse- Counter Discriminator

By M. G. SCROGGIE, B.Sc., M.I.E.E.

THE design of this receiver from the i.f. output on- the valve input capacitances,
wards (comprising limiter, discriminator and de- frequencies, attenuation of the upper
emphasis circuit) has been discussed in detail in the from the frequency eliminating v.h.f. components received
April issue. So far as the rest of the receiver is con- changer and limiting the i.f.
bandwidth at the top end, improving stability. There
cerned, the most significant feature of this discrim- is of course
inator is that its working frequency has to be of the capacitances further top cutting by the valve output
shunting the coupling resistances. Fig. 2
order of 150 kc /s. To meet this requirement there is the amplification
are two alternative methods. One is to use a con- frequency characteristic. With
reasonable care in the layout of the i.f. and discrim-
ventional f.m. receiver as far as the i.f. output and inator stages and the wiring of the decoupling
then interpose a second frequency changer to reduce capacitors, it is unnecessary to use inter-stage screen-
the standard 10.7 Mc /s i.f. to 0.15 Mc /s. This pro- ing.
cedure would combine the usual standard of sensi- The signal required at the grid of V, is at least 3V
tivity with the low distortion and absence of align- and preferably
ment that are the chief attractions of the pulse - gain of V, and about V,
10V. The combined voltage
counter discriminator. It may be that in some cir- V, should be at is about 4,000, so the input to
cumstances they would be worth the extra complica- least a millivolt or two.
tion of the second frequency changer.
The other alternative is to accept 150 kc /s (or The Frequency Changer
thereabouts) as the i.f. This extends the no- align- To reduce the incoming v.h.f. signal to 150 kc /s
ment benefit to the i.f. amplifier also, and the receiver in one step requires the heterodyne frequency to
as a whole is simple, but its sensitivity is low. How- differ from the signal frequency by only about
ever, where an adequate field strength exists this is 0.17 %. This is small enough to suggest the use of
no disadvantage. a self- oscillating coupling circuit between r.f. and i.f.
In the interval since the publication of the dis- stages, but although the idea might be worth follow-
criminator article some readers seem to have been ing up it rejected on the ground of liability to
troubled in their minds about the problem of radiate. was Second-harmonic operation, though it
amplifying a bandwidth of 150 kc /s centred on an i.f. would greatly reduce this liability, was also rejected,
of about the same figure. As it happened, however, in the interests of televiewers on Channel 1, who are
no difficulty at all was experienced with the i.f. somewhat vulnerable to an oscillator fundamental
amplifier; nearly all the trouble occurred in connec- frequency in the 44 to 47 -Mc/s band! Third -
tion with the frequency changer. harmonic operation puts the fundamental in a less
Fig. 1 shows the complete circuit diagram, from awkward band and at the same time still further
which it will be seen that two resistance -coupled i.f. reduces radiation, so was adopted. An incidental
stages are used in front of the limiter. The resistors advantage of the lower oscillator frequency, especially
in series with the grids provide, in conjunction with valuable if automatic frequency correction is
Fig. I. Complete circuit diagram. The portion shown in Fig. 9 of the previous article appears on page 259. Coil L3 is the
tapped coil below L4

258
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 74
Fre- Pro -
72
quency Station gramme
70 (Mcls)
68
88.1 N.H. Tor
66 .3 S. Coldfield
dB 64 F R2,6 B .5 Pontop Pike
7 Medlrum
62 { Blaen Plwy
60 3
.9
Va
89.1 Wrotham
"."'""*........%**********N\
38

. _ __ . _ . - .,..., .n.. .nn ann inn o

.3
.5
.7 Norwich
-
Holme Moss Light

6 8 10 20 30 .9 Wenvoe
FREQUENCY (kcfs) 90.1 Divis
.3 N.H. Tor
Amplification /frequency characteristic of the i.f. amplifier, measured .5 S. Coldfield
Fig. 2. .7 Pontop Pike
between A and B. Meld rum

-
9
Blaen Plwy
91.1
respect, however, Wrotham Third
employed, is the greater latitude in the design of the
.3

oscillator. Against this, the efficiency of frequency would clearly not

be enough to dis-
.5
.7
Norwich
-
Holme Moss

changing is about one -third less than with second-

.9
92.1 Wenvoe
harmonic and two- thirds less than with fundamental. criminate appreci- .3 Divis
For domestic broadcast reception by all and sun- ably against the .5 N.H. Tor
dry, station selection by a continuously variable second channel, .7 S. Coldfield
tuning control belongs, in the writer's opinion, exclu- which with a .9 Pontop Pike
Meldrum
sively to the archaic era. What is required is an 150 kc /s i.f. is 931 Home
Blaen Plwy
instantaneous switch-over from one correctly tuned only 0.3 Mc /s off - .3
5 Wrotham
programme to another. Especially is this true of tune. This comes .7 Holme Moss
f.m., to the tuning of which the non -technical listener half way between .9
Norwich
94.1
has even less of a clue than usual, and where the the carrier fre- .3 Wenvoe
provision of switch tuning is facilitated by the fact quencies of two .5 Divis

that all the programmes to be received come from a other stations, but
group of three transmitters having the same location, both of them give
power and frequency spacing. heterodyne frequencies well within the pass band of
The pattern is shown in the accompanying Table, the i.f. amplifier. That these do not cause inter-
where the three transmitters comprising any group ference is due to a combination of two things : the
low sensitivity of the receiver, and the " capture
can be picked out by vertical movement of the three-
pronged fork seen on the left, with a fixed inter- effect " in f.m. reception. Interference is not
prong spacing of 2.2 Mc /s. (Incidentally, it is help- appreciable in f.m. receivers unless the interfer-
ful to remember that the Light programme comes on ing signal approaches the same order of magnitude
as that being received; but this happens only in
the Lowest frequencies and the Home programme on
the Highest.) fringe areas, for which the sensitivity of this receiver
The tuning of the r.f. stage can be made broad is insufficient. The only signals of comparable
enough to cover all three frequencies in any group strength will therefore be the other two from the
and still give a useful gain, so for simplicity this was local station, and they come beyond the i.f. pass band
on both heterodyne channels. Several types of fre-
done, leaving the oscillator circuit as the only one to quency changer were tried, including all those shown
be varied. Substantially better sensitivity and selec- in designs published in Wireless World during the
tivity can be obtained by variable r.f. tuning, if one last few years, but much the most satisfactory was a
cares to provide it. Whatever were done in this type not shown in any-the ordinary diode with
separate oscillator, as used for microwaves. Attempts
to involve V, in any part of the frequency -changing
process-and especially oscillation-gave very in-
+208V ferior results. Whatever frequency changer is
°28.1A adopted, care must be taken to ensure the absence
Rte of squegging, which at an ultrasonic frequency can
22k11
1W easily be mistaken for instability of the i.f. amplifier.
The oscillator is fairly conventional, and supplies
Rt, about 6 to 7V in series with the signal from the r.f.
4120 stage to the diode D,. Best results were obtained
Ctl,-- with a thermionic diode (type EA50) and rather sur-
Rei
V, 6opF prisingly there was no trouble with hum from the
"live" cathode. Nevertheless, there is obviously a
R90 lot to be said for a germanium diode in this position,
with the exception that the performance at this fre-
C26
4.7kR EF80 P1{µFp
4kfl 47kfl
' OUT
UTP
A F
quency of those types that can take the oscillator
D,
voltage appears to be somewhat uncertain. The
R2,
47kii _ c25 samples tested gave 50% to 70% of the output of
the EA50, but because the best of these provided
C24
=°2 l000pF TioopF
ample level at the limiter it was adopted for its con-
iµFT
R24
venience. The d.c. component across R. is about 1V
o 2701)
o- less than the r.m.s. oscillator input to D,.
259
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 An important component in the signal chain is Substituting 1.8 for SC, and trying C4 =10, we get
the coupling from the r.f. stage L2. Of a large C5=45.5 and C6=29. Now with a fixed L, the
number of couplings tried, the one shown in correct tuning for the Home programme can be
Fig. 3 (b) is the most satisfactory. It is very similar pre -set by adjusting C,. The precise values of SC
to the one described by Amos and Johnstone in the for the other two programmes depend on the total
October, 1952, issue. The rest of the r.f. stage, in-
cluding the input coil L, which is also shown in
tuning capacitance and hence on L and as 0.76 µH
is only a target value for L, there is need for some
Fig. 3, is quite conventional. adjustment of SC to ensure correct tuning of the
For the purpose of calculating the oscillator other programmes. One does not want to have to
tuning circuit values for the three programmes, the pre -set all three capacitors in the switching group;
difference between the signal frequency and three the most convenient is C4, but does adjustment of
times the oscillator frequency can be neglected. it vary both steps equally? Let us assume for the
Each of the outside signal frequencies (Home and moment that it does. Then the calculated values
Light) differs by 2.2 Mc /s from the middle one of C, and C6, 45.5 pF and 29 pF, are rather odd for
(Third), or 2.4 %. This necessitates very nearly 5% fixed components. But putting C, =50 we get
capacitance difference. Theoretically the two steps C4 =10.4 and C,=33, which if not directly available
are not exactly equal, but the difference is small can be made up from 100 and 50 in series. Assum-
enough to neglect. The most obvious arrangement ing these values, and plotting the two capacitance
is that shown in Fig. 4(a). A suitable value for the steps against C, we get Fig. 5, which shows that
total capacitance needed for one -third of any of the although exact equality is not maintained the values
Third programme frequencies is 36 pF, for which of SC can be varied over sufficiently wide limits
the corresponding inductance is about 0.76µH. Of without serious discrepancy.
this, the stray capacitance is likely to be about
12 pF, leaving 24 pF to be provided. Five per cent
of 36 pF is 1.8 pF, so the capacitances required are Automatic Frequency Connection
approximately as shown. The switched capacitances To correct for any such error, and to render un-
are so small as to be complicated by the stray capaci- necessary any particular care in the choice of com-
tances of the switch, and also (because the whole ponents or other precautions for counteracting
oscillator voltage is applied) by its losses. oscillator drift, automatic frequency correction was
Fig.4(b) shows an alternative that avoids these adopted. There are various possible ways of pro-
disadvantages, for the switched capacitances C, and viding it, such as the milliammeter- operated system
C6 can be made relatively large and consequently
low in potential. If the capacitance steps are described by C. H. Banks in the February issue.
denoted by SC, then For the fun of it, the writer experimented with the
C5 =
C4(C4 - SC) core -saturation system shown. The oscillator coil
L, is wound on a small ferrite core placed in the
SC gap of an electromagnet L4, which is energized by
_(C4 - SC)(C4 -2tSC) the anode current of the spare half of V2. A z.f.
SC controlling voltage exactly proportional to the i.f.
is available at the output of the discriminator; as
shown in the previous article, the voltage corre-
V. H.F.
CORE sponding to 150 kc /s is 2;. R6 is adjusted so that
V.H.F.
CORE
with an equal voltage applied from an independent
2 I/ source to the grid of V2b the anode current is a con-
POLYSTYRENE
FORMER
venient value (in this case about 21 mA, mid -scale
on the tuning indicator M) and the oscillator
4 trimmer is adjusted so that with the switch at "H"
the Home programme is tuned in to give 2; V out-
ALUMINIUM L2
put from the discriminator. If this cannot be done
SCREEN
PRIMARY I-2, 3TURNS
with a total capacitance reasonably near the design
CLOSE WOUND & INTERWOUND value, the number of turns on L, must be altered.
WITH 31/z TURNS SECONDARY 3 -4;
THEN 5 TURNS SPACED AS SHOWN When satisfactory, the discriminator output can be
2 3
ALL 30 D.S.C. COPPER WIRE substituted for the independent source. The i.f. and
a.f. must of course be filtered out, for which purpose
I

3/2 TURNS
L
18 TINNED COPPER WIRE.
TAPPED AT ITURN
(b) R, and C are provided.
To hold the tuning, an increase in i.f. must affect
(a) the inductance L, in such a way as to reduce the

Above: Fig. 3. Constructional

details of the v.h.f. tuning
coils, L1 and L2.

Fig. 4. Two alternative L

arrangements of the programme
selector switch, of which (b) is
preferred. The capacitance
values shown are calculated
for L3 -0.76 pH and stray
capacitance 11 pF.
(a) (b)
260 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 LT03
i.f., and a little consideration will show that for this
to happen it is necessary for the oscillator to be
tuned so that its third harmonic is 150 kc /s lower 25
3 TO H
in frequency than the signal.
The oscillator inductor L, consists of 12 turns of
30 d.c.c. wire wound more or less toroidally on a
core of type B4 Ferroxcube, list No. FX.1595, which
is ;in square with a central hole and resembles a ?O

k -in square nut. For experimental purposes a

10,000 c Type 3,000 P.O. relay was used as the
polarizing system L with the armature and springs
(pF)
removed and the circular pole piece filed on one
SC

side to make a $ -in parallel gap to take the L, core. 15

This gave an extremely tight control of tuning with

only about 0.4mA normal polarizing current, but
the low reluctance of the closed magnetic circuit
was found to be very vulnerable to stray magnetic
fields frequency -modulating the oscillator at 50 c/s 1.0

and so causing hum. This could presumably be

eliminated by placing the whole relay in a Mumetal
screen. Another disadvantage was that the ferrite 3TOH

is subject to hysteresis, and if the polarizing current

LTO3

accidentally went outside its normal control limits 05

8 9 10 11 12 13

(say in the setting -up) it was necessary to switch

6 7

C4 (pF)
off and start afresh, in order to ensure that the core
Fig. S. With the values of Cb and Cs as in Fig.
4(b), the
was working on the rising- from -zero part of its
jumps of capacitance between the positions of
the
magnetic cycle.
Because of its lower hysteresis (and higher perme- programme switch vary with C4 as shown.
ability) a ring core of B2 Ferroxcube was tried as an
alternative to the B4, but its higher losses stopped
oscillation at 30 Mc /s. (Incidentally, even B4 would (a) POLYSTYRENE
WEDGES Ly
be unworkable at the oscillation frequency of a
fundamental frequency changer.)
By turning the oscillator core at right angles as
shown in Fig. 6 and wedging it in position with
small pieces of polystyrene as magnetic gaps, the
Poi t
liability to hum was greatly reduced, and the looser
COIL
(L4)
a.f.c. and larger polarizing current (22 mA) reduced \\

the extent to which the tuning could be shifted by MULL ARD

hysteresis. It is still necessary to keep mains trans-

FERROFX.I14XCUBE CORE
6 (b)
formers a foot or two away and check their orienta-
tion, and of course the anode supply to V2b-as in
1 -3: TURNS 30 D.C.C. COPPER,
12
CENTRE- TAPPED AT 2

fact to the other valves-must be thoroughly Fig. 6. Details of the experimental oscillator tuning
coil
smoothed. The actual h.t. voltage is not at all and polarizing magnet for the automatic frequency
critical and can be varied over wide limits. correction system described.
.
Adjusting the Oscillator instant change -over with little or no click, and of
The a.f.c. system and the oscillator circuit having course the limiter ensures that all three programmes
give exactly the same a.f. voltage for a given depth
been set up to receive the Home programme as of modulation. Incidentally, the output is sufficient
described, C, must be set to give correct tuning of to drive a Leak TL /10 amplifier fully without pre-
the other two programmes on working the switch. amplifier.
Correctness is indicated by the reading of M re- For adjusting the cores of L, and L2 and checking
maining constant, or nearly so. If switching over to the adequacy of the i.f. signal, a valve voltmeter from
Light makes the reading rise, C, is too large; with
re- grid of V, to earth is very helpful; if a proper instru-
the switch back at Home it should be slightly ment is not available, one can be extemporized,fre- or
duced and at the same time C8 increased to keep the an oscilloscope used. The low intermediate
total as before. The process is repeated until tuning quency facilitates such measurement. To tune L,
is reasonably correct for both programmes. If the
values of C, and C8 are close to the specified values, and L,f the cores should be adjusted to peak on the
Third programme; then slightly staggered readings(one up
the tuning of the Third should then be good enough. and one down) to give approximately equal
Because with the pulse -counter discriminator the on Home and Light.
i.f. is not at all critical, it is not a serious matter
if
the control current changes even half a milliamp
The receiver here described is not to be regardedinter-
of as a final design but rather as a suggestion for
between stations, but it does reduce the margin esting experiment. The relay as a polarizing magnet
a.f.c. available for dealing with warming-up drift, is only a makeshift that happened to be handy;
no
etc. The meter M is not essential as a permanent designed
doubt a much smaller component couldanbealtogether
feature of the receiver, but is useful for indicating for the purpose. Or one might prefer
departure from correct adjustment. This programme different system of a.f.c. Then the question of how
switching system works extremely well, giving an
261
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 this type of receiver compares with the conventional
as regards freedom from noise and interference has
there should be no appreciable amplitude modulation
not been fully investigated. The writer suspects that to affect the discriminator; but any amplitude varia-
on a level comparison it might be somewhat less good, tions at the input might slightly vary the ratio of
but, just as in the matter of its inferior selectivity, it positive to negative "half " cycles, and in that way
can afford to give something away because of its lower
introduce a certain amount of phase modulation,
sensitivity. The hybrid type -the double super- which would distort the f.m. waveform. This too
heterodyne- combines the advantages of both. has not been fully investigated, but a rough calcula-
Lastly, remembering that one of the chief claims of tion seems to indicate that with the 1 dB drop at the
the pulse-counter is its low distortion, one must upper peak of 100% modulation from 150 kc /s
consider possibilities of distortion elsewhere. One shown in Fig. 2 the distortion would be small. It
such possibility is inequality of amplification over the could be reduced (along with the very slight dis-
i.f. band. Provided that the amplitude at the input criminator distortion indicated in the previous
to the limiter does not vary beyond the limits of the article) by reducing the i.f. to, say, 100 kc /s; but it
almost perfectly level part of the output voltage is necessary to leave enough margin for there to be
characteristic shown in Fig. 8 of the previous article, no chance of the i.f. drifting so low as to clash with
the a.f., for that causes most evident distortion.

LONG RANGE ON V.N.F.

iJRING recent months the solar activity, which Apart from the general effect of the solar activity
was expected to increase at a rapid rate, has done in increasing the ionization of the upper atmosphere
so even more rapidly than was anticipated. The there is another effect which is of importance in
response of the F2 layer to this increase in the radio communication; i.e., as the activity increases
activity of its producing agent has been most marked. there is an increase in the frequency and severity
Its ionization has risen rapidly, and, during the of ionospheric disturbances.
period February-April, 1956, the monthly means of Ionospheric disturbances are of two distinct
the noon F, critical frequencies as measured at and we may deal first with the short -lived kinds,
Slough were of the order 9 -10 Mc /s. ance- lasting disturb-
perhaps for an hour-known
So far as radio propagation is concerned, there- " Dellinger " fadeout. These are produced by as a
fore, it may be said that quasi-maximum conditions; flares (usually associated with sunspots) and solar
i.e., those associated with a sunspot number greater in a complete fadeout of short -wave signals onresult
than 100, already exist. And they are producing cer-
tain frequencies which traverse the daylit hemisphere
some interesting, if sometimes unwanted, results. of the earth. During the first six months
The higher frequencies in the h.f. band are being of
only two such fadeouts were reported and during1955
regularly propagated by the F, layer over long second half of the year only five. But during the
distances, whilst those in the lower part of the period January -April, 1956, no fewer than twenty the
v.h.f. band are similarly affected. such fadeouts were reported, some of which were
The prediction curves published monthly in Wire- of great intensity. "Dellinger" fadeouts are
less World under " Short -Wave Conditions " have likely
to be relatively frequent during the next few years.
given some indication of the very high frequencies But it is the other kind of ionospheric disturbance,
expected to become usable; for example, for 25% known as an ionospheric storm, which constitutes
of the total time over the paths to Montreal and the major disruptive phenomenon in short -wave
Johannesburg. In point of fact these have been communication, since its effects often last for several
somewhat exceeded. U.S.A. amateurs on 28 Mc /s days. This again originates in the sun and is thought
were receivable in this country for almost half the to be due to the emission of a corpuscular stream
days of March and April, whilst mobile radio signals which, travelling at a velocity of the order of 1,000
on 30-35 Mc /s were frequently received from miles per second, enters the earth's atmosphere and
southern U.S.A. From several places in South produces, among other terrestrial effects, the iono-
Africa and Southern Rhodesia, including Johannes- spheric storm. Since the corpuscular streams appear
burg, reports were received of the frequent reception
-almost daily during late March and early April
of the sound channel of the London television ser-
- often to occur in association with sunspots -some-
times reaching the earth about 30 hours after a solar
flare -their frequency and severity can be expected
vice on 41.5 Mc /s. And very often the video chan-
nel on 45 Mc /s was receivable after a fashion. to increase with the increasing solar activity. In-
The solar activity is likely to continue to increase deed, there has already been a marked increase in the
at a rapid rate, though not, perhaps, without con-
frequency of these disturbances.
siderable fluctuations. The seasonal effect in the The next maximum of solar activity is expected
Northern Hemisphere is, however, at present tend- occur sometime during the first half of 1957, and toto
ing to produce lower values of daytime F, layer be one of outstanding intensity. The F2 m.u.fs
ionization, and this may cause some slight reduction may therefore be expected to continue to increase
towards that time, but the occurrence of ionospheric
.

in the daytime m.u.fs towards midsummer. After disturbances to become more frequent and, perhaps,
that, however, the seasonal effect will be operating more prolonged and severe. This state of affairs
in the same direction as the increasing solar activity, should continue during the course of the Interna-
and the m.u.fs are likely to rise to even higher tional Geophysical Year, which is a circumstance
values than at present. By the autumn, therefore, favouring the acquisition of new knowledge during
it is probable that the F, layer will be capable of that period of intensified scientific effort.
propagating phenomenally high frequencies.
T. W. B.
262
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 This article is based on certain sections
Characteristics of of the author's book, "Radio and Electronic
Components, Vol. 1 : Fixed Resistors,"
recently published by Sir Isaac Pitman
fixed Resistors and Sons at 28s.

Sources of Noise, Resistance at High Frequency and Some Measurements on

Carbon Composition Resistors By G. W. A. DUMMER, M.B.E., M.1.E.E

ALL resistors generate noise due to thermal agi- the higher the direct voltage across the resistor the
tation of electrons. This is known as Johnson noise, higher the noise.
after J. B. Johnson who discovered the effect in 1927. Another important factor is the size of the resistor
The random motion of free electrons in a resistor in relation to its wattage; with a given value under
is in equilibrium with the. thermal motion of the similar current conditions the noise in a small resis-
molecules. This random motion is superimposed tor will be greater than that in a large resistor Fig.
on the electron drift current arising from the poten- 2 shows the relative noise voltage developed in three
tial difference across the resistor. This produces a different sizes of carbon composition resistors using
fluctuating voltage, or noise, and its r.m.s. value is an a.f. amplifier with a bandwidth of about 10 kc /s.
only proportional to resistance and temperature. Measurements have shown that for carbon cotn-
It has a uniform spectrum of frequency distribution
and if different bandwidths are selected for a fixed Io Mn

value of resistance the relation between generated

noise and bandwidth for a 10 -kt resistor will be
approximately as given in Table I. I M!1
For a given bandwidth the relation between John-
son noise and resistance value is approximately as
shown in Fig. 1. In general the higher the resist-
ance value, the greater the noise.
Carbon composition resistors also generate noise a`
due to the passage of current through them. This
causes random changes in the material of which
the resistor is made. This noise, sometimes called
"Bernamont" noise (after Bernamont's work in kn
1934), is also greatly dependent on bandwidth.
I

Unlike Johnson noise, however, which is uniform

at all frequencies, current noise decreases with in- loon
creasing frequency from below about 10 c/s up to 0I I0 10 loo
at least the kilocycle range. The generated noise JOHNSON NOISE VOLTAGE p.V(r.m.$)
(E) is related to bandwidth in the following way:
Fig. 1. Relation between Johnson noise and resistance

E' -
a /2-
value.

where f, and f, are the lower and upper frequency

limits of the noise -measuring equipment or ampli-
fier, so that for a constant bandwidth the square of
this noise voltage is inversely proportional to the
frequency.
Current noise is particularly dependent on the
physical construction and materials used in the
manufacture of the resistor and can vary greatly
between resistors of similar types and even between
those of the same value and manufacture. In general
TABLE I

Bandwidth of Johnson Noise

Amplifier or Measuring Voltage (r.m.s.) in TWO -WATT
UN NSULATED
Equipment µV (approximately)
//////////////////////////////////// /// ///
JOHNSON NOISE LEVEL

1 kc/s 0.6 50 0 ISO 2.0 250 300

10 kc/s 1.5 VOLTS ACROSS RESISTOR
100 kc/s 6.0
1 Mc/s 20.0 Fig. 2. Noise voltage from carbon composition resistors
10 Mc/s 60.0 of the same value but different wattage ratings.
Crown Copyright Reserved

263
WIRELESS WORLD, PINE 1956

www.americanradiohistory.com
 position resistors current noise increases linearly and inductance as well as resistance. A simple equi-
with current up to about 15µA. With greater cur- valent circuit of a resistor is that of Fig. 3. R is
rents the noise curve approximates to a parabola. the "pure" resistance, C, may be considered the
On a range of carbon resistors of different values sum of many small distributed capacitances along
up to 1.5 Mn the current noise generated at normal its length and L is the lead, and any other induct-
voltages may vary approximately as in Table 2. ance in the actual resistor. C, is the mounting
It is of interest to note that the total noise allowed capacitance when the resistor is included in a circuit.
in composition resistors for Service use is given At any one frequency the resistor can be repre-
by sented by the simple parallel circuit of Fig. 4. Here
2 +logs µV /V Co may be positive or negative. For low values of
'\ 1,0001 resistance the variation of Cri with resistance value
where R is the resistance in ohms. The noise - takes the form of the curve in Fig. 5. The r.f.
measuring equipment has a bandwidth of from 200 resistance, Rrj, again for low values of resistance,
to 10,000 c /s. On this assumption the maximum and plotted as the ratio of Rrl/Rdc against the value
amounts of noise allowed for various values of car- Rd, (d.c. resistance) is as shown in Fig. 6. It can
bon composition resistors are given in Table 3. The be seen from this curve that, due to the series
third column in this table gives the maximum noise inductance, the r.f. resistance of very low value
generated in cracked carbon resistors under equiva- resistors rises above the d.c. values.
lent conditions and is included here for comparison. In general the r.f. resistance, or impedance, of all
R.F. Performance. -A resistor must, by its con- resistors decreases with an increase of frequency
struction, contain a certain amount of capacitance because of the distributed capacitance (C Fig. 3)
and the higher the value of the resistor the greater
the fall in r.f. resistance. The performance of typical
carbon composition resistors in values from 1 ks2 to
1 M12 and at frequencies up to 40 Mc /s is shown in
Fig. 7. Here the fall in resistance is plotted as the
ratio of r.f. to d.c. resistance against increasing
frequency.
Low -value resistors can be used as load resistors
up to several thousands of megacycles and their
Fig. 3. Simple equivalent circuit of Fig. 4. Equivalent behaviour at frequencies up to 1,500 Mc /s is shown
a resistor. circuit of a re- in Fig. 8. A useful rough rule is that when R, is
sistor at a given less than 0.03 (where R is in Mil and f in Mc/s)
frequency. the r.f. resistance of a carbon composition resistor
is within 10 per cent of the d.c. value.
Measurement of Noise. -The usual method of
C*./
measuring the noise generated in a resistor is to
feed the noise voltage into an amplifier of known
gain and bandwidth and to detect the noise voltage
APPROXIMATELY 50 TO 80.51 on a thermo -couple voltmeter. The noise voltage
is generated by applying a direct voltage across the
resistor under test and a matching load in series.
Certain precautions must be taken to ensure that
Fig. 5. Variation of Crr with Rda the extremely small voltage to be measured is not
masked by stray pick up or instrument- generated
noise voltages. Such precautions include screening
of the input circuit, effective mains filtering and the
APPROXIMATELY 50 TO 80 !i use of low -noise valves in the early stages of the
_y RA
amplifier. Resilient valve mounting might be
needed in the first amplifying stage to avoid
microphony effects.
Measurement of R.F. Resistance.-There are two
main systems of measuring the r.f. resistance of a
Fig. 6. Variation of Rrr /Rdo with R. resistor. One is of a more fundamental character,
involving standards of capacitance (resonance
TABLE 2 TABLE 3
Volts (D.C.) Approx. Current Johnson Maximum Noise (µV /V)
applied Noise (t V) Noise
to resistor (V) (µhr) Resistance Carbon Cracked
Min. Max. Value Composition Carbon
Resistors Resistors
50 250 1,500 30-35
100 350 2,300 30-35 1 kf2 2.0 0.03
150 375 2,900 30 -35 10kí2 3.0 0.15
200 400 3,400 30-35 100 kit 4.0 0.35
250 450 3,600 30-35 1 MO 5.0 0.50
300 450 3,800 30-35 10 Mû 6.0 0.52

264 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 10
In all measurements on resistors at radio frequen-
1kí1 cies the method of mounting the resistor is important.
2kí1 The end-to -end capacitance of the resistor and the
0.8 capacitance of the. two leads to the resistor body
are included in the total capacitance of the resistor
being measured and the resistor should therefore
0.6
R,./ Fig. Variation of
7.
be mounted as nearly as possible as it is when in
Rdc 50kí1 R 'Rd,with frequency for use.
04 different values of carbon Summarizing, for a resistor to be suitable for
composition resis.ors. operation at radio frequencies it should meet the
following requirements:
02 (1) Its dimensions should be as small as possible.
Mf! (2) It should be low in value.
(3) It should be of the film type.
0 10 20 50 40 (4) A long thin resistor has a better frequency
FREQUENCY (Mc /s) characteristic than a short fat one.
(5) All connections to the resistor should be made
25 as short as possible.
(6) There should be no sudden geometrical discon-
tinuity along its length.
Mc /s
loo
200
500

40 60 80 100 120 14.0 160 180 200

1.500

220
Origins of Radio Telephony
RESISTANCE (OHMS)
IN the April issue " Free Grid " enquired when radio G.
Fig. 8. Behaviour of low value carbon composition telephony was first used. A Danish correspondent,
resistors at high radio frequencies. Bramslev, happens to have been conducting some
researches into this obscure branch of radio history, and
sends some notes on his findings. of
In Fessenden's U.S.A. Patent, No. 706,747 by
September 27th, 1901, relating to generation of r.f. the
OSCILLATOR E means of an alternator, there was a claim covering
use of the apparatus for telephony, with a telephone
receiver at the receiving station. However, no details
of the receiver were given.
In 1903 Valdemar Poulsen of Denmark developed the
Fig. 9. Substitution method for r.f. measurements on oscillating arc, and in his Danish patent, No. 8208, of
resistors. September 27th, 1904, stated that wireless telephony
could be realized by putting a microphone inof the the
oscillatory circuit and so modulating the amplitude
methods) or lengths of standard line (transmission- oscillations.
line methods). The other uses resistors as standards Thus far, there were no claims to practical realiza-
(bridge substitution methods). Resonance and tion, and the first descriptions of actual experiments
transmission -line methods make no preliminary which Mr. Bramslev has been able to trace were pub-
assumptions about the resistors to be measured; lished in 1906. In the German periodical Elektrotech-
bridge methods are in most cases comparison sys- nische Zeitschrift of November 15th of that year, Ernst
tems. A simple substitution method can take the Ruhmer wrote about radio -telephony tests he had con-
ducted with a Poulsen generator, carbon microphone
form shown in Fig. 9. R, is a low value standard " and an electrolytic detector in the receiver. The distance
resistor and R is the resistor to be measured. R covered was only 30 yards. An interesting fact is that
and R, are brought alternately into the circuit until Ruhmer mentioned both amplitude and frequency
a value of R, is found which results in the voltage modulation, and, apparently, fully understood the nature
across the resonant circuit being the same with of both systems.
both R, and R. The unknown resistor (R) is then Fessenden, after having successfully demonstrated the
given by use of his rotary r.f. generator, turned to practical experi-
ments in telephony, and in December, 1906, transmitted
both speech and music over quite long distances from
z
R =R, (1+ Cz) his station in Massachusetts, U.S.A.
What Mr. Bramslev says about Fessenden is borne
Transmission -line methods use slotted -line im- out by Blake's " History of Radio Telegraphy and Tele-
pedance- measuring equipment in which the termin- phony " (1926), which states categorically He was the
ating impedance of the line is determined by plot- first to use a microphone in the aerial circuit." Blake
ting the voltage standing wave pattern along the adds that, by the use of telephone relays of his owna
line. The length of slotted -line required is equal design, he demonstrated the possibility of connecting
to half a wavelength at the frequency being used. land -line telephone to a radio -telephone station. Accord-
ing to a quoted statement of Fessenden's, speech Rock was
The standard, with which the resistor whose r.f. " transmitted over land -line to the station at Brant
resistance is required to be found is compared, is and re-transmitted wirelessly by a telephone relay,
a length (or rather number of lengths) of accurately received wirelessly at Plymouth, and there relayed on
made short-circuited line. These " standard lines " another land-line." In 1907 Fessenden successfully
are used to calibrate the measuring line. All transmitted speech over a distance of 200 miles. arc
measurements are then made in terms of lengths Undoubtedly both the r.f. alternator and the
and the method is applicable up to about 3,000 Mc /s. played important roles long before the valve was used.
265
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 MORE ABOUT NOISE
Answers to Some Questions By CATHODE RAY"

BEFORE going on with this subject we had better blame you if you aren't -you will have to refer to
recapitulate last month's findings. the books, because frankly I have yet to find any
" Noise " in the title is not a class of sound, but way of showing it that is both simple and convincing.
" unwanted energy " in a communication system. Certainly the noise is greater the greater the frequency
True, it can cause unwanted sound, but it can also band accepted by the measuring device-that is
cause unwanted marks on paper or unwanted flecks what one would expect, and it is confirmed by experi-
on a cathode -ray tube. Our particular inquiry is ment -but if the relationship is as per Nyquist it
confined to what is called thermal noise -the energy looks as if over an unrestricted frequency band the
of movement of free electrons in solid matter. This noise power is infinitely large, which no one will
mechanical energy is what happens to the heat believe. I did point out that in practice the frequency
energy that has been put into the matter to raise its band is never unrestricted, because any resistance
temperature from absolute zero (- 273 °C). And
because (according to the first law of thermo-
has some stray capacitance which increasingly
shunts it as the frequency goes up. But quite apart
dynamics) there is a fixed rate of exchange between from that the assumptions on which Nyquist's
the two kinds of energy, and temperature is pro- formula is based cease to apply somewhere in the
portional to heat energy, noise is proportional to region of 6,000,000 Mc/s. So even if there were no
absolute temperature. And because (according to other restriction of bandwidth the maximum noise
the second law of thermodynamics) one resistor power available from anything would be limited
connected in parallel with another cannot raise the (at 27 °C)to 1.38x 10- 23x300x6x1012 = 2.5x10 -8
other to a higher temperature than itself, it can easily watts = 0.025 microwatt, which is not going to
be shown that the square of the noise e.m.f. in any blow anyone's head off.
part of a circuit must be proportional to the resistance
of that part, and the noise current squared must be
inversely proportional to the resistance. The same Maximum Noise Power
conclusions follow via a more complicated chain of The next question that may be forming in the
reasoning (which I only outlined) from basic assump- mind is: why this emphasis on maximum available
tions about electrons in matter, such as that in a noise power? Shouldn't we be more interested
uniform piece of material the free electrons are in obtaining the minimum? True, but the minimum
uniformly distributed, and that the conductivity of is obtainable by short- circuiting the resistor respon-
the material is proportional to the number of these sible for the undesired noise and thereby bringing
electrons per unit volume. Because the number of it to zero. This would also bring the desired signals
electrons is enormous and their movements are to zero so would be of no practical benefit. The
completely random, the average net number moving condition for maximum noise, on the other hand, is
in any particular direction can be calculated with normally the condition for maximum signals; and
considerable precision on a basis of probabilities, in any case it is a standard condition for ready com-
in exactly the same way as the average departure parison and it has the simplest formula-kTB.
from exact equality of heads and tails when vast Next, if the noise e.m.f. is proportional to the
numbers of coins are tossed. The maximum noise square root of the resistance, it would be greatest in
power that any part of a circuit can deliver to another material of the highest resistivity, which has least
-which happens when, by another basic principle, free electrons. A perfect insulator, with no free
the resistances of the two parts are equal -is the electrons, would presumably generate infinite noise
same for all resistances. It is kTB, where k = 1.38 voltage! Does this not contradict the theory that noise
x 10 -E3 (Boltzmann's constant), T is the absolute energy is the energy of free electrons? If one rashly
temperature (centigrade degrees above - 273), and
B is the frequency band in cycles per second. This
pursues matters to infinity or zero, however, one
must be prepared for absurdities. Any finite amount
is the key formula for thermal noise, and is due of power, however small, developed in a resistance,
to Nyquist of negative feedback fame. Because the would result in an infinitely high voltage if the
terminal voltage when a generator is supplying a resistance became infinite. At the same time the
matched load is half its internal e.m.f., it follows current would become infinitely small. But there is
that the internal r.m.s. noise e.m.f. in any resistance no possibility that this extreme state of affairs could
R is \/(4kTBR). ever be reached in practice, because even the best
Even if you have read the previous instalment and insulators contain some free electrons; and moreover
possess at least a glimmering of an idea why the noise
voltage squared should be proportional to the absolute
temperature and the resistance, and are prepared to
accept the need for some constant to make the units Fig. I. This can be regarded as the
right, you may be inclined to query B. All I said simplest noise -making circuit, because
about it was that since the electron movements are any resistance is shunted by stray copa-
completely random their fluctuations responsible for citance, such as the grid capacitance of
noise occur equally at all frequencies, so noise power the amplifier that reveals the noise.
is uniformly distributed over all frequencies. If you
are not satisfied with that explanation-and I don't
266 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 There are several alternative ways of dealing with
this problem. According to one point of view,
sticking strictly to Fig. 1, the resistance is a fixed
iRc.
quantity R9, which contains within itself the equiva-
Fig. 2. This circuit is equivalent to Fig.1at any lent noise e.m.f. E; this is loaded by C, so thereofis Ca
particular frequency, if the values arc correctly voltage drop in R depending on the reactance
chosen for that frequency. and therefore on frequency. That is where B ccmes
in.
Tc According to another point of view the voltage
drop can apparently be dodged, because the impedance
of RD and C measured between the terminals can
the higher the resistance the narrower the frequency be exactly imitated at any given frequency by R,
band, owing to the influence of stray capacitance. and C' in Fig. 2, provided that the values of R, and
And that is the cue for dealing with another C' are correctly chosen for that frequency. The noise
possible cause of doubt: the effect on noise of circuit e.m.f. in R, supplies no external current, so appears
reactance. But before going on to that there are in full at the terminals, and there is therefore no
still some possible queries about resistance. Although need to allow for the capacitive loading. But the
I talked a lot last month about " resistors " as the value of R, depends on frequency, so in dodging one
source of thermal noise, it was not meant to imply that complication one runs into another. As a matter of
noise of that kind originated only in components fact it is exactly the same complication!
sold in shops as resistors. Just as much noise arises
in a 500 -ohm transformer winding as in a 500 -ohm Comparison of Methods
resistor, except in so far as the resistance may change
with frequency. That proviso raises the question Let us compare the two methods by lumping
of how, for purposes of noise calculation by means \/(4kTB) together and calling it K for short, and
of Nyquist's formula, one reckons the resistance. calculating the terminal voltage (call it V) by each
Is it the resistance of the wire or whatever the method in turn.
circuit is made of, or does it include the effects of Call the noise e.m.f. in Fig. 1 E9, and the reactance
losses in iron cores, capacitor dielectrics, etc? of C X,. Then
The answer is that in reckoning the part of the E,, = KA/R,
total noise coming within any narrow frequency band
the resistance that counts is the value actually the X°
measured at that frequency. It would includeresis- and V = EDVXc2
effects of all incidental losses-even radiation ,,2
tance. This may seem rather surprising, andone it
can lead to some quite involved arguments if
tries to explain theoretically how some of the more V = KX°
R,,
" fictitious " components of resistance make their So
is/ Xc2 + Rp2
noise contribution; but the fact remains that it is
supported by careful experiment. Now call the noise e.m.f. in Fig. 2 E,. Then
And now we come to reactance. Of itself it doesn't E, = KA/R,,
cause any noise, but it can affect resistance noise in
two ways. It always brings with it a certain amount The formula for finding the value of Rs equivalent to
of resistance, corresponding to power losses; for Rp is
X °2Rp
instance, eddy currents in the wire of an inductor R,-
(and in the core, if there is one). And even a pure X °2 +Rp:
reactance affects the a.c. resistance of a circuit
measured between any two points. So V=E,=KX, X °2RP-f- Rq=
Calculating Noise Voltage which is exactly the same as by the otherthe method.
in fact, the simplest it doesn't matter whether we reckon noise
Fig. is a simple example; So
1
like real life, for even an isolated e.m.f. for the actual resistances in a circuit and then
that is anything taking
resistor has some stray capacitance, which can find the voltage between any two points by or
by C, and which in into account the rest of the circuit, or calculate
approximately be represented resistance between the two
practice would be augmented by the input capacitance measure the equivalent
of the amplifier without which the noise wouldnoise not points and apply the Nyquist formula directly to it.
Having disposed of that, we can tackle the band-
be noticed. How does one calculate the According to the Fig. 2 point of
voltage between the terminals? width problem. zero to
The noise formula, once again is view the bandwidth goes all the way from
E = V4kTBR infinity -or at least 6 MMc /s -but the equivalent
falls
or E2 = 4kTBR resistance, and hence the noise voltage, at very
with frequency, so the contributions
E being the e.m.f. of an imaginary noise generator high frequencies are negligible. According to
in series with the resistance R. Of the factors -2a in the Fig. point of view the resistance is definitely
involved, k is a known constant, 1.38 x 10
1
proportion of noise voltage that
in R,, but the
the usual system of units; T is the temperature ordinary actually appears between the terminals falls with
degrees absolute or " Kelvin " (°K), and for as
frequency, so the bandwidth can be regarded
room temperature can usually be taken as is290; we have seen, both come to the same
and R in this case is R-or is it? And what the limited. As
thing. If we use the Nyquist formula to calculate
frequency bandwidth?
267
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 the terminal voltage squared per cycle of B at various Before we get too excited about this, however,
frequencies, we can plot a graph like Fig. 3. The we
proper way of calculating the total noise voltage is should remember it is only a guess, and to see
to add together all these voltages-squared for every the result of the consultation with Integralwait
Calculus.
single cycle from zero frequency to infinity and take This consultation is reported in full as Appendix I,
the square root of the result-or to be perfectly and the result is remarkably simple:
correct we should narrow the slices of bandwidth hT
from whole cycles per second to infinitely small
fractions and add together the infinitely large number C
of them. The only thing wrong about the guess, then,
To some readers such a task might seem to be in that it was - /(2/77) or 0.8 times the correct value, oris
the same class as those set by unbelievably beautiful 20% low. In practice, however, it might well be
princesses in fairy stories in order to liquidate tire- more nearly right, because even the widest-band
some suitors. But, just as in the fairy stories, there amplifier doesn't go up to infinite frequency, so some
is a magic formula provided by a powerful wizard, of the noise included in the " true " formula is bound
named Integral Calculus. Those who are not on to be cut off. Of course, if the amplifier bandwidth
good terms with this wizard will no doubt be obliged is even less, say, than A, then one should use the
to make as good a guess as they can. They might basic noise formula, reckoning the amplifier band-
pick out the frequency where the voltage -squared per width as B, and R,, as the resistance (if it is reasonably
c/s had fallen to a half, and reckon as if it existed constant over the band).
at full strength up to that frequency (call if fi) The V2 = kT /C result is so simple and con-
and then dropped to zero. In other words, they venient, getting rid as it does at one stroke of both
would substitute the shaded area in Fig. 3 for the the rather hazy factors in the Nyquist formula, that
area under the curve, reckoning R in the formula as one would expect it to be included in all the books
R , and B as f,, and trusting to luck that and articles on the subject. It may be that I have
all the bits of noise beyond fi would leaving out
just about
looked up the wrong ones, but the nearest any of
cancel out the error of reckoning them in full up to them go to it is to state the integral equation without
ft. bothering to work it out. No doubt innumerable
irate authors will now write to point out that I have
TERMINAL
VOLTAGE
inexcusably overlooked their clear presentation of the
SQUARED matter. However, what I have overlooked others
PER
4k1R, may have too, so it may help to press the point home
if we see what sort of noise voltage to expect across
(4kTR,)
FREQUENCY
a resistor-any resistor -at
the input of a valve
amplifier. The factor k is absolutely fixed, and for
Fig. 3. Graph of the square of the noise voltage ordinary situations T is virtually fixed at about
per CIS 290° K, while C is likely to be of the order of 10pF.
between the terminals of Fig. I plotted against frequency.
Total noise is represented by the area under the
curve;
So
an approximation to it is the shaded area, the boundary
being the frequency at which the curve is half
ft =1.38 x 10 -23 X 290
maximum height.
its V
10 -'1
20 x 10 -'
If the voltage -squared has dropped to a half at =20 microvolts.
It is seldom likely to be much more than this,
fi, the voltage must have dropped to .N/f or 0.707. because to double it would necessitate reducing C to
Either way, it is a drop of 3 dB. This may remind 2.5pF, and almost any valve's C,o would contribute
us of the common practice of regarding the frequency more than that. It might be much less, however, and
band of a single resistance -reactance combination this formula is useful for estimating the noise from a
as ending where the loss is 3 dB. Besides " looking crystal or electrostatic microphone. But except for
about right," this practice is convenient because the high resistances the noise voltage actually producing
frequency chosen is the one at which the reactance results at the output would usually be less-perhaps
is equal to the resistance; i.e. much less -than in this formula because of the
1 restricted bandwidth of the amplifier. So it is
2ITfiC
= R, useful to have some idea of how much of the
theoretically infinite bandwidth represented in the
so fi = 1
kT/C formula does actually contribute appreciable
2nCR noise. And this is almost where we came in, for
But is this the fi we want ? Is the frequency
which the impedance of Fig. 1 to signals is ./¡ at we have already done something like it before-.
times R,, necessarily the frequency at which the noise with the help of Fig. 3. But we can do it rather
voltage is reduced in that proportion? If so, then more scientifically now by taking as the effective
Rs at that frequency must be ¡RD. bandwidth of the Fig. 1 circuit the bandwidth that
Well, one
only to equate IR to the formula for R3 in termshas would include the same amount of noise from a
of
R,, to assure ourselves that it is. So we can substitute constant resistance equal to R,,. This can easily be
our formula for fi in place of B in the Nyquist noise found by equating the noise-squareds given by the
formula, to discover that the terminal noise voltage two formulae:
works out at kT
4kTR,, B
2kT
nC Whence B =
The interesting thing about this is that resistance 4CR,,
1

has completely disappeared, and so has frequency.

Since B begins at zero this really gives the
268
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
" effective " top frequency; call it fe. (An alternative bandwidth as B. To get an idea of the sort of
way of getting at it is from what we already know, values, let us consider a v.h.f. circuit with a dynamic
that it is it /2 times fi.) Here, from the use of this resistance of 10k S1 followed by an amplifier with a
bandwidth of 500kc /s. Then ß/(4k TBR) =
result, are some representative values of fe on the
assumption that C = 10pF (see table, left). V(4 x 1.38 x 10- 28 x 290 x 5 x 105 x 104) = 9ILV.
So for the lowest values Typical v.h.f. single tuned circuits with Rd -10161
of resistance the noise would generate appreciable noise over a wider band
R, f. voltage indicated by the than 500kc /s; so in this case the amplifier selectivity
kT/C formula has no great is cutting the noise somewhat. This is just as well
1 S2 25 kMc/s practical significance, be- when calculating by Nyquist; because it ensures that
10 2.5 cause no amplifier would the resistance is not too far below Rd at the edges of
100 250 Mc/s have sufficient bandwidth the frequency band.
i kS2 25
to make all of it evident, However, if we want to have an idea of the noise
10 2.5
and even if it could its own voltage at the terminals of the circuit itself, irrespec-
100 250 kc/s
noise would swamp the tive of how it may be pruned by the selectivity of the
1 MO 25
10 2.5 resistance's. Even the amplifier, we can approximate to it by the method
medium values are likely we used for Fig. 2 in Fig. 3. Of course the curve
to be limited more by the falls away in both directions from the resonant
amplifier than by C. But with resistance of the frequency, which we will call fr, as in Fig. 5. So
megohm order the effective noise voltage may well there are two frequencies at which the noise voltage-
be as much as 20µV. squared (and therefore the equivalent series resist-
A still more important noise maker is the ordinary ance R,) is half the maximum (Rd), and the distance
tuned circuit, which can be represented by Fig. 4. between them is what we will regard as B.
Its resistance, measured between the terminals, Now it can be shown (see Appendix 2) that, as in
increases from the relatively small value r at zero Fig. 1, these fi points are also the frequencies where
a signal would be reduced by 3dB as compared with
frequency to a maximum Rd at resonance and then
down almost to nothing at very high frequencies. spot -on tuning. And it is well known that the
(Incidentally it should be remembered that r itself frequency band B between these points is very
varies somewhat with frequency.) This is a good nearly 1/Q times fr. And as Q = 2afrL /r we can
example of how the apparent resistance rather than calculate B thus:
the actual resistance in the circuit counts for noise
voltage; for the smaller r is, the larger Rd. In this
it
Q
- irr = r
2rrf,L 2irL
particular case it is easy to see why the noise voltage Then we fill this value for B into Nyquist:
should correspond to Rd rather than r, because those
parts of the noise voltage generated in r at or near the V = V4kTBR
resonant frequency are magnified by the resonance 4kT rR
effect, just as would signals injected into the circuit 2irL
at the same place. the
We are assuming that R has the value Rd over this
band B, and Rd is known to be L /rC, so putting
V.H.F. Tuned Circuit in we get
On presenting this problem to our wizard, we 4/4kTrL
receive from him some rather awkward -looking V = 2 arLrC
instructions, and the usual procedure is to byinput
-pass
these by arguing that an amplifier with a tuned less 2 kT
circuit will be almost certain to have a more or that n C
clearly defined frequency band no wider than at This has a strangely familiar look-yes! it is
over which the input circuit dynamic resistance is identical with the result we obtained for be Fig. 1 by
or near its maximum value, Rd. So the Nyquist the same method, and which we found to
of very
formula is used, with Rd as R and the amplifier much the same order as the theoretically exact

(Right) Fig. 4. Another important circuit

the parallel tuned circuit.
- figure \/(kT /C) obtained by integration.
should not be surprising to learn that this too.
the tuned circuit
So it
is the
theoretically exact figure for
Again, the value of this delightful simplicity tends
and
to be rather lost because so often the amplifier that
what have you restricts the frequency band then
applies so far as results are concerned. And
Nyquist may be more helpful.

R5
Other Resistors ?
Mention of the amplifier may leadalong to question
on why I have been assuming
that the
noise comes from one particular resistor -the one at
the input of the amplifier-as if all resistances in the
-makers.
lvz ld l FREQUENCY
amplifier or indeed anywhere were not noise their
Well, of course they are, but the point is that
11z

noise is usually amplified so much hess

Fig. 5. This diagram can be used in
estimating the
neglected without making that it n Ae
difference.
Fig. 4.
total noise voltage between the terminals In
269
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 possible exception would be the coupling between
first and second stages, if the first had an unusually APPENDIX I
small amplification, as might be so at very high Calculation of the noise voltage V at the terminals of
R,, and C in parallel (Fig. 1) over an infinite frequency
frequencies. A more pointed objection to any band.
assumption would be that the first valve itself
contributes noise, which is rarely negligible. It is R,
" shot noise," which is similar in some ways to V2 = 4kT Rw,.df where R, _ (caCR,)2 + 1
thermal noise and different in others. There is no o

space here to go into these differences, except to = 4kTR c>"'

df
mention that the noise current is generally pro- °J (27rfCR,)2 + 1
portional to the square root of anode current. In 4kTR, df
practice, first -valve noise is usually lumped in with (27rCR,)2 2 1
first- circuit noise by specifying it as " equivalent
noise resistance "; that is to say, the resistance o f (27TCR,)2
which, if connected at the input of an imaginary 2kT[ la
noiseless but otherwise equal valve, would cause the
o
same noise at the output. So if the equivalent noise = 2kT n
resistance of a valve were 3,50011 and the input arC 2
circuit had a resistance of 8,0001Z, the noise output kT
would be the same as that given by a noiseless valve C
with a 11,500 S2 input circuit. /kT
Lastly, if the noisiness of a given resistor depends So V =
C
only on the things already mentioned -temperature,
k, etc.-and not at all on the quality of resistor, why APPENDIX II
are some classed as " low noise "? Is it just The series resistive component R, of the impedance of
advertisers' licence, without any scientific basis? the Fig. 4 circuit between the terminals is
By no means. The thermal noise we have been r

discussing at such length is only one of the kinds in

non -metallic resistors, which are by far the most Let R1
R'- w2LC)2 + w2C2r2
(1
be the value of R, at the points on the
slope of
commonly used. In carbon composition resistors it the resonance curve where the whole impedance is V4
may contribute only a small fraction of the total
noise. The reason it has been (I hope) worth
times that at resonance (Rd). Unless resonance is very
flat, the ratio of the resonant frequency, to the frequency
difference between these points, B, is very nearly equal
f
discussing is that it is an irreducible minimum, so it
forms a standard by which total resistor noise can be
to the Q of the circuit. With the same proviso, 1
w2LC at either of these points is very nearly equal to
-
judged. The other noise is caused by variations in B /fr.
resistance when current is flowing, and these of
course result in fluctuations of voltage across the So 14 r rQ2
1 1 = 2
resistor. Unlike thermal fluctuations, they are worst
at lowest frequencies; they tend to predominate over
thermal noise below about 10kc /s. As one would
At resonance,
Rd
1-
=
w C a 0, so
Q2+-
rQ2
expect, they are proportional to current. In " high -
stability " cracked- carbon -film resistors the current So Rl ^- Rd
2
noise is very much lower, being almost unmeasurably
low except for low and medium resistances but rising Mr. Wiktor Tell has pointed out a slip in the algebra
rapidly above 1M O. One should of course never use at the end of the May instalment. Instead of saying y is
proportional to T it should have been VT. The final
a composition resistor across the input of a high -gain
amplifier. conclusion then works out to I oc v/T /R, which really is
in agreement with previous results!

C. H. L. ED WARDS, author of the art-

icle in our March issue describing an
amateur transmitter- receiver, is well
known in amateur circles as operator of
G8TL, at Theydon Bois, Essex. On the
left in this photograph of the station
are two receivers (AR88 and CR100), in
the centre is a 160 -m transmitter to
the right of which is an all -band 75-
watt t,ansmitter and a 35 -watt
modulator. The transmitter- receiver
described in the March issue is the
smaller of the two sets behind the main
transmitter.

270
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
Physical
Society's
Exhibition
NEW ELECTRONIC DEVICES
AND TECHNIQUES

Valve and allied devices as well as

test and measuring instruments
shown at this exhibition are
described in a separate report in
this issue.
Decca Radar digital computer using magnetic -core circuits.

RESEARCH for use as a frequency standard in the field. The

resonator is a gapped quartz ring about 1 inch in
Communications. -The assessment of multi- channel diameter which is maintained in vibration by a
communication systems from the point of view of noise transistor amplifier. Sound is emitted by a hearing -
and intermodulation is a tedious process when carried aid earpiece, loaded by a quarter-wave pipe. In the
out by normal single frequency measurements. Quicker region of 20° C the temperature coefficient is 5 parts
and more realistic results are obtained in a method in 10' per degree C.
developed by G.E.C. Research Laboratories in which For the maintenance of oscillations of very low fre-
the total multi- channel signal is simulated by noise of quency (less than 1 c /s) thermistors may be used to
the appropriate bandwidth. If band -stop filters are simulate inductance by virtue of the thermal lag of
introduced at the source to represent unoccupied current behind the applied voltage. The Radio Research
channels and the output in these channels is examined Station showed a simple para',lel capacitor-thermistor
after passing through the system and related to the combination with a period of several seconds working
signal power in an " occupied " channel of equal band- in conjunction with a pen recorder. It was stated that
width, the ratio enables a rapid assessment of per- the equivalent inductance of a type A5412/100
formance to be made. thermistor under these conditions was 8,550 henrys!
A time -division method of transmitting two -way
audio signals on a line without the use of amplifiers Semi- conductors. -Research into semi -conductor alloys,
was shown by the Post Office Research Station. The and in particular indium antimonide, has been extended
system has been developed for use with electronic ex- by the Services Electronics Laboratory. The electron
changes where the power required for switching is small mobility in this material is 17 times that of germanium
compared with conventional electromagnetic relays. and the Hall effect, for a given transverse magnetic field,
The energy stored in the terminal capacitors of filters gives an increase of available power at the side elec-
at each end while the line is open is transmitted in trodes, proportional to the square of the mobility, of
pulses when the line is closed by applying a 3 -µsec about 300. With Mumetal rods used as flux concen-
pulse at 100 -µsec intervals to a biased diode switching trators the device makes either a sensitive electrical
system. If there is no dissipation in the filter com- compass or a " gradiometer " for detecting anomalies
ponents all the energy is concentrated in the pulses, in the earth's field in magnetic survey work. A devia-
and in practice useful distances can be covered without tion of one degree from the null position, at right angles
any amplification. to the field, gives a power output of 0.002 µW, which
in- will operate a galvanometer or sensitive relay without
Oscillators.- Another Post Office development of amplification.
terest is a 1,000 -c /s tone generator of compact design
ELECTRONICS
Computing. -Decca Radar have now entered
this field with a digital computer which is not-
able for its small size, reduced power consump-
tion and heat dissipation and low cost. These
features are the direct result of using two -state
magnetic cores instead of valves, not only for
storage purposes but throughout the arithmetic
circuits as well. Valves are used merely for
generating pulses for driving the core circuits.
The equipment is constructed on the "package "
principle, with the 2 -mm diameter cores
mounted on printed circuit cards which plug
into racks. The main store is a magnetic drum
and this rotates at a speed (6,000 r.p.m.) which
Portable 1,000 -c!s tone generator with gapped quartz ring vibrator makes its digit rate three times that of the com-
(G.P.O.) puting circuits, so that the normally long access
271
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 time is reduced to a third. For scientific work the input Machine Tool Control. -A system for setting the
and output medium is normally punched paper tape, co- ordinates of a machine tool work table was demon-
but for business applications magnetic tape equipment strated by B.T. -H. The scale for each motion consists
is available as well and this can also serve as long -term of a composite bar with magnetic and non -magnetic
storage for reference data. segments with interfaces spaced exactly 1 inch apart.
Representative of electronic analogue computing was As the table moves the bar varies the flux in a differ-
a "real-time " simulator for the study of control ential pickup head, any difference in the two paths being
systems, shown by Elliott Brothers. Designed on the amplified and used to control the servo motor driving
unit construction principle, it was actually built up the lead screw. When the flux is balanced a move-
from three of the Elliott general- purpose machines ment of 0.00002in causes a detectable error signal.
described last year but had additional apparatus includ- Intermediate dimensions are registered by a subsidiary
ing a non -linear function generator and an electronic micrometer screw servo drive which alters the setting
multiplier of the crossed-fields type. of the pickup head.
Storage Systems.-A new type of two-state storage
device based on the non -linear dielectrical properties Ultrasonics.- Specialized echo sounding equipment for
of barium titanate was demonstrated on the Plessey locating the bearing and range of whales has been
stand. This material has an electric field-strength/ developed by Kelvin Hughes. The horizontal scanning
flux -density characteristic which takes the form of a beam can be varied in vertical width and alternative
square hysteresis loop, giving two states of remanent frequencies and pulse lengths are provided to give
induced charge. The storage cell itself consists of a optimum conditions of detection from 1,800 metres
single crystal of barium titanate with a small matrix down to the firing range, Both aural and visual indi-
system of electrodes on either side of it (X co- ordinate cation of the returning echoes is available.
electrodes on one side and Y on the other), and applica- Ultrasonic abrasion (drilling) of hard materials has
tion of a voltage pulse (say 40 V) to one X electrode been carried a stage further by the development by
and one Y electrode induces a charge at the point of Mullard and Plessey of small portable drills. The
intersection. To " read out " the stored information a Plessey drill uses a new piezoelectric ceramic material
higher voltage pulse (100 V) is applied. If this is of (" Casonic III ") with enhanced power-handling
the same polarity as the induced charge it produces no capacity and frequency stability.
output, but if it is of opposite polarity the reversal Equipment for the ultrasonic cleaning of small parts
causes an output pulse to appear on the appropriate on a laboratory scale has been developed by Dawe
co-ordinate electrodes. The cell has the advantage Instruments. To ensure effective cavitation in the
over magnetic two -state stores of smaller size and lower cleaning liquid at the surface of the objects, a pulsed
power consumption, but requires higher voltages. A oscillator is employed which also operates with less
"read -out" speed of 0.2 µsec can be obtained. power.
The use of ferrite magnetic cores for binary storage An ultrasonic thickness gauge designed to work with
devices, using the square hysteresis characteristic, is a 50 -ft connecting cable has also been introduced by
now well known, and examples of matrix stores based Dawe. It is designed for us in shipyards where it is
on this principle were shown by both Plessey and Mul- often inconvenient or hazardous to take the main equip-
lard. The Mullard magnetic matrix is particularly inter- ment close to the part of the hull to be examined. The
esting for the use of printed circuit frames for equipment works on the resonance principle and
terminating the threaded wires, giving a neat and simple the frequency is adjusted until the wavelength in the
form of construction. material is a function of the thickness.
Magnetic tape is one of the most obvious media for The calibration of barium titanate accelerometers at
storing information but has not hitherto been used a high frequencies and small amplitudes (of the order
great deal in this country. However, Ferranti were of 10-` inch) calls for special methods of measuring the
showing a new magnetic tape handling equipment for amplitude, and E.M.I. Electronics demonstrated an
computers, and this was notable for its high tape optical interference principle which has been found
speed of 100 inches per second and its ability to start useful in this application. An interferometer of the
and stop within 10 milliseconds. Four channels can Michelson type is mounted above the vibration table
be accommodated across the width of the *-inch tape to which one of the mirrors is fixed. With monochromatic
and in each channel the digit handling rate is 10,000 light interference fringes are visible in the interfero-
per second. meter eyepiece, and as the amplitude is slowly in-
Transistor Instruments.-The use of transistors in creased these alternately disappear and reappear. Each
oscillator circuits to give small and compact e.h.t. sup- disappearance is sharply defined and is a subjective
plies was exemplified by two instruments shown by the phenomenon based on the time average of the instan-
Atomic Weapons Research Establishment -a gamma
radiation monitor and a charging unit for quartz-fibre
taneous brightness. This can be calculated, and related
to the amplitude and wavelength of the light.
radiation dosimeters. Both are powered by two 1.35 -V Photographic. -In photographic printing the range of
cells and are small enough to be carried in the pocket. densities in the negative is often too great for the
The portability offered by transistors is also of great printing paper to reproduce all the detail in the pic-
value in physiological amplifiers, enabling them to be ture. To overcome this Cinema- Television have intro-
brought nearer to the patient to avoid pick-up of hum duced an electronic contact printer which smooths out
and other interference. Here, examples were shown such large variations by a feedback principle. The
by Edison Swan and Guy's Hospital Medical School. negative, with printing paper behind, is scanned by a
In the field of time measurement, B.T. -H. had a spot of light from a c.r. tube in television fashion,
transistor timebase calibrator which can measure inter- with overlapping of lines to give complete and even
vals from 10 sec to 2,000 sec with an accuracy of illumination. A photocell then picks up the transmitted
0.5 µsec, while Mullard demonstrated a transistor light on the distant side of the printing paper and gives
counter-chronometer which can also be used for fre- an output signal which varies with the density of the
quency measurement. In the last -mentioned instrument negative. This signal is amplified and fed back to
a quartz -crystal oscillator and transistor frequency control the brightness of the c.r. tube scanning spot
dividers produce reference timing pulses with accurately so that it decreases in intensity when the less dense
known repetition rates from 100,009 to 1 per second. areas of the negative are scanned. In this way the
A time interval to be measured (between two input effects of the density variations can be reduced to any
pulses) is then determined by counting the number of desired extent by controlling the degree of feedback.
reference pulses which occur during the period. Alter- Scanning of a kind is also used in a new method
natively, frequency measurement is performed by of high -speed electronic photography developed by
counting pulses derived from the input waveform during Mullard. This gives a sequence of very short exposure
the period of 1 second. pictures and uses a device similar to an image con-
272
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
verter called an image dissector tube. The photo-
cathode of the tube, which is exposed to the pheno-
menon to be photographed, is active only in patterned
areas, and the electrons from these are swept across
the fluorescent screen by a deflection system. This
results in a composite picture on the screen which is
recorded by a camera. The individual pictures from
the composite record can be sorted out afterwards by
viewing through a patterned transparency or by means
of another dissector tube. Recording speeds of up to
10 pictures per millimicrosecond are possible.
Displacement. -A convenient method of measuring dis-
placement has been evolved in the Mechanical Engineer-
ing Research Laboratory of D.S.I.R. A Ferroxcube
core attached to the moving part (e.g., a valve tappet)
is arranged simultaneously to increase and decrease
the inductance of a pair of coils forming two arms of
an inductance bridge. The other pair of arms is
formed by a similar transducer in which the core is
adjustable by a micrometer. The bridge is excited at
10 kc /s and any out-of- balance component is detected
by a one -cycle response demodulator* and indicated
on a centre -zero meter. Alternatively, the balance point " using
can be used to brighten the trace on a c.r.t. display Ferranti transistor digital computing "package
and the amplitude measured at any point by reference printed circuit.
to the micrometer reading.
Volume.-The principle of the Helmholtz resonator,
familiar in connection with loudspeaker cabinets, is used
in a " Volumometer " which was shown by the research
department of the Morgan Crucible Company. The
change in resonant frequency resulting from the intro-
duction of a body into the resonator chamber depends
only on its volume and the reduction of compliance of
the remaining enclosed air, and is independent of the
shape.
Pressure. -For investigations and routine testing of
hydraulic transmission systems over a range of pressures
from 25 to 50,000 lb /in' Cossor Instruments have in-
* E. A. Johnson. Selected Government Research Reports,
Vol.
5, Report 9, pp. 128 -133. (H.M. Stationery Office.)

Cossor" Hydraudyne "

pressure testing oscillo-
scope.

Below :-Measuring bar and

electro- magnetic head for co-
ordinate setting of machine
tools (B. T. -H.).

Mullard transistor counter -chronometer.

Apparatus for calibrating

accelerometers at high
frequencies by observation
of optical interference
(E.M.I. Electronics).

Plessey matrix store

using a single barium
titanate crystal

27.?
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 troduced the " Hydraudyne," which comprises a pres- voltages are required. They are also superior to metals
sure transducer, amplifier and c.r. oscilloscope with a such as nickel in having low conductivity and requiring
long-persistence screen. The timebase has a scanning no lamination to reduce eddy currents. Plessey as well
velocity variable from 0.001 to 10 sec and the amplifier as Mullard are actively engaged in developing these
response has a rise time of 10 µsec. materials.
The radial pressure built up on textile thread bobbins
by successive layers can be measured by a transducer Magnetic Tape. -As an alternative to magnetic iron
developed by the Rayon Research Association and made oxide, pure iron in powdered form is being used as a
by H. Tinsley (Type 30 -A). A wire resistance strain coating medium in ;in -wide tapes made by Salford
gauge records the transverse distension of a pad of Electrical Instruments. When the particle size is
resilient material which is inserted under the layers of sufficiently reduced pure iron (in bulk a " soft " mag-
thread. netic material) acquires " hard " magnetic properties
Flow. -A flowmeter (Type ND.31) for conducting and exhibits a high remanence. The tapes shown
liquids, developed by Elliott Brothers, works on an in- have a sensitivity 6dB below normal oxide tapes but
version of the electromagnetic principle used for pump- 15 dB higher maximum output and, it is claimed, lower
ing liquid metals. An alternating magnetic field harmonic distortion. They can be made in a wider
traverses the flow tube at right angles to a pair of range of coercivities and they cause less abrasion of the
diametrical probe electrodes, and any movement of the recording heads than oxide types.
liquid induces an e.m.f. which can be amplified and Laminations. -The production of awkward shapes or
applied to a meter, recorder or servo controller. The small quantities for which the cost of press tools would
indication is independent of conductivity over a wide be prohibitive is facilitated by an etching process which
range (tap water was used in the demonstration) and has been developed by the Telegraph Construction and
an accuracy of ± 1 per cent of full scale is claimed Maintenance Company. The process is similar to that
over the range 0 -400 cm' /minute. Originally, the used for printed circuits and can be used for laminations
development was carried out in conjunction with up to 0.004in thick.
A.E.R.E., Harwell, for measuring rates of flow of radio- Miscellaneous Exhibits.- Sub -miniaturization continues
active nitric acid. to be one of the main trends in component develop-
ment, and most of the newest types exhibit this feature.
Plessey have extended their range of " Castanet " tan-
MATERIALS talum electrolytic capacitors and added some new models
Microwave Absorption. -Sheet material (AF 11) con- primarily for transistor applications. These are housed
sisting of prepared foam rubber with absorption between in metal cases only fin in diameter and under fin long.
A 45 -SF, 10 -V type, for example, is long.
99 per cent and 99.7 per cent in the range 5 to 35 kMc
has been introduced by Plessey. It is available /s in
The same trend was observed in the new Type SD
wirewound potentiometer shown by Salford. It is a
panels lft square and is suitable for lining laboratories semi-precision type in a metal case, yet measures only
where field measurements are to be made at these lin in diameter and fin deep, and at present is available
frequencies. Another range (" M " type) depends on in resistance values of from 2 kf2 to 8 kf2. The rating
an interference principle in which the energy reflected is 0.5 W.
from a metal backing is absorbed in facing material of Two very small moving -coil relays were seen this
critical magnetic and dielectric properties and high year; one made by Elliott is argon -filled and hermetic-
refractive index; the total thickness is considerably less ally sealed and plugs into a B9A valveholder. It
than the free -space wavelength of the radiation. operates with only 10 mW input and its contacts will
Magnetostrictive Ferrites. -The use of these materials handle up to 2 W. The other is housed in a small
as an alternative to electrostrictive ceramics such as rectangular case with flying leads and is made by
barium titanate is based on the fact that lower operating Electro Methods. It, also, operates with only 10 mW

Above :Miniature " cable -layng '

machine shown by Fortiphone.

Electromagnetic flowmeter (Elliott

Brothers). Left : Salford sub-miniature pre-
cision potentiometers, Type SD.

274
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 of it resembles the early balanced- armature loudspeaker
input and handles 2 W. Despite the sensitivity movements, but is very much smaller and to far more
these units they are said to be particularly robust and sensitive. The frequency response is linear
within
resistant to shock and acceleration. with the
±3 dB from 200 to 3,500 c /s. A similar unit,used
An aid to the further miniaturization of mobile v.h.f. frequency range extended to 4,000 c /s, is as a
radio- telephone equipment is the extension of the over- transmitter (microphone).
tone operation of quartz crystals into the regions of 100 Among the exhibits of Fortiphone was a miniature
and 200 Mc/s. Crystals operating on their 7th overtone (for its type) " cable -laying" machine. Its function
is
(180 Mc /s) were shown by Standard Telephones and using very
some 5th -overtone models (up to 75 Mc /s) by Salford. to produce a twisted two -wire insulated cable strands of
Apart from permitting a reduction in size of equipments fine wires. In one form it " lays " two enamelled
these crystals lead to a worth-while economy in opera- 0.0036in diameter (43 s.w.g.) heat -stripping
0.005in dia-
copper wire with, initially, two strands ofoverlay
tion as in many cases no frequency multiplying stages meter nylon monofilament and a final of 7
are required. Reliable and economical crystal control strands of nylon. A plastic adhesive is automatically
of f.m. receivers becomes a possibility. applied and a small oven dries it off before reeling.
A new telephone receiver (earpiece) was shown by created
Standard Telephones in which higher sensitivity (about Activities in the ultra- miniaturization field haveunobtain-
the a demand for such a cable, which was hitherto
5 dB above the average) is achieved by separating by
magnetic and acoustic functions. The former is per- able anywhere- A four -way cable is provided
formed by a rocking armature and the latter by a light- cementing together two twin cables. The breakdown
voltage is around 5,000.
weight metal cone-shaped diaphragm. In some respects

TEST AND MEASURING GEAR

New Exhibits at Recent Shows

One design trend should be mentioned at the start,

This report embraces instruments shown at
the
because it is not confined to any particular kind of R.E.C.M.F., Physical Society and 2nd International
instruments-the use of transistors. An increasing
number of indicators employ them in place of valves instrument exhibitions without distinction. The
for amplification, with great saving in space and power last -named show introduces some unfamiliar
consumption. A number of transistor oscillators also marques. As in previous reports, many of the
appeared; crystal controlled and otherwise. Transistors instruments mentioned were shown in prototype
are particularly appropriate for bridge oscillators and
in- form and are subject to modification before they
null indicators. Other transistorized instruments Those previously
cluded a distortion -factor meter, a frequency meter, become available, If they do.
a time -base calibrator and a complete oscilloscope. reported are not mentioned again, even if now
Circuit printing has not yet made much headway the available for the first time, unless the modifications
among measuring instruments, presumably because Avo are substantial.
production quantities are seldom large, but the -level
AM /FM signal generator and the Dawe sound for
meter have made a start with this technique, which
instrument work has an obvious advantage in reducing
manufacturing variations.
Right: Grayonics Developments in unamplified meters were less notice-"
millivoltmeter, Type able this year than last, but to the Pye " Scalamp en-
1277. series of electrostatic voltmeters has been added anused
larged model for measuring the higher tensions level the
for cathode -ray tubes, up to 40 kV. At this attention.
prevention of brush discharge calls for special to 32 V
The Labgear r.f. millivoltmeter, covering 20 mV
in six ranges, at frequencies from 50 kc /s to 250 Mc /s,
might easily be mistaken for a valve -aided instrument,
rectifier
but actually employs only a germanium crystal
operated on the slide -back principle, and its power
Below : Pye "Scalamp "
needs are confined to a small bias battery. from the
electrostatic voltmeter. The evolution of indicating instruments especially
original valve voltmeter continues however; and either
in the two directions of higher sensitivity
wider or narrower frequency bands. The Grayonics voltage
millivoltmeter has no fewer than 12 linearover the
ranges, from 1 mV to 300 V full- scale,
frequency range 20 c/s to 2 Mc /s. Its meter can be
adjusted to the most convenient angle for reading. are
For audio frequencies much higher sensitivities
possible; there is the B at K (Bruel and Kjaer; Den-
mark) Model 2408 with full-scale readings from 31.6 µV
to 1,000V over the range 20 c/s to 20 kc /s. Although
the lowest f.s.r. on the Peekel (Holland) Model 051B
is higher -10 mV -this instrument is notable for
its
0.15 c /s,
frequency range going down from 10 kc /s to available.
for which pointer instruments are not usually Even
Three steps of pointer damping are provided. same firm
this is not quite zero frequency; for it the µV
lists Model 30B, with 10 ranges from 30 per tohour. 1V
f.s.r. Zero drift is claimed as less than 2 AV
275
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 Frequency control gear in Sanaers
XT312 signal generator.

New London transistorized

voh- ohmmeter.
Marconi Instruments alignment
oscilloscope, Type TF1104.

tuning to 1,025 c /s. A smaller instrument by the same

firm uses 0071 junction transistors to give a sensitivity
of 20 µV f.s.r. at 3,200 c /s, or slightly better on a flat
range from 100 c/s to 10 kc /s. The B & K Model 2002
employs the superheterodyne principle to provide
ranges down to 15 f.s.r. from 20 c/s to 30 Mc /s,
the i.f. being 1,650 kc /s. A very different B & K selec-
tive indicator is their Model 2105, for a.f. analysis over
47 c/s to 12.5 kc /s in eight bands and with four steps
of selectivity in a degenerative RC amplifier. Another
B & K instrument that can be used for frequency
analysis is their filter set comprising 27 one -third octave
filters covering in all 40 c/s to 16 kc/s. A useful filter
is the Krohn -Mite (U.S.A.) Model 310AB, a RC net-
work with amplification, having the unusual facility of
Dawe Type 442 v.l.f. oscillator. separate adjustment of low and high cut-off frequencies
over the range 20 c/s to 200 kc /s; input impedance 6 M12
in parallel with 50 pF, and output 50012.
The B & K Model 2423 megohmmeter, 0.1 M12 to The new oscilloscopes have been designed for im-
10 MM12, can also be used as a voltmeter or a micro - proved refinement of control or facilities, such as width
microammeter. Elliott make a micro -microammeter of frequency band or ability to select small portions
(which presumably can be used as a megohmmeter) of complicated waveforms. Signal delay, so
covering 10 µµA to 1 f.s.r. in six ranges, using a
ME1401 electrometer valve in a sealed and desiccated
repetitive waveforms can be observed in full,that non-
is now
sub -unit. Further developments in the Pye galvano- quite frequently provided, even in general-purpose
meter-modulator system of sensitive z.f. amplification models such as the Mullard L140 precision type. The
enable a large number of low voltage and current ranges dual trace L101 oscilloscope now appears with certain
to be obtained. improvements as a Mark 2; for example, time -base
The New London (U.S.A.) transistorized volt - triggering from either positive or negative wave fronts,
ohmmeter has d.c. and a.c. (10 -5,000 c /s) ranges with only and continuous sweep expansion. The latter facility,
5 µA full-scale load; voltages 0.03 to 1,000V f.s.r. in 10
over a 10 to 1 range, is included in the Solartron AD557
ranges, and resistances 1, 10, 100, 1,000 and 10,00012 with d.c. amplication for both X and Y. Nagard, who
mid-scale. specialize in wide -band oscilloscopes, showed their
DE103 with an improved specification, including signal
Versions of an a.f. wattmeter of N.P.L. design are delay line and frequency range from 0 to 15 Mc /s.
offered by Cambridge Instrument and by Tinsley; the Their R103 is displaced by a new design in which good
numerous ranges down to 2.5 mW full-scale on a con- use is made of a 20th Century precision c.r. tube with
ventional dynamometer instrument are made possible the high sensitivity of 14 V /cm.
by negative -feedback amplifiers for both current and In contrast to these is the Cintel " Synchroscope"
voltage coils. Another instrument, in which an ampli- without amplification (sensitivity, 60 V /cm) and thereby
fier is used to bring noise voltages up to the level at able to respond to frequencies up to 1,000 Mc /s ( -3 dB
which a mean -square detector can be efficiently oper- point).
ated, is the Wayne-Kerr video noise -level meter. Its Although described as an oscilloscope, the Marconi
bandwidth is flat from 10 kc /s to 1.5, 3, 6 or 10 Mc /s Instruments TF 1104 is much more, since it includes
according to the setting of a switch. a signal generator covering Bands I, II and III and
Frequency- selective amplification is used mainly for appropriate v.f. and i.f. ranges, and crystal- calibrated
two purposes; to enable a sharp minimum to be marker pips; in fact, all required for viewing television
obtained in bridge work; and for frequency analysis. and v.h.f. frequency characteristics.
The Microwave Instruments Type 3100, which would To demonstrate the present possibilities of transistors
be very effective as a null indicator, was actually in this field Mullard showed an experimental oscillo-
developed primarily for use with crystal detectors in scope providing all the usual facilities of the simpler
microwave gear, and achieves the exceptional sensitivity kinds of instrument (see block diagram), with no
of 1µV f.s.r. clear of noise by the sharpness of its thermionic valves and a total power input of 6 watts at
276
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 SAWTOOTN SAW TOOTH
D.C.
VOLTAGE AMPLIFIER
CONVERTOR GENERATOR

VOLTAGE
12 V D.C.
STABILIZER

D.C. SYNC. "BRIGHT -UP"

CONVERTOR PULSE PULSE
GENERATOR AMPLIFIER

12V
POWER SUPPLIES AMPLIFIER

Block diagram of Mullard experimental oscilloscope. SIGNAL ATTENUATOR

IMPUT

12 V. Another interesting transistor unit is a B.T. -H. dards, but also phase angle against a reference voltage.
time -base calibrator. It generates an output super- The advantages of transistors for bridge oscillators
imposing a time scale on any oscilloscope screen, with and balance indicators were shown in a number of
small divisions at 10 µs intervals, double-size at 50 µs, exhibits by Tinsley.
and triple at 100µs, the whole being controlled by a Not only for ordinary oscillators but also (in con-
100 -kc /s, ± 0.01 % crystal. junction with quartz crystals) as frequency standards
A delay circuit enables the whole width of the screen are transistors being applied. A good example is the
to be used for observing any small section down to Labgear oscillator, giving a single standard frequency
JO µs of a waveform up to 2,000µs in duration and
up to 10 Mc /s, with point- contact transistor, complete
measuring it to within 0.5 µs. Other accessory equip- with battery in a case 3áin x 2in x l in overall.
A somewhat similar " potted " calibrator by Elliott
ment includes the Mullard television line selector for is stable to better than 1.5 in 10' per °C. With a
displaying any individual line on an oscilloscope, the 100 kc /s crystal, useful harmonics are available up to
Furzehill two -way beam switch for converting a single - 30 Mc /s. For low frequency standards, Elliott have
trace oscilloscope to a dual -trace type, and the Cossor developed transistor -maintained tuning forks. The
monitor with separate c.r.t. and brightness and focus Furzehill G410A frequency standard uses a 5 -Mc /s
controls for duplicating the trace of an oscilloscope to crystal with valve multivibrator dividers to provide
enable one to be observed while the other is photo- pulses at 5 Mc /s, 1 Mc /s, 100 kc /s, 10 kc /s, 1 kc /s and
graphed. 100 c /s, modulated if desired at 400 c /s, and selected
A new attenuator was shown by Advance Com- by push buttons. A Labgear prototype standard con-
ponents: the A64 for a.f. and low r.f., giving a range tains an oven -stabilized 200 -kc /s crystal and a t.r.f. radio
of 0 to 70 dB in steps of 1 dB at 600û. A series of receiver tuned to the B.B.C. 200 -kc /s carrier, enabling
fixed attenuators by Hatfield Instruments, consisting of the source of frequency to be known within about 1
disc and rod resistors arranged in coaxial " T " forma- in 10'. Finally, a version of the famous G.P.O. fre-
tion have the exceptionally wide frequency rangeline of
quency standard with a daily stability of better than
0 -1,000 Mc /s; the standing wave ratio in a 72 -f2
5 in 10'° was to be seen on the Airmec stand.
at the highest frequency is given as 1.1, and the in- Coming now to signal generators and considering
accuracy of attenuation ±0.3 dB. One might have them in ascending order of frequency we have first the
thought that the now well -known Sullivan decade air Dawe 442 v.l.f. oscillator, 0.1 c/s to 10 kc /s, notable
capacitor would have satisfied all requirements as for the placing of the main heat-producing components
regards setting accuracy, but the very small residual though pro-
-the valves -more or less in the open air,feature
errors are now corrected on the 100 pF steps bysame an is the
tected by recessing. Another interesting
auxiliary set of curiously shaped sectors. Theinduct- provision of two outputs in quadrature. The Wayne -
firm showed an example of their new mutual- Kerr wide -range a.f. oscillator now appears at a further
ance standards, temperature- compensated in a similar stage of development, the most interesting feature being
manner to the Sullivan -Griffiths self-inductance stan- its frequency and amplitude controls with scales
dards. Decade inductance boxes are less commonly arranged for fool -proof direct reading; the frequency
available than resistance or even capacitance boxes, but control combines the advantage of continuous variation
Muller- Barbieri (Switzerland) showed one of three and main -point working. The instrument meets
models, giving 1 mH to 10 H in 1 mH steps, each step Ministry requirements with a clearly visible mains -
being a toroidal coil on a molybdenum -permalloy voltage adjustment mounted on the front panel under
powder core. a transparent slip. The Salford 50 c/s to 50 kc /s signal
New bridges were not much in evidence; one of the generator is unusual in being provided with crystal
exceptions was the Avo Universal Measuring Bridge, all calibration, checking by which is facilitated by scale
with 24 ranges of R, C and L, covering almostThe markings. For use with their Acoustic Calibrator, Dawe
component values except inductances below 1 mH. have produced a compact transistor oscillator to pro-
source is an internal 1 -kc /s oscillator for C and L, vide a signal at 400 and 1,000 c /s, but it can also be
and d.c. for R; and the indicator is a meter. Capacitor used as a bridge source, etc.
leakage can also be measured; appropriate test voltages The Muirhead D783 Square -Wave Shaper, for use
are provided. A Pye inductance bridge is interesting in conjunction with sine -wave sources in the range
for its direct -reading scale, with thumb- operated main 20 c/s to 300 kc /s (or with 50 c/s from its own power
and fine controls strongly resembling a well -known type supply for that frequency) gives an output with a rise
of radio tuning control. The indicator is a miniature time less than 0.08 µs. This parameter is of even
c.r.t., which facilitates phase and amplitude balance. greater importance in pulse generators, of which the
Range 0.1 mho to 10 mhos with accuracy at worst Nagard Type 5001 is a new example. The recurrence
± 0.2% of reading. A deviation test bridge
shown by
B and K has provision for interchangeable meter
frequency of the internal oscillator can be varied from
scales marked with appropriate limits. Not only can 0.1 c/s to 1 Mets, and pulse width and delay from
production components be rapidly tested against stan- pre -pulse are variable from 0.2 µs to 2s.
277
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 In new r.f. signal generators the accent is on v.h.f. geared to a reflector voltage control. A wide range of
The Avo AM /FM model already mentioned is a modi- microwave equipment shown by Sivers Lab (Sweden)
fied production form of the prototype seen last year, included a signal generator which, like all their variable -
and covers5 to 220 Mc /s in eight ranges with a.m. and frequency equipment, was provided with a cyclometer
65 to 120 Mc /s with f.m. Special arrangements are type of frequency indicator direct reading in Mc /s. For
included for increasing the accuracy of the scale close measuring noise factor of X -band receivers, the Marconi
to any desired frequency. The Airmec Type 204 cover- TF1070 noise source consists of a discharge tube mounted
ing 2 to 320 Mc /s, is notable for having nine different in the E -plane of a length of wave guide, generating a
modulation facilities, including a.m. or f.m. or both to- standard noise signal 15.5 dB above thermal noise.
gether, internal or external, or pulse modulation. Two Other microwave gear by Sanders included a No. 16
v.h.f. signal generators were shown by Hatfield: the waveguide test bench (8.2 to 12.5 kMc /s), a calibration
LE120C covering 3 to 300 Mc/s; and the LE250B for receiver with first detector crystal mounted on a three -
3 to 20 Mc/s and 81 to 105 Me/s intended for pro- stage low-noise head amplifier, and a high -power dummy
duction testing of Band II f.m. receivers. Modulation load. B.T. -H. equipment included an X-band automatic
is by either sine or sawtooth waveform, the latter being s.w.r. and impedance measuring equipment in which a
intended for display of selectivity and discriminator Smith chart display is given on a c.r.t. screen, an X -band
curves. A special feature of the Marconi TF1066 dielectric test set for relative permittivity and tan ò,
FM /AM signal generator is an incremental tuning and a Q -band " Enthrakometer " for measuring power
system in which small frequencies are read from a meter at 35 kMc /s by the change in resistance with tempera-
with a direct calibration valid at all carrier frequencies. ture of a very thin platinum film suspended across the
The lower microwave frequencies are covered by incoming waveguide. Another solution of the power -
another Marconi signal generator, the TF1058, built measuring problem, applicable to all frequency bands,
around a klystron oscillator of a new type in which the is the E.M.I. Precision Microwave Wattmeter Type 1,
length of a coaxial line and the reflector voltage are which operates in a novel manner, making use of what
simultaneously varied by the frequency control to enable is in effect an a.f. phase -shift oscillator amplitude -
the unusually wide range of 1.7 to 4 kMc /s to be covered stabilized in the usual way by a thermistor. In this
in one band. A similar technique is adopted in the case the thermistor is used as the r.f. termination, and
Sanders XT312 and XT314 signal generators which be- the rise in temperature due to the power being measured
tween them cover 1.5 to 12 kMc /s. Again, the direct- is automatically compensated by a change in a.f. cur-
reading wide -band frequency control is of particular rent through it; this current is therefore a measure of
interest; linearity and accuracy of the scale is achieved the r.f. power dissipated in the thermistor, which is
by an adjustable cam controlling the coaxial piston and (Continued on page 279)

Taylor Type 94A television alignment

generator.

-
Right: Cintel Signal Tracer."

Airmec " Radivet " receiver

tester.

Left:-Microwave equipment made by Sivers Lab.

278 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
kept at practically constant resistance. Favourable in coils, and a power unit containing the indicating meter,
accuracy and ease of control are claimed. The power working on the " grid -dip " principle. For recording
range is 0 -4 mW in steps of 0.1 mW. A thermistor is audio frequency characteristics, B & K showed equip-
also the basis of a method by Microwave Instruments ment consisting of a manual or motor driven beat -
in which it forms one arm of a Wheatstone bridge, frequency oscillator in conjunction with an output
calibrated by d.c. substitution. recorder with a selection of 10 paper chart speeds from
The highest -frequency microwave equipment seen was 0.003 to 100 mm /s.
an experimental test bench for 75 kMc /s (4 mm wave- Production test equipment included a shorted -turn
length). When it is considered that at this frequency detector by Nash and Thompson capable of indicating
a half -wave dipole aerial would be only 2 mm long some one shorted turn in a coil of 40 s.w.g. wound to a radial
of the instrumental difficulties can be imagined ! One depth of ;in; continuity of the coil is also shown. The
of them is the production of such waves; actually an object is to detect faults in transformer coils before
8 mm klystron was used, with a silicon crystal to generate assembling the cores. For automatically inspecting
the second harmonic. Very close tolerances, of the order strength of permanent magnets at high speed, B.S.A.
of 0.0001in, are necessary in the " plumbing." The usual showed test gear working on the principle of upsetting
slotted -line type of s.w.r. indicator being impracticable, a balance by saturation of one of two small mumetal rods
a rotary indicator is used, the phase of the standing wound with coils forming twin arms of a bridge. A relay
wave being measured directly as a physical angle. operates a rejection mechanism for magnets below stan-
Another fascinating microwave exhibit was the Polarad dard. A B.T. -H. production tester for point -contact
spectrum analyser, which was shown displaying the germanium diodes displays on a c.r.o. the forward and
spectrum of a pulse- modulated signal in the 0.91 to reverse current /voltage characteristics, simultaneously
4.56 kMc /s band. This consists of a series of vertical but with appropriately different scales; and by applying
lines (carrier and sidebands) spaced at frequency inter- a square wave to the diode under test enables recovery
vals equal to the pulse recurrence frequency and having time (hole-storage effect) also to be assured. The same
an envelope width dependent on the pulse width. firm exhibited a transistorized direct- reading frequency
Lastly, there were exhibits not falling clearly into any meter of the charge -discharge type covering 0.3 to
of the foregoing categories; for example, a selection of 100 kc /s f.s.r., for signals of any waveform and any
the Heath (U.S.A.) instrument kits described by C. B. amplitude between 0.1 and 500V.
Bovill in the October 1955 issue of Wireless World. An A rather unusual instrument, designed for use in a.c.
interesting feature of the valve-voltmeter kit is the use testing of magnetic materials is the form- factor meter by
of an etched circuit. The Taylor 94A Television Wave- Tinsley, having two ranges : 1 to 1.5 and 1 to 3. The
form and Alignment Generator provides an exception- Airmec phase -meter Type 206 enables both phase and
ally comprehensive selection of signals for testing gain to be measured in the frequency range 20 c/s to
television, f.m. and short-wave receivers, including a 100 kc /s. For amplitude modulation in signal generators
number of television patterns -Continental and Ameri- and similar instruments, the Hatfield balanced modulator
can as well as British standards- with synchronizing, Type LE90A is claimed to be completely free from
interlacing, etc., signals, c.w. with and without a.m. or frequency modulation. It makes use of the firm's wide-
f.m. at various output levels, a wobbulated signal for band r.f. transformers in conjunction with a balanced pair
alignment in conjunction with an oscilloscope, and a of germanium diodes, and covers frequency ranges of
variable- output a.f. signal. The r.f. frequency coverage 0 to over 100 Mc /s for the modulating signal and 3 to
is 8 to 230 Mc /s. As a complement to the well-known over 400 Mc /s for the carrier. Finally, one of the largest
Airmec " Televet " there now appears the receiver tester instruments, the Ellott " A -PAP " microwave aerial near -
Type 211 or "Radivet," giving a r.f. signal in the field phase and amplitude plotter. In contrast to radiation -
frequency ranges 0.1 to 15 Mc /s and 85 to 100 Mc /s field polar- diagram equipments, which require a fair
with a.m. or f.m. or both, and an a.f. signal 100 c/s to amount of space and precautions against reflections, and
15 kc/s. A crystal calibrator is incorporated for accurate which do not necessarily indicate the cause of anv
frequency checking, and an oscilloscape for wobbulator undesired feature in the radiation pattern, this
display or general purposes. A considerable amount of measures the field close up across the front of the aerial
signal and circuit testing can be performed by the Cintel being tested. It consists of a r.f. source to feed the aerial,
Signal Tracer, consisting of a probe unit containing a and a pick-up automatically traversed, the output ampli-
calibrated oscillator covering 0.1 to 100 Mc /s with plug- tude and phase of which are recorded on a chart.

NEW VALVES AND SEMI- CONDUCTORS

"Glassware " and Allied Exhibits at the R.E.C.M.F. and Physical Society's Shows

Special and Transmitting Types. -The use of ceramic to as high as 900Mc /s and at this frequency will give
instead of glass for valve envelopes is not new, but a useful c.w. power output of 600 watts. English
appears now to be on the verge of greater development. Electric were also showing a new magnetron, M541,
It gives greater mechanical strength, smaller size for a designed for pulse operation at about 1,200 Mc /s. It is
given power dissipation, enables the valves to work at mechanically tunable over a 10% frequency band and
higher ambient temperatures and permits more effective will give a pulse power output of 0.5 megawatt.
de- gassing during manufacture so that greater emission The backward-wave oscillator is one of the latest
current can be obtained under pulse conditions. A types of velocity -modulation valves and is comparable
series of ceramic valves shown by Ferranti included with the ordinary travelling wave tube except that the
triodes for oscillators and amplifiers, with anode dis- r.f. field energy travels in the opposite direction to the
sipations ranging from 15 to 100 watts; low -g triodes electron beam. A notable feature is the wide frequency
for use in stabilized power packs, and indirectly heated variation which can be obtained by altering the beam
half -wave rectifiers. The maximum operating tempera- accelerating voltage, and in the Mullard valve on show,
tures are in the region 180 ° -250 °C. type MS1203, this amounts to 7,000 Mc /s- 11,500 Mc /s
A somewhat larger ceramic valve was shown by (with 50 mW output). A more familiar velocity -
English Electric. This was a coaxial transmitting modulation oscillator valve is the klystron, and this is
tetrode, CR1101, with an air -cooled anode capable of not normally known as a tunable device. However,
dissipating 2 kW. The valve will take full ratings up E.M.I. Electronics were showing a new reflex klystron,
279
WIRELESS WORLD, TUNE 1956

www.americanradiohistory.com
 20th Century Electronics tube with spiral p.d.a. electrode.

Above :- Mulla:d photo-

transistor compared in size
with a threepenny bit.

English Electric tunable

magnetron.
Left :- G.E.C. r.f. heating
power valve.

E.M.I. Electronics reflex

klystron in tunable tional junction type is the " avalanche " transistor, and
cavity. an experimental model was shown by Mullard. The
action depends on the fact that the collector is operated
at a higher voltage than normal, so that the holes flow-
ing towards it reach high velocities and collide with
germanium atoms nearby, producing additional holes
and electrons. These in turn produce more current
carriers by the same effect, and a sudden multiplication,
or " avalanche," occurs which results in a very rapid
build-up of current through the transistor. Moreover
the high collector voltage causes the hole -depletion layer
to stretch almost all the way through the base to the
emitter and consequently the holes only have to diffuse
a very short distance through the base before being
swept into the collector. The avalanche process is
analogous to the breakdown in a cold-cathode discharge
valve except that the base retains control during the
conduction. The device is particularly suitable for pulse
applications, for it will handle rise times as short as
0.01 µsec.
A power transistor (germanium junction) that can be
Ferranti 15 -W ceramic triode for use up to 2,000 Mc /s. used to switch as much as 5 amps was shown by
B.T. -H., while G.E.C. had a new junction type, EW53,
in which the base is connected to the metal can to give
R5222, which operates with an external cavity resonator lower thermal resistance and improved mechanical
tunable between 8,400 Mc/s and 10,300 Mc/s. The strength. English Electric have now entered the field
power output rises from 28 mW to 55 mW at mid -band with some junction transistors for low -power audio
frequency and falls again to 25 mW. applications, while Brimar have introduced three new
Receiving Types.-In multichannel television recep- types for similar applications, TS1, TS2 and TS3, which
tion the receiver often has to handle signals of widely are hermetically sealed and replace their earlier TJ
different amplitudes, and one unfortunate result of this types.
can be cross -modulation distortion between vision and Rectifiers. -The most important development in this
sound, giving an objectionable buzz on sound. G.E.C. field has been the emergence of the silicon junction
have now produced a new variable -mu pentode (Type diode. This device is characterized by its small size
W729) for television i.f. amplifiers, which minimises for a given current rating, its ability to work at high
this distortion while at the same time maintaining a temperatures and its low reverse current. A typical
high slope and giving a better control characteristic small unit about the size of a '+-watt resistor might give
than conventional variable -mu valves. For a.m. /f.m. a rectified current of 200 mA and have a reverse current
reception this firm also displayed a new triple -diode of 5 µA with a peak inverse voltage of 100V while
triode DH719 /EABC80 with one diode having a operating at a temperature of 100 °C. Examples were
separate cathode, while Mullard had a variable -mu r.f. shown by B.T. -H., Ferranti, G.E.C., and S.T.C. Heavy -
pentode with two diodes (EBF89), which is notable for duty types capable of passing currents of 30 -50 amps at
a low anode -to -grid capacitance (less than 0.002 pF) and 100 -200 volts were also on view and these were still only
reasonably high slope (3.6 mA/V); this makes it pos- about in in diameter.
sible to obtain high gains without instability in the i.f. Germanium junction rectifiers, by comparison, have
stages of a.m. or f.m. receivers. Another new Mullard the advantage of lower forward resistance. As an
valve for a.m. /f.m. reception was the miniature battery example of what can be done with them, G.E.C. were
r.f. pentode DF97, intended for portable sets. showing a water -cooled germanium diode capable of
Transistors. -A new kind of junction transistor which giving a rectified current of 300 am res ! Mullard
overcomes the inherently slow response of the cQnven- had a new germanium junction diode, Type OA10,
280 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
notable for its extremely low hole -storage characteris- Another new material which has been found to have
tics. photo -electric properties is germanium, when used in
A new method of depositing selenium on metal plates junction devices, and an embodiment of this was the
by a vacuum evaporation process has led to metal rec- photo-transistor, type OCP71, shown by Mullard. It
tifiers of somewhat improved performance. The process is very small and will operate from a low voltage
gives a more intimate bond and makes possible a thinner (about 10 V), giving an output current when illuminated
layer of selenium, as a result of which the forward in the region of 5 mA. The " dark " current is not
resistance is lowered and the current rating for a given more than 300 µA.
size of rectifier is increased by about 25%. The life C.R. Tubes. -For television the 21 -inch tube with a
of the rectifier is also said to be improved. 90° deflection angle is the latest thing on the market
A range of selenium rectifiers notable for their ability and two new types were on show -the Brimar C21KM,
to work at ambient temperatures as high as 85 °C was with a tetrode gun, and the G.E.C. 7501A with a triode
displayed on the Westinghouse stand. One type suit- gun. Both have aluminized screens. As a contrast,
able for television receivers, giving 300mA at 270V, Ediswan have reverted to the 9-in screen size in their
has elements contact -cooled at the edges on to the new tube CRM93-although they have also produced
aluminium case and is " potted " in a block of resin of a 24 -inch rectangular type.
about 1 cubic inch. It will operate successfully even Among oscilloscope tubes 20th Century Electronics
when the chassis to which it is fixed is as warm as were showing a new type which has the interesting
60 °C. feature of a spiral post -deflection acceleration electrode
Photo-electric Devices.-An improved version of the instead of the usual series of rings. It is formed by a
conventional " venetian-blind " photo -multiplier tube coating of resistive material on the glass and the accel-
was shown by 20th Century Electronics. In this a care- erating potential (10 kV) is applied across it. This gives
fully designed electrode system gives some degree of a potential gradient which increases evenly towards the
focusing from stage to stage, and so leads to improved screen and so avoids the lens effect which occurs be-
extraction of slow secondary electrons and higher inter - tween separate rings and also the need for a series of
stage efficiency. Increased efficiency was also the separate acceleration voltages. The writing speed of
feature of the new cadmium sulphide photocells exhi- the tube is 1,000cm per microsecond. Another tube
bited by B.T.-H., which are several orders of magnitude with post -deflection acceleration, the Cintel G601 -C4,
more sensitive than selenium cells. Both single -crystal is capable of recording frequencies as high as 1,500
and powder -layer types were shown. Mc /s.

LETT THE E DUO

The Editor does not necessarily endorse the opinions expressed by his correspondents

increase eft, by 0.58 %. My own timer is constructed

" Precision Photographic Timer" on a conventional tinplate chassis, completely enclosed
MAY I reply to your correspondents, Messrs. Hercock, in a plywood box. The temperature difference of the
Neale and Askew (April, 1956)? Changes in contact chassis underside rises on a 50 min. exponential
potential are usually a limiting factor in the design of (approx.) to an asymptotic value of 16 °C (approx.). If
low- frequency apparatus such as the photo timer. The used immediately after switching on, the timer would
input voltage et was chosen to be a compromise be- therefore give intervals 9.2% longer than in the tem-
tween high values giving little or no increase in CR perature- stabilized condition. On this account it appears
and low values where contact potential changes are advisable to allow about 30 min. warming -up time
relatively more significant. In regular use, the grid where it is desired to keep to better than 5% accuracy.
voltage of a valve such as the EF37 will return to within G. A. Askew is quite correct in stating that the Miller
100 mV of its mean value and, with intermittent use, integrator valve can stay on its grid base without an
to within 200 mV, falling to 100 mV in less than an anode load (for negative " run -down " only). In the.
hour. With the suggested value of 4.7 V for ern. 100 mV design of a single valve Miller time -base to work faster
corresponds to only about 2% change in interval. Larger than about 20 V /µsec the feedback capacitor current
changes due to ageing may be allowed for using the and the usual load capacitor current are, of course,
2.5 -K potentiometer (Fig. 4). larger than the anode resistor current. I should have
Changes in the integrator valve internal gain have mentioned that the only reason the 1.5 -M anode resistor
only a small effect on the interval (conventional negative is used in my circuit is to set up the recommended
feedback theory). The effect becoçnes more important conditions for low grid current (cf Mullard ME1400).
as etn is reduced, due to the relatively greater signifi- The recommended reduction of the heater voltage to
cance of the changed grid voltage excursion, but this 4.5 V has been noticed by myself and others to be
is a price paid for the effective increase in CR, as with ineffective and possibly harmful to valve life.
the contact potential changes. J. G. THOMASON.
The steps of 4/2 (or 1.58) are logarithmic, which Malvern, Worcs.
seems more appropriate than arithmetic steps of, say,
15 seconds, and though admittedly rather coarse, were
chosen for two reasons -(a) a large total range had to Single or Double Sideband?
be covered using a standard 11 -way switch, (b) most
modern printing papers (such as Ilford's Plastika) are BEFORE the opening of the new London television
supplied with a leaflet advising that there is no harm station I wrote to the B.B.C. in terms similar to those
in using developing times of -50% to +100% of the used by your correspondent, " Lambda " (April issue).
optimum if desired-this is more than adequate to inter- Nevertheless, the event has proved both of us wrong,
polate between steps of 1.58 in exposure time. although I imagine the B.B.C. were unaware that this
Measurements on an Atmite disc type L275702A would be so. In spite of " Lambda's " curves, I at
show a temperature coefficient of 0.23 ,hA per °C least am able to receive a satisfactory picture using a
measured at 27 V, 40ÁA (cold). Thus a 1 °C rise will receiver turned to the upper sideband. In fact, it is
281
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 possible to resolve the 2.5 -Mc /s bars on Test Card C of developments and techniques, for there are no
(previously I could just see the 3- Mc /s). electronics or radio technical journals published in
I am sot suggesting that the B.B.C. were right in Braille in this country; in fact, the only one published
their decision to use the vestigial sideband system, and anywhere in the world is the Braille Technical Press in
it would obviously not have been possible to consult New York.
the viewing public. It is perhaps fortunate for the You will appreciate that when an article is read aloud
B.B.C. that the thousands of U.S.B. receivers still in by someone not interested in the particular subject, the
use were not rendered obsolete overnight. resultant impression can be confusing and often
F. R. ESTALL. misleading.
South Benfleet, Essex. In the U.S.A. this has been overcome to a large ex-
tent by club members reading articles, together with
their comments, on to tape and sending them to fellow
B.B.C. Reply members for discussion.
May I suggest that some of your readers might like
YOUR correspondent, " Lambda," sets out in the April to participate in such a service over here, and regularly
issue some of the arguments against the change to read items and articles from Wireless World and so help
vestigial sideband transmission that has been made with us to keep up to date, and to overcome to some extent
the opening of the Crystal Palace television station. our handicap.
These points were all considered, but the factors that Also, may I ask for suggestions for an aural modula-
principally influenced the decision were: tion indicator to substitute for the usual visual indicator
1. The desirability of using the same type of trans- on tape recorders?
mission at all B.B.C. television stations; all the CHARLES H. STANDEN.
permanent post -war stations have a vestigial side - London, S.W.9.
band characteristic.
2. The improved power- conversion efficiency and
greater power output that could be achieved in
the transmitting equipment.
3. The improved performance of the transmitter and
Terminology; Warlike -
I WISH to make a complaint. Why must we be fired
aerial system over the band transmitted. at and triggered-off in television circuits? Are we
4. The saving of 2 Mc /s of spectrum space, which required to assume that the authors have all had Army,
may possibly be useful for some other purpose Navy or Air Force training?
in the future. Cannot we leave these matters to the war-lords and
Before the decision was made, tests were carried out, simply speak of the electro-activation pulse.
with the co- operation of the radio industry, on a num- A. HARDWICKE.
ber of commercial receivers designed to favour the Glossop, Derbyshire.
upper sideband. The tests were made with a filter
rather more severe in attenuating the upper sideband
than the actual characteristic of the Crystal Palace -and Misty
transmitter. It was found that the loss of detail in the
picture was much less than would be expected from IT seems to me that the terms " demodulate," " demodu-
Lambda's " theoretical diagram. Even without any lator," are bad ones and ought to be dropped, as their
modification of the receivers the resolution was only meaning is the reverse of that intended. If the wind-
slightly degraded, and the difference was barely notice- screen of my car becomes misted I switch on a demister
able on most of the receivers tested. which removes the mist and leaves the windscreen clear.
The number of receivers favouring the upper side - Yet some radio people, when they have a carrier which
band that are still in use is certainly less than the figure is modulated, use a demodulator to remove the carrier
of 60,000 quoted by " Lambda " from Sir Noel Ash - and leave the modulation. This is surely as absurd as
bridge's paper of 1951 as the total number of sets in saying that a demister is a device for removing the
use at the time to which he referred (shortly after the windscreen and leaving the mist.
reopening of the Alexandra Palace station after the war). P. E. K. DONALDSON.
Nearly all the receivers of this type are at least six Cambridge.
years old, and many considerably older; a number of
them have, in any case, been replaced by sets capable
of receiving the I.T.A. transmissions in Band III, which, Tape Amplifier Design
incidentally, also have a vestigial sideband charac- I HAVE recently constructed tape recording equipment
teristic. based on the design of A. F. Fischmann, published in
The B.B.C. regrets that the owners of some of the Wireless World as long ago as November, 1954.
older receivers favouring the upper sideband may need It is difficult to understand the use of an 8 -µF electro-
to modify them, but believes that such cases are few and lytic capacitor, Cs to decouple the screen of V,. Unless
that the decision was justified in the general interest. the polarizing current is much below average it will be
E. L. E. PAWLEY. sufficient to reduce seriously the screen voltage, and a
Head of Engineering Services Group, B.B.C. capacitance of 1 pF would be adequate.
The record amplifier has a very high degree of feed-
back, and two identical intervalve couplings. An ampli-
Helping the Blind fier to Mr. Fischmann's specification showed a peak of
PRIOR to losing my sight in 1946, I was a regular reader + 12 dB above the mid -band gain at 0.2 c /s. This leads
of Wireless World. For the past eight years I have been
employed by Decca as a mechanical inspector, using
instruments specially adapted for the use of the blind.
to a tendancy to instability. The peak can be reduced to
C
+2 dB by the use of the following component values :
0.005 µF; C omitted; C 0.25 iLF; R8, 1 MU.
-
The high- frequency peak is smaller, due to the reduc-
I have now become interested in electronics and par-
ticularly in tape recording, which I use mainly for cor-
tion in feedback by
in parallel with R,.
C and can be dealt with by 100 pF
respondence with other recordists in all parts of the It was found impossible to replace V, by a pentode, as
world. I am a member of two tape clubs, both centred suggested in the text, because of the difficulty of provid-
in the U.S.A. I would be interested to hear of any ing a screen supply that introduced sufficiently little
similar British club. phase shift to avoid low- frequency oscillation.
Many of my correspondents are blind, and you will R. C. MARSHALL.
appreciate that our greatest difficulty is to keep abreast Wheathampstead, Herts.
282 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
Gascode A.F.
Amplifier
"Long-tailed Cascode Pair'' as Combined

Pre -amplifier and Phase Splitter

By L. B. HEDGE, Ph.D.

THE " cascode " amplifier -a series connection

Amplifier with "replacement" output transformer.
of two triodes which operates much like a single
triode, with characteristics practically unattainable
in a single triode -has been extensively employed importance only in a stage (the first, barring ex-
as a high- frequency amplifier during recent years, ceptional circuitry) where the input signal is of
and more recently as a first-stage, low- level, audio - sufficiently low intensity to make the signal-to-
frequency amplifier (so- called " pre- amplifier "). stage- generated -noise ratio critically small. In the
Although the cascode was developed as a direct - next -to -final stage of an audio -frequency power
current amplifier for voltage regulator control amplifier, only exceptionally bad design could
application', its recent uses have been largely based make the noise generated in the stage a factor of
on the inherently low level of stage noise. The significance in the performance of the system.
importance of minimizing the signal-to -noise ratio High quality in audio -frequency power amplifier
in a variety of high -frequency applications, including
radar, television and many others, has served to
performance- uniformity of response and low dis-
tortion over the spectrum of audible frequencies
depends in large measure on ' a few closely inter-
-
keep attention focused on this low -noise feature as
the distinguishing characteristic of the cascode, related design elements; the output transformer, the
and its use in the audio-frequency field has also been feedback circuitry, and the frequency, phase- shift, and
based largely on this feature. attenuation characteristics of the inter -stage couplings
The amplifier here described (on which patents which establish the limits within which feedback
are pending) is the result of a return to an earlier may be used as an overall correctives. In general the
view of the cascode stage; it is used here because of output transformer is the effective limiting element
the characteristics for which it was originally in amplifier performance, and recent impressive
developed -its triode -like performance and its improvements have been based on special transformer
high equivalent amplification factor. Although low designs4.
noise is no disadvantage in any amplifier, it is of In exploring the problem of evolving an amplifier

Complete amplifier and power supply. (UTC LS -55 output transformer.)

283
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 design which would make most effective use of an
output transformer of non -critical design-one
which would make the best use of any output trans-
former built into it-it soon became clear that some
major changes in " conventional " circuitry would be
required. A feedback loop to support a high level of
corrective feedback which would include the output
transformer and go back at least to the phase-
inverter stage seemed a minimum reasonable re-
quirement, and with conventional circuitry this leads
to something very much like the basic " Williamson " Ein
layout. With the low gain of most popular phase -
inverter stages, and the high drive requirements of
the output stage, at least one driver stage is required
between the phase inverter and the output stage, and
an additional stage which may be either before or
after the inverter. One direct coupling between µR1,
stages (as in the Williamson scheme) is quite prac- Epp Etn
ticable, but more than one adds serious complications
to the power supply and isolation filter problems. Fig. I. Cathode- coupled phase-inverter (long-tailed pair).
The result is a feedback loop which contains two R -C
coupling networks and the output transformer, with
a possible maximum phase shift or 270 °. Stability Below:- Fig. 2. The cascode amplifier. (A) Cascade con-
+H.T.
nection (B) Signal equivalent -V2
as load for VI (C)
Signal equivalent, VI (D) Triode equivalent of cascode.

Eout
V2

+ H.T.

Ei,
Ein

i (C) (D)

of the amplifier requires that the loop gain be re- _ tp(rp+Rt)+E1

duced to less than 1 before the phase shift reaches (B) E1P

180 °, and, in view of the phase -shift and attenuation E1(/c+I)

characteristics of the couplings and the transformer, fp = (2)
rp+Rt
the frequency range over which feedback can be kept
high must be considerably smaller than the usable rp + R1
R = E1/ip (3)
range of the transformer itself5. The search for a /1 +1
reasonable way out of this vicious circle of conflicting
constraints led to the analysis of the cascode and the (A) in{z _ ip(rp+R1) (4)
cathode-coupled phase- inverter, and finally to the
combination of the two -the " long -tailed cascode ip Ein /1 (s)
rp +R1
pair " (l.t.c.p.).
The aethode-coupled phase-inverter is well known Ein/1 /t(/[+I)Ein
and has been extensively used (Fig. 1). The un -by- ip =
+(/t+2)rp (6)(365)
passed common- cathode resistor provides degenerative R/1 +1 +rP RI.

feedback to the input tube as well as driving potential

for the grounded -grid inverter. The anode -to -anode Eout ipR1 tt(/1 +I)R1 fc'R1
output of this stage is independent of the value of the Eta E;.n Ri +(/c+z)rp Ri +rp
(i)
284 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
cathode resistor if the two valves are matched and
the anode load resistors are equal, and the ratio of
the two anode -to -earth output voltages is6:
EA RL + rp
(1)
Es (µ +1)Rk
Precise balance can be provided by selection of Rk
and RL for given tube characteristics, but if high gain
and reasonable power supply voltage requirements
are to be realized, must be exceptionally large.
IX

The cascode amplifier consists of a conventional

triode with a cathode -driven triode as its anode
load (Fig. 2). Analytically the cascode takes the form
of a fictitious triode with characteristics µ', r', and
g'm the values of which, expressed in terms of the
characteristics of the component triodes (assumed
identical) µ, rp, and gm, are:
IL' = µ(µ +1)
= (µ + 2)r,, Under-chassis view of the complete I.t.c.p. amplifier.
a' ii(µ +1) i+ +1 gm
r..
_ (µ + 2)r,, µ + 2 (the approximate value normally available in an
Typical twin-triodes in cascode connection should
thus provide characteristics as follows:
Type µ rr, gm µ' rP gm
- audio- frequency power amplifier, and the maximum
available from my regulated adjustable supply unit)
check reasonably well with the curves, and even
better with the computed values. Within the regions
6SN7 of good linearity to the two cascodes the 6SN7/7N7
7N7 20 7 k S2 2.9 420 0.15M S2 2.8
should provide a gain of approximately 128 with a

l
6SL7 load resistance of 66k S2 and a anode supply of 475
7F7 70 44 k S2 1.6 5000 3.2M S2 1.6 volts, while the 6SL7/7F7 should provide a gain of
Anode characteristic curves for these two types were about 360 with a load of 250k S2 and the same anode
constructed for design reference (Fig. 3). The curves supply voltage. On the basis of this analysis the
represent measurements on one valve of each type, experimental amplifier was laid out using 7F7's
and may not be good averages in the accepted sense. in the l.t.c.p. stage.
They do provide, however, an approximate basis for The final circuit of the amplifier is shown in Fig. 4.
selection of operating points and load-line construc- Type 1625 output valves (12 -volt heater versions of the
tions. Dynamic checks with loads as indicated on the 807-similar in general characteristics to the KT66)
curves and anode supply voltage (Ebb) of 475 V were used because they were at hand-as were the
Fig. 3. Cascode amplifier anode characteristics and dynamic check test.
3
15
7N7 (a) \ 7F7 (b)
28

---
14
RL,.. 100,000
26
13 E9=o
24
12
R133,000 Eq-0
22 LINA

;-=
II
MI 1= REGION
OF GO00

\ ,=
EAR-
GOO \-2'.
-
r
2
I0
-
G10
DRE
OF
-RL=z5o,000.-,

9
Eq=-1 18

R\ 6,000
I6

li I E9=-I

`1Ir
-
7

12

MIN1111
MOE=
$
-,
.I r. E9=-4
O8

P6

o
HIM=
1.
E9=-1.5

,
E. Eq=-2s
'
0
r
o 100 200 300
.. o 100 200 300
.
S00

ANODE VOLTS
ANODE VOLTS

CHECK -R1 =33,000, Ebb =475V,E9 2=120V,E5 --zV, DYNAMIC CHECK - Rt.= 100.000 Ebb= 475V, Eq¿= 150V, E9 -- I V.
DYNAMIC
EIN° 0.1V r. m. s, EouT - 7.5V r. rms. EIN =0 IVr.m.s, EouT ° I4Vrm.s

285
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 INPUT
Y
Fig. 4. Complete amplifier with
inverter -driver stage.

LIST OF PARTS
C1, C7 204 450V electolytic R5, R11 400kí2, watt
C2, C3 0.14 600V R10 2000 rheostat, 10 watt (Output bias adjust-
C4 120µf, 150V electolytic ment)
40µf, 450V
C5, C6 R12, R15 1kí2, watt
Cs 404 350V R13 100ít, 5 watt
C9 40µF, 450V
Clo 104, 600V R14' l000 pot., 5 watt (Output cathode balance
Chi 5H, 300 ohm, 40mA choke adjustment)
Che 10H, 90 ohm, 200mA choke R16 1Mí2 pot. (Feedback adjustment)
R1 50k0, 1 watt R17 4.7kí2, + watt
R2 68kí2, watt R18 15kí2, 10 watt
R3 500k0 (volume control) Si S.P.S.T. switch (Feedback disconnect)
R4 47kí2, } watt t Cascode 2nd grid voltage Ti Output transformer-(See text)
R5 100kí2, 1 watt j divider T2 Power transformer 375- 0 -375V, 200mA,
R6, R8 220k0, 1 watt 1 Cascode load and balance heater as required
R7 50k02 pot., i watt ( adjustment VI, V2,7F7; V3, V4, 1625; V5, 574; V6, OB2; V7, 6X4.

7F7's. The essential symmetry of the l.t.c.p. stage using a United Transformer Company's LS -55
suggested immediately the closure of the feedback
loop through the grid circuit of the grounded -grid
transformer as a reference -a
typical " good "
transformer (reference 7 covers its use in the " ultra-
inverter, since satisfactory introduction of the feed- linear " connection)-and a " universal replacement"
back voltage into the input grid circuit is somewhat type, unidentified by manufacturer's name or model
complicated by the presence of the volume control. designation, culled from the shop " junk box," as a
Pentode, triode, and so- called " ultra- linear " opera- kind of " worst possible " unit for evaluation of the
tion of the output stage is provided by the alter- system. Fig. 5 indicates the effectiveness of the
native connections (A, B, and C, Fig. 4) for the screen system in providing adequate drive and stable
grids of the 1625's. operation at high corrective feedback levels.
Performance of the complete amplifier was checked The complete amplifier -a
" bread -pan layout "e-
286 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 +2 design consistent with high quality and exceptional
o stability. With a real tc dog" for an output trans-
-4 dB former, this tc tail " will wag it so that it will perform
WITH LS-5S TRANSFORMER 0dB -8W 6 like a thoroughbred !
8
I ! l I I ! ! !
10
o o
r ó g ó
á.
ó
r ó
REFERENCES
(c /s I Hunt, F. V., and Hickman, R. W., " On Electronic
Voltage Stabilizers," Review of Scientific Instruments,
FREQUENCY

January 1939, p.6.

2 Wallman, H., Macnee, A. B., and Gadsden, C. P.,
" A Low-Noise Amplifier," Proc. I.R.E., June 1948,
Ei p.300.
3 Terman F. E., " Radio Engineers Handbook,"
D
-WITH REPLACEMENT TRANSFORMER 0dB -3W McGraw Hill, 1943, p.395.
Nyquist, H., " Regeneration Theory," Bell System
Technical Journal, January 1932, p.126.
M Ñ o Q o
O 4 Williamson, D. T. N., " Design of a High Quality
o Amplifier " Wireless World, April 1947, p.118, and May
FREQUENCY (c /s) 1947, p.161.
Performance characteristics of complete Williamson, D. T. N., " High Quality Amplifier-
Fig. 5.
New Version " Wireless World, Aug. 1949, p.
amplifier. 1949, p.365, and Nov. 1949, p.423.
6 Terman, F. E., /oc. cit., p.224.

Curve A 0dB feedback 0.16V r.m.s.input-U-L connection s Langford- Smith, F., " Radio Designer's Handbook,"
r.m.s.input- pentode connection
0.1IV
Curve B 10dB feedback 0.52V r.m.s.input-U-L connection
p.347. Iliffe and Sons Ltd.
Curve C 10dB feedback 0.38V r.m.s.input -pentodeconnection
connection Richter, W., " Cathode Follower Circuits," Elec-
Curve D odB feedback 0.25V r.m.s.input -.triode connection tronics, November 1943, p.112.
0.12V r.m.s.input- pentode
Curve E 10dB feedback 0.80V r.m.s.input- triode connection
7 Brewer, F. T. C., " A Linear Power Amplifier,"
Curve F 10dB feedback 0.40V r.m.s. input-pentode connection Radio and Television News, Aprii 1953, p.64.
%: s Hedge, L. B., " The Bread -Pan Layout," Audio
Maximum output watts with harmonic distortion less than
1

Engineering, August 1953, p.24.

Transformer Replace- Transf. LS -55 Transformer

ment Pentode U -L Pentode
Connection
Triode COMMERCIAL LITERATURE
A.M. /F.M. Signal Generators, an e.h.f. spectrum
Feedback 0dB 10dB 0dB 10dB 0dB 10dB 0dB 10dB

0.5 0.1 0.5 12 18 10 15 analyser, a v.h.f. alignment oscilloscope, s.h.f.andandan v.h.f. f.m.
wavemeters, a counter -type frequency meter,
30 c/s 0.1
6 8 12 18 12 15
station monitor are among the 25 new instruments described
100 c/s I

8 12 18 12 15
1,000 c/s 3 6 3
10 kc /s 3 6 3 8 12 18 12 15
in the Marconi Instruments 1956 " Electronic Measure-
ment " catalogue. From the company's address at Long -
to 20dB with increase in input acres, St. Albans, Herts. -9A and
Note: Increase in feedback from 10 characteristics less than IdB with Voltage Stabilizers; a.c. for output currents of 0of
voltage of approx. x3 changes output 0 -30A at mains voltage; and d.c. with outputs
0 -7A,
LS -55 transformer and less than 2dB with the replacement transformer. con-
1 -30V and 0 -2.5A, 1 -15V. These and other electrical
from
As may be surmised, trol instruments described in a 1956 general catalogue
is shown in the photographs. Servomex Controls, Crowborough Hill, Jarvis Brook, Sussex. for
neither construction, layout, nor wiring is critical Also a data sheet on an i.f. waveform generator suitable
-
in any sense. The 105volt supply required for the testing servo mechanisms, etc.
Electronic Thermometer, for industrial or medical
use,
cathodes of the 1.t.c.p. stage is an exceptional re- with quick response. Uses germanium thermo -sensitive
quirement, but it is easily met by a simple modifica- device and has accuracy of ±1 °C. Available in four types:
°C; manu-
don of a conventional power supply, as shown in the 25 ° -45 °C; -10 °- 110 °C; -50 °- 160 °C; 0 °- 210
Ultrakust Geratebau of Germany. Leaflet from
wiring diagram of Fig. 4. Since each d.c. connection factured by
the distributors, Headland Engineering Developments,
164-
to the amplifier is to a symmetrical and balanced 168, Westminster Bridge Road, London, S.E.1.
load, isolation, hum and ripple filter can be quite Decimal H.P. Electric Motors, for sound recording
and
reproduction equipment. Shaded pole DHP1 induction motors for
simple. speed 1,345
The output stage cathode bias scheme shown is 100/130V or 200/250V a.c. Type :
Type DHP2D:
r.p.m. at 50 c /s; running torque 2 in-oz. torque
simple and effective for providing final stage balance, speed 2,800 r.p.m. at 50 c /s; runningand Manufacturing 3.5 in-oz.
but it is not in any way a special feature -the William- Leaflet from The Garrard Engineering
son -type network should be equally effective. Co., Newcastle Street, Swindon, Wilts. capacitors with rela-
is not Silvered Mica Plates for assembling
The cathode bypass condenser in this stage tively simple equipment. Description, diagrams and
tables
necessary either, but the author prefers to use it of data in an enginering data sheet from Johnson
Matthey
since it tends to reduce distortion if and when the London, E.C.1. Also
and Co., 73 -83, Hatton Garden,
output tubes, by ageing or for other reasons, depart similar sheets on contact materials, their choice, properties
and availability.
from perfect balance. No provision has been made Vibration Isolators, air- damped, for mounting isequipment
for static balance of anode currents in this stage, designed
to withstand mechanical shocks. A standard type and a rug -
since the author's experience and tests indicate that for protection against a series of 15g shocks Technical
gedized type for protection against 30g shocks.
dynamic balance will produce lower distortion, and data sheet from Cementation (Muffelite), 39, Victoria Street,
that dynamic and static balance frequently occur at London, S.W.1. voltmeters,
different bias adjustment settings. Waveform Analysers, signal generators, valve equipment and
The " long -tailed cascode pair," by eliminating frequency monitors, counters, waveguide test Company of Cali-
other products of the Hewlett- Packard country provided
one inter -stage coupling without reducing gain or fornia, U.S.A. Technical service in -formthis
seriously complicating the power supply require- by Livingston Laboratories. A short illustrated cata-
ments of the conventional power amplifier system, logue from the British representatives, Lithgow Electronics,
1, Grange Court, Sudbury Hill, Harrow, Middlesex.
makes the use of output transformers of non -critical
287
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 WideBund Television Aerials
Review of the More Interesting Types Current in North
America
By M. G. O'LEARY

THE recent inauguration of television in Band III generally a good vertically polarized one if it is
has introduced at least one problem at the receiving rotated 90 degrees about its axis. Therefore it
site-that of an aerial system which will efficiently follows that a successful American aerial design for
receive signals in both bands. Preferably this system dual -band reception will also be successful for this
should bring in both bands on a single transmission in Britain, if mechanically and structurally it is
line. British manufacturers have introduced feasible to rotate it through 90 degrees. Neverthe-
numerous designs to do just this. In the light of less, American designs which do not meet this re-
these events, a review of practice in North America, quirement will be reviewed here also if their design
where television on both bands is several years old, and theory are unusual. Ingenuity will always sug-
should be of interest to students of aerial design. gest adaptation in one form or another.
There are several differences between British and The basis of the problem is the three -to-one
North American television and since the reader is (approximately) frequency ratio between Band -I and
presumably familiar with British standards, only the Band -III channels. A half-wavelength dipole cut
American ones need be reviewed. Channels are for a Band-I channel will be one -and -a -half wave-
6 Mc /s wide. Channel 2, the lowest, is 54 to lengths for Band -III channels. The polar response
60 Mc /s, Channel 3 is 60 to 66 Mc /s, and so and current distribution of this dipole for the Band -I
on to Channel 6 at 82 to 88 Mc /s, with a 4 -Mc /s channel are as in Fig. 1, whereas these character-
gap between Channels 4 and 5. Channels 7 to 13 . istics for the same dipole on Band III are shown
on the high band cover 174 to 216 Mc /s. Other in Fig. 2. The split lobes in Fig. 2 could be used
differences include transmission of 30 pictures per for reception of horizontally polarized signals, at the
second with 525 lines. Nevertheless, these differ- risk of ghosts, but such an aerial would give
ences do not alter the basic aerial problem for extremely poor results on vertically polarized Band -
Band I /Band III reception. However, in North III signals, even when rotated through 90 degrees.
America television signals are transmitted with Nevertheless, no American designs known attempt
horizontal polarization and this difference presents to receive Band -III frequencies in this way. Instead,
an additional consideration to the aerial designer, an attempt is made to alter the Band-III polar
but only with respect to structural and mechanical pattern to obtain a single forward lobe in line with
factors. A good horizontally polarized aerial is the lobe for Band I.

Fig. 2. Polar diagram (a) and

current distribution (b) of a
Band-1 dipole used as a 3'2 wave-
length aerial on Band I 1. 1

(b):
Fig. 3. Current distribution on
a tilted -element aerial; (a) Band
1 (b) Band III.

(a)

(a)

(b)
Fig. I. Polar diagram Fig. 4. Two forward tilted con-
(a) and
current distribution ical dipole aerials (with reflect-
(b) of a
half-wave dipole cut for Band I. (b) ors) stacked vertically for fringe
areas.
288
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 A widely used means of accomplishing this is to LOW PASS FILTERS
cut a Band-I dipole and tilt its elements forward
as in Fig. 3. Doing this does not adversely affect
the Band -I polar pattern, but it combines the split
lobes of the Band -III pattern into a single lobe.
In doing so, small split lobes remain in the opposite
direction, which could cause ghosting troubles in
difficult localities. Usually these tilted aerials have TRANSMISSION
broad-band, conical-type dipoles with a reflector
favouring the Band -I frequencies. In fringe areas Fig. S. Band 1'Band III
combination aerial embodying low -
two are stacked vertically with a spacing favouring pass filters for coupling elements. (Lapointe Electronics Co.)
Band III, as in Fig. 4. Such stacked arrays give
excellent reception at distances of 75 miles from
transmitters on both bands (100 kW e.r.p. Band I;
200 kW e.r.p. on Band III). Fig. 6. In -line combined
Another design popular some years ago functions aerial with two folded
on a different principle (see Fig. 5); A is a half - dipoles.
wavelength Band -I dipole, B is a full -wavelength

.
Band -III dipole, " T "- matched to a balanced trans-
mission line through C. On Band I, A is doubly
" T "- matched to the transmission line through B
and C, but the connections between A and B are
low-pass filters acting as a short circuit on Band I,
and as an open circuit on Band III. Thus A is
: ........
..
............. .......:.

Fig. 7. Collinear arrange-

not directly effective on Band III other than to act A ( : IBI U : D ment of dipoles with (a)
as a band broadening device through mutual coup- closed - end connecting
ling to B; B in turn acts as a broadening device (a) stubs and (b) folded
stubs. Current distribu-
on A for Band-I frequencies. These aerials were tion shown also. Dashed -
supplied with a reflector effective on Band I and a .` line Band 1, dotted-line
director favouring Band III, and were similar to ,r........ .............. ;......w. Band 111.
the tilted conical dipole and reflector in performance.

Collinear Designs
(b'
The " inline " aerial, Fig. 6, consists of two folded INSULATORS
dipoles, one for Band I behind the other for Band
III. The larger acts as a reflector for the Band -III
frequencies and, in turn, is itself aided by a reflector
cut for Band I. The two dipoles are served by a
single transmission line, which, along with the dipole Fig. 8. Two -stack
spacing, is so arranged that there is a minimum of yogi embodying
interaction between the dipoles when acting on their collinear ele-
respective frequencies. ments of Fig. 7.
Collinear designs as in Fig. 7(a) and (b) have also (Technical Appli-
been popular. In Fig. 7(a), A, B, and C are each ance Corp.)
a half -wavelength for Band III, separated by closed -
end quarter -wavelength stubs (Band III). The D
current distributions for each band are indicated
(compare with Fig. 2) and the result is a single -lobe BAND I
polar response on both bands. A popular con-
figuration takes the form of four such aerials stacked
vertically and backed by a screen or reflector spaced
to give good reflector action on both bands. Re-
.....
.

ferring again to Fig. 7(a), the total length A +B+C,

and the separating stubs, is effective for Band -I ` BAND DI

response. Stacking spacing of these multiple aerial Fig. 9. Current distribu-

systems is a half-wavelength on Band III, and thus B
tion on director A of Fig.8.
favours this band.
Figure 7(b) illustrates a different way of accom-
plishing this collinear effect. Rather than separate
the three collinear elements with quarter- wavelength (a)
stubs, small quarter -wavelength (Band III) folded Left: Fig. l0. Triple
dipoles, D and E, are used. These reverse the folded dipole aerial
BAND I CURRENT
(a) with current
phase of the Band -III currents by 180 degrees and distribution (b)-
thus all three collinear elements act in phase to .........
give a single-lobe polar response. On Band I the
combination acts as a single dipole. Fig. 8 shows IAI
how this design can be used for a two -stack yagi
system. BAND III
Figure 8 calls for some additional explanation- (b) CURRENT

WIRELESS WORLD, JUNE 1956

289

www.americanradiohistory.com
 HEAVY LINES= LOW BAND ELEMENTS Fig. 12. In -line design using conical
dipoles on hoth bands and several
parasitic elements. (Kay - Townes
Antenna Co.)
1
(a)

Below : Fig. 13. Dual -band yogi using

"T "- matching and collinear elements.
(Winegard.)

(b)
Fig. 11. Current dist-
ribution on Bands I and
III of triple folded dipole
aerial of Fig. 10 with
HEAVY LINES = HIGH BAND ELEMENTS added parasitic elements.

the " fed " elements (those to which the feeder is be much more sensitive than the simpler designs,
joined), are spaced a quarter -wavelength for Band I such as the one illustrated in Fig. 6. Fig. 12 shows
and three -quarter -wavelength for Band III. Twin- a typical example. The larger conical dipole receives
fed dipoles spaced a quarter -wavelength and fed Band-I signals while the smaller catches the
at the rear are well known for having a good front - Band-III signals. The inter-connections between the
to -back ratio (not so true when vertically polarized), two, in combination with spacings and associated
broad bandwidth, and fairly good sensitivity.' parasitic elements, neutralize the normal split lobes
Reasonably similar results could be expected on of the conical dipole on Band III and shape a single
Band III with three -quarter-wavelength spacing. forward lobe. This " inline " design has a half -
Director A is actually two collinear Band -III direc- wavelength conical dipole for Band I behind a full -
tors separated by a quarter-wavelength transmission wavelength one for Band III.
stub (Band III), the whole acting as a Band -I director
as well. Reflector E is cut for Band I. Reflector D Complex Systems
is three collinear Band-III reflectors insulated from
each other by fibreglass rods. D has no significant An extremely interesting type of yagi has recently
effect at Band-I frequencies; E is spaced too far been introduced based on a somewhat different
from C to be effective at Band -III frequencies. principle to any previous designs. Fig. 13 illustrates
Figure 9 illustrates the action of the director A in the use of this principle in a dual -band yagi design.
Fig. 8, and shows the current distribution for the Since the theory of this yagi is a combination of
two bands. This idea has been used also for stacked this and several other principles, some elaboration
dipoles where both elements are fed and backed by seems indicated. On Band I the active portion, A,
a screen -type reflector. The dipoles are made of B and C, is a twin -fed yagi, stagger -tuned to give
angle section for structural as well as wide band better cover of the entire band. The forward
considerations and slotted to reduce wind resistance. dipole, C, is connected to the transmission line, and
The stacking distance favours the Band -III chan- the inter -connection, E, to the rear dipole is trans-
nels, while the reflector spacing favours the Band-I posed. The spacing between the two dipoles is
frequencies. one -eighth wavelength for the median frequency
Figure 10 (a) and (b) illustrate the development of Band I. This configuration gives a good front-
of the so- called " tripole " type of aerial. Two to-back ratio, broadband characteristics, and, for
Band -III folded dipoles are coupled with a single normal dipoles, very low aerial feed -point resist-
Band -I folded dipole and connected together so that ance.' This latter property, normally not desirable,
dipoles D and E not only neutralize the out-of- is countered with a specially designed " T" match
phase Band -III current in the A section, but pro- (or dipoles B and C) to raise the resistance. Incor-
vide a resultant current from the combination D, porated into the "T" match are two collinear half -
A, and E which is in phase with those at B and C. wavelength dipole elements for Band III reception.
D and E have negligible effect at Band -I fre- The three -eighths-wavelength spacing which results
- quencies. This " tripole " has been used both with for Band -III frequencies is not quite as effective
screen -type reflectors, and in yagi configurations. In as the one -eighth -wavelength spacing of Band I,
the latter form directors and reflectors for both but the large number of parasitic elements on
bands are interspaced with the other elements. Band III more than make up for this. Close ob-
Those for Band III are three collinear elements servers will discern that the long directors are, in
isolated from each other with fibreglass couplers. principle, the same design as those in Figs. 7 and 8,
Fig. 11 (a) and (b) illustrate the action of the but here the quarter-wavelength shorted transmis-
various elements on the respective bands. sion stubs, x (for Band III), are folded for com-
A number of more complicated " inline " designs pactness.
have been introduced recently which are alleged to The designs discussed so far have all attempted
290 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 BAND III
POLAR DIAGRAM
/

Fig. 14. Unusual design

of in -line, dual-band yogi
using "wing" dipoles.
(Trio Mfg. Co.)

PARASITIC
DIRECTOR -- BAND I
POLAR DIAGRAM
INSULATOR \ INSULATOR
Fig. 16. Polar diagrams
of "wing" dipole shown
in Fig. 15 on Bands I and
111 respectively.

Fig. 15. Details of wing - Fig. 17. Another dual -band system with
shaped elements in the three folded dipoles, and unusual
TRANSMISSION
aerial of Fig. 14. phasing arrangement (Sabre).
LINE

to obtain single forward response lobes on both designs available in North America. Emphasis has
bands exclusive of parasitic action; that is, in the been placed on principles of operation rather than
primary (fed) elements only. Lately several designs on the many variants of a given principle that are
have been introduced in which parasitic action has available. But, before closing, the writer cannot
been coupled with primary element characteristics resist the temptation to include Fig. 17, a dual -band
to accomplish the same result. One such is shown design with claimed " Miracle Phase." Not having
in Fig. 14. Here Band -I action is that of a stagger- seen the patent, or other explanation of its theory,
tuned, twin folded -dipole (A, Ai) yagi with one the writer hesitates to put forward his analysis of
director (D) and two reflectors (R, R1). Fig. 15 its action. It is presented mainly as a bit of mental
illustrates the configuration of the wing-shaped exercise for students of aerial design.
folded dipoles. Two of the three directors (X, Y) REFERENCES
are actually triple collinear Band -III directors as ' T.V. and Other Receiving Antenna, A. B. Baily
discussed for previous designs. Fig. 16 shows how (Rider).
the dipole configuration, plus the built -in parasitic The A.R.R.L. Antenna Book.
director, act together to produce a single forward Radio and Television News, October, 1955, p. 91

lobe for Band -III frequencies.' There are other

such designs which allegedly correct Band-III
directivity parasitically. "Analysis- Synthesis" Telephony
All the designs discussed are claimed to be broad-
band types having good aerial characteristics on all ECONOMIES in bandwidth of the order of 100 1 are :

twelve channels. The writer has witnessed that this envisaged in a system of speech transmission, under
is so for a good many of them in an area where
development by the Post Office Research Station, which
was demonstrated at the Royal Society Conversazione in
the following channels could be received: 2, 150 May.
miles; 3, 75 miles; 4, 150 miles; 5, 80 miles; 7, 40 Speech sounds can be synthesized by applying pulses
miles; 8, 75 miles; 10, 80 miles; 11, 16 miles; 12, of different amplitude and repetition rate (larynx excita-
160 miles. Extending the responses of an aerial to tion) to resonant tuned circuits (cavity formants of the
cover the high -band channels is not excessively diffi- mouth, etc.) and adding bursts of white noise (hissing
cult, but the low -band channels are not quite so consonants). When circuit elements of this kind are
easily covered as the frequency ratio is wider, the connected to a loudspeaker and energized in the proper
sequence by signals originating from an equivalent
band being 54 to 88 Mc /s. In some of the designs analysis of the speech at the sending end, intelligible
discussed the broadening devices are quite apparent and often realistic speech is heard.
and in general these take the form of conical -type In the Post Office analyser three formant frequencies
dipoles, stagger -tuned dipoles, director lengths were selected by tuned circuits, the larynx tone by isolat-
favouring the high end of the band, reflector lengths ing the peaks, and the hissing sounds (fricatives) by their
favouring the low end of the band, parasitic spacing high-frequency content. Some ambiguities are inevit-
favouring one or another portion of the band, twin able in a simple analyser of this kind, but the results so
yagis, large diameter dipoles, folded dipoles, etc. In far achieved are undoubtedly promising.
For convenience a six -way cable was used in the
other types the broad -band devices are not so ob- demonstration to connect the transmitter and receiver,
vious; apparently element interaction and mutual but there is no reason why the information should not
coupling having been used in these cases to give the be encoded for transmission on a single channel.
desired results. Also all these designs are claimed
to be a good match on all 12 channels to the 300 -ohm
transmission line which seems to be the most com- "Magnetic Tape Amplifier." In Fig. A, p. 123 of the
monly used type in North America. March issue a 0.1µF blocking capacitor should be
No attempt has been made in this discussion to included in the lead from the junction of R,oa and R,ob
include all of the multiplicity of dual -band aerial to the twin -T network.
291
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 'Manufacturers' Products
NEW EQUIPMENT AND ACCESSORIES FOR RADIO AND ELECTRONICS
Decade Oscillator due to gravity) gives an output of about 20mV and
BOTH sine and square waves over a frequency range this is maintained within ± 10% over a range of 40 c/s
of 10 c/s to 100 kc /s are provided by the decade to 10 kc/s. The transfer characteristic is stated to be
oscillator made by Winston Electronics, Ltd., Govett linear up to 1000g. To minimize spurious readings
Avenue, Shepperton, Middlesex. It is of the Wien the transverse sensitivity has been limited to 5% of the
bridge type with thermistor amplitude stabilization axial sensitivity.
Like the accelerometer, the strain gauges are intended
for alternating displacements (originally, vibration in
turbine blades) and it is claimed that their sensitivity
is over 2,000 times higher than conventional resistance
elements. They are 0.035 -in in thickness, i-in long
and either ;--in or }-in wide; the frequency range quoted
is 20 c/s to 50 kc /s. An important advantage is that
they can be used as driving elements to excite as well
as detect resonances. Temperature limits are -50 °C
to + 100 C.
°

Dual Moving -Coil Loudspeaker

THE B.T. -H. Type K10A moving-coil unit used in
cinema sound installations can now be bought for use
with high -quality domestic equipment. Frequencies up
to 1700 c/s are radiated from an 18 -in cone and above
Winston Electronics decade oscillator, 10 cis -100 kc /s, that frequency from a coaxial horn- loaded pressure unit.
sine or square waves

B.T. -H. dual coaxial loudspeaker

within 1%. The output attenuator is calibrated (Type KIOA) and cross -over filter
0 -10V with switched multiplier of x 0.1, x 0.01 and
(Type G2A).
x 0.001.
Harmonic distortion on sine waves is said
<1% and on square waves the rise time is toabout be
0.3 µsec at 100 kc /s. At 10 c/s the maximum drop in
the horizontal part of the wave is 2 %. Frequency
stability is <1% for ±10% change of mains voltage
and about 0.02% for ambient temperature changes.
Valve replacement is simplified by the use of 12AT7s
throughout.
The price of the decade oscillator is £57.
Barium Titanate Transducers
AN accelerometer and strain gauges for vibration test-
ing are now being produced by the General Electric
Co., Ltd., Kingsway, London, W.C.2, in which piezo-
electric barium titanate elements perform the conversion
from mechanical strain to electrical output.
The accelerometer (Type E) makes use of a disc After emerging from the centre pole of the large magnet
element in contact with a 10 -gram mass. An alternat- the horn divides into twin flares, arranged to maintain
ing acceleration of "g" (equivalent to the acceleration wide distribution in the horizontal plane at high fre-
quencies.
The I.f. drive unit has a magnet with a flux
density of 14,300 gauss and total flux of 285,000
maxwells; the figures for the h.f. unit are 12,700
gauss and 48,000 maxwells. The power handling
capacity of the unit as a whole is rated at 20 watts.
Particular attention has been paid by the makers
to the delayed resonance response, and this is
claimed to be free from anomalous effects.
A two -section cross-over filter (Type G2A) with
an attenuation of 12dB /octave at cross -over is
included in the price of £45. The makers are the
British Thomson -Houston Co., Ltd., Rugby.

We regret publication of this issue of

Wireless World has been delayed. The next
issue (July) should be dispatched on June 29th,
but with the August number we hope to
resume normal publication (on the fourth
(Left) G.E.C. barium titanate accelerometer in unscreened
and screened versions, and (right) vibration strain gauges
ITuesday of the preceding month).

292
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 Aerial Cross -Over Network
Unit for Combining Band and Band III Aerials
Design and Construction of a I

By L. S. KING, B.Sc.(Eng.), A.M.I.E.E.

THE cross -over network is an electrical filter unit up to cut-off will suffice. The filter will then have
two distinct portions, one called the low -pass (LP)
which enables Band I and Band III aerials to the be
section which will allow the passage of Band I fre-
together to a common downlead to
coupled quencies with low loss, but which will offer a high
television set. It consists of precise values of impedance to Band III frequencies so as effectively
inductances and capacitances to meet certain known to cut off the latter. The other portion, called the
frequency and impedance data. high -pass (HP) section, will allow the passage of
The advantage of such an arrangement lies in the Band III frequencies with low loss, but will effectively
single downlead, and the ability to select either cut off Band I frequencies.
Band I (B.B.C.) or Band III (I.T.A.) by simply With these points established, there is no reason
turning the knob of the tuner on the set. Of why the relatively slow rising characteristic curve
course, where a combined Band I /Band III aerial of each portion should not overlap as shown in
is used, the cross -over network is not required
and
Fig. 1, the only necessary condition being that each
the said advantage does not arise -or for that frequency in turn is passed by its own filter section
matter perhaps, where the set (or set and convertor) and not by the other.
has two inlet sockets. This allows the use of the simplest form of filter
A disadvantage in the use of the cross -over net- and, in general, of one that would be useless in the
work is not operational, but electrical. The intro- telecommunication art; that of the half section as
duction of any network must involve electrical shown in Fig. 2.
losses because the network components are not It will now be seen from the general shape of the
pure reactances (which alone would be non- dissi- characteristic curves of this type of simple filter
pative) and the position that arises is whether any
losses at all in signal strength can be afforded. On
Band I, in the service areas, some loss can usually
be accepted, but this state of affairs is often not
the case on Band III with its considerably higher
radio frequency and greater attenuation in signal
strength.
The two aerial systems to be coupled may indi-
vidually be producing pictures on I.T.A. as brillianta
and as clearly defined as on B.B.C. and with
background almost as steady, but the condition inside
the receiver circuitry may not be similar and may
be very different in the two cases. The receiver and
is provided with automatic gain control (a.g.c.),
this will attempt always to match up the signal to the
required level by a varying amount of amplification, 'co
but to this, of course, there is a practical limit. 45
(Mc /s)
If the I.T.A. signal seems poor and on lookingofcare- the
FREQUENCY

fully at the picture there is a faint stirring

background as compared with a steady background Fig. I. Attenuation, or insertion loss, of the high- and-low-
filter sections comprising the aerial cross -over net-
on the B.B.C. channel, then it might be wise to pass
work described in the text.
leave well alone and not introduce any further
losses at present on the I.T.A. channel. However,
there is no reason why such a filter should not be t, C2
tried as it is relatively easy and cheap to construct;
also, it may be possible to improve the signal strength
by means of a larger and higher aerial. M

The network is an electrical filter, although the

¢ ,
W

hardly a filter as would be recognized as such in filters

OUT P UT
á
z
telecommunication field, where steep -sided -á
are enforced by the closeness of adjacent com- m

munication channels and the necessity of non-

interference between these close- spaced channels.
In the case of Band I/Band III, we have something
H.P SECTION
L.P SEECTION

like 140 Mc /s separation between the relatively

low frequencies of Band I and the higher frequencies
TO
TELEVISION
of Band III. Consequently only a few inductors RECEIVER

and capacitors are required in the relatively cut simple

-off circuit diagram of the aerial cross-
filter which can be used, since steepness of Fig. 2. Theoretical
over network.
is not a necessity, and a much more gentle slope
293
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 section, Fig. 1, that in the case of the LP section
(Band I), say, the cut-off of that section must be CZ
well up to the Band III operational frequencies for
T..,+....
the rather gentle slope of the curve to produce a
sufficiently low attenuation, A, at its own operational
frequencies. Imagine a curve of the same family
drawn more to the left and shown dotted in Fig. 1.
It would still be effective to Band III frequencies
as it would cut off at an even lower frequency, but
the attenuation, or insertion loss, at Band I fre-
quencies has now gone up from A to A'. This
means that the insertion loss at its admittance fre-
quency has probably increased to a prohibitive value
and so the filter would be unsuitable for our purpose.
Similarly, the HP section of the filter should be
designed for a cut-off frequency near to Band I
frequencies, and here, low insertion loss to Band III
frequencies is even more imperative.
Having determir_ed on this, as the basis of design, INPUT INPUT
we will set the cut -off of the LP filter section at BAND I TO BAND III
TELEVISION
190 Mc /s and that of the HP filter section at 50 Mc /s, RECEIVER
both to match into 70 ohms impedances.
Formula for the calculation of filter components Fig. 3. Practical version of the aerial cross -over
are given in the Appendix, and using these, we get network.
The thin laminated base measures 2f x 4fin and
the values for capacitance C shown in the following should
be mounted in a screening box about tin deep
table :-
benzine). The stock type formers are internally
CAPACITOR TABLE threaded for dust cores but no cores are used with
these coils.
Position Calculated Value Use as below Obviously, it will not be possible to achieve
(pF) (pF) these fractional portions of a turn but they are
given here so that the constructor knows what
cl 24 25 should be aimed at. It is not possible to achieve
C2 23 50 40 in series 3.85 turns but 3.8 turns can be obtained and so a
= 22.2 compromise is made.

Now it is not usually possible to get capacitors Final Design

with capacitances as calculated and some com- Distributed capacitance in the coils will be very
promise has to be made. Also, some series or small but it is still appreciable when dealing with
parallel arrangement may have to be built up ; an extremely small values. This capacitance would
expedient that is adopted here. have the effect of lowering the impedance to some-
For example, 23 pF is made by putting 50 pF thing lower than the expected impedance as cal-
in series with 40 pF to give 22.2 pF as the nearest culated. It is extremely difficult to get an idea
to 23 pF. Remember the rule for adding capacitors of the capacitance due to the coils, so it would be
in series is similar to resistors in parallel, viz.; well to design for an impedance somewhat higher
Product 50 x 40 2,000 than 70 ohms, say of the order of 85 to 90 ohms.
22.2 For a nominal 90 ohms impedance, the component
Sum or 50 -F 40 90 =
Low loss silvered -mica capacitors should be L AND C VALUES FOR 90 -OHM IMPEDANCE
used. Making the inductors is liable to give trouble FILTERS
owing to the fractional number of turns as calcu-
lated and the small diameter of the former that is Sec- Com-
tion ponent Value How Derived
required. The writer found that while 4 turns on
Lt, produced a fair picture, 3.8 turns gave a better C, 18.6 pF 30 pF and 47 pF capa-
one, so the inductance apportioning is fairly critical. citors in series = 18.3
Formula for simple -layer inductors are well known LP pF nearest value.
and using No. 26 s.w.g. insulated wire on fin. dia. L, 0.15 pH 4 turns No. 26 s.w.g.
polystyrene formers the number of turns required double silk covered
are given in the inductor table. The wire should wire close wound on
be secured to the formers with a small blob of fin dia polystyrene
" Styrene " cement (polystyrene dissolved in a little former

INDUCTOR TABLE C2 12.6 pF 20 pF and 30 pF capa-

citors in series = 12
pF nearest value
Position Inductance (pH) Turns on lin dia HP L2 0.102 pH 4 turns No. 22 s.w.g.
former double cotton covered
wire close wound on
L, 0.12 3.85 fin dia polystyrene
L2 0.112 3.75 former
Capacitors to be 1 5% tolerance or better.
294
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
values are as follows, but here the wire gauge for APPENDIX
L2 has been changed to No. 22 to obtain an even
The filter types used are constant -k, half -section
number of turns for the inductance required.
The writer has made up networks to both 70
are as follows:
Low Pass Section
-
networks, and the formulae used for these half-sections

and 90 ohms but the nominal 90 -ohm network was L1==f:

preferable. This network is illustrated in Fig. 3
>r

and is shown assembled on a thin laminated base

with tag spacings of lin. and 2in. centre-to -centre
Cl - Ufo Ro
1

across. Judging from the B.B.C. picture there High Pass Section Ro
L2 =
was no evidence from the brightness or definition 41r f,,
that the filter had been inserted, but on Band III C2
where, in the writer's case, little signal strength 477. f. R,
can be sacrificed, the insertion of the filter did where L is in henrys
produce some reduction in brightness, but generally C is in farads
to the extent that it was noticeable only when Ro is the equivalent terminating resistance in ohms
compared with the brightness with the filter removed. L is the relevant cut -off frequency in c /s.

Abnormal V.H.F. Propagation

Determination of Radio Refractive Index Structures from Weather Data
By A. H. HOOPER

IT is now well known that there is a considerable necessary both for the latter case and for the more
degree of association between v.h.f. propagation and usual case of intermediate conditions to determine
weather conditions. On the seasonal scale, for approximately the vertical structure of radio refrac-
example, the general level of signal strength is tive index. Precise evaluation of radio refractive
weaker in winter than in summer, while over periods index values for quantitative work is necessary only
of several days duration marked departures from on the proportion of occasions when significant
seasonal averages are found to develop. On the variations in structure occur.
latter occasions it is frequently found that a spell The method to be described enables a graph of
of fine settled weather is being experienced. On a radio refractive index with height to be prepared
still smaller time scale, signal strengths are found very rapidly from meteorological data. The state
to increase and decrease from normal for a matter of the lowest layers of the atmosphere is clearly dis-
of hours. played, and on those occasions when numerical
In time these effects come to be regarded as values of radio refractive index and height are
associated with the weather conditions observed. required, use of a measuring scale enables them to
Detailed examination has revealed, however, that be read with as much accuracy as the basic observa-
such associations exist in only a proportion of cases; tions justify. As the numerical computation of a
there are, for example, many spells of fine settled series of values of radio refractive index from the
weather with nothing unusual in the way of propaga-
tional effects occurring. In consequence assess- 8,000
ments of propagational conditions cannot usefully
be made from studies of local weather conditions or
of weather charts alone.
It has been found that fluctuations of signal
strengths at metre wavelengths can, to a very large 6,000
extent, be explained by variations in the amount of 800 mb
downward bending of the radio waves in passing
through the lower levels of the atmosphere. The
significant quantity in such circumstances is the
W
L
vertical structure of radio refractive index. The 4,000

only data from which this can be determined on a _ 900 mb

day -to -day basis are the results of meteorological
soundings of the atmosphere given in the form of a
series of values of air pressure, dew point and 2,000
temperature.
On a day of vigorous atmospheric motion it is 1,000 m b

often found that the radio refractive index decreases

with height at a fairly steady rate and that signal
strengths are sub-normal. On occasions of stagnant
MEAN 1,050 mb
SEA LEVEL
300 350 400
atmospheric conditions, however, large departures 00 250
M
from a steady rate, and with them enhanced signal
strengths, may occur. While the former condition solution for the
Fig. I. Elementary form of graph giving a
can, with practice, be ascertained by inspection of first term of the basic equation.
the meteorological sounding data it has been found
295
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 700 reported data is a laborious and time -consuming task
300 350 it is evident that the proposed method offers con-
M siderable advantage.
Derivation of Graphs.-The radio refractive index

750
i 8000
µ is often expressed in "M-units ", given by
M' =(µ- 1).106
This yields values ranging in the lower troposphere

. C
7000
from 360 to 280 units.
In terms of meteorological parameters, M' is
approximately given by
800 6000 79.P 379200.e
T T2
where P =total atmospheric pressure (millibars),
5000 e= (partial) water-vapour pressure (millibars) and
T= temperature (degrees absolute). The error in
M' is less than 1% at below 336 °A (= 63`C).
1111111151111111 The graph adopted for displaying radio refractive
I S11111111,1111
1
index structure has height as ordinate and M' as
abscissa, both being linear and increasing conven-
tionally.

ia
3000
A graph of this type is shown in Fig. 1, together
with superimposed curves of P and T derived from

Illl'l%'l'i
110,\\\\1i\\\0fi\\i\\
2000
the expression P . A plot of observed values of
950
P and T using the superimposed grid gives, upon
1000
reference to the underlying grid of M' and height,
1111 a direct solution for the first term of the given

\
V\\\\ ML 30011111111
,
11\\MM\ equation. For the rapid evaluation of the complete
111011111M\N MIMI'
1000
li \\\\S\\f
nIII 3/12/54 CAI\MU equation, however, there are advantages in replacing

MS 4/IZ/KMU
MS these superimposed lines by a sufficiently accurate

1050
11 Mt IM
0°
"REFRACTION TEMPERATURE"
-50°
1\\\
l\\\ 0200 G.M.T.

Tr (ARBITARY ZERO)
\\ approximation in the form of Fig. 2, from which the
basic grid of M' and height has been omitted for
clarity. This alternative grid comprises a set of
horizontal pressure lines and straight, parallel, tem-
perature lines. The magnitude of the approximation
Fig. 2. Plotting chart in terms of pressure and the is considered later.
"refraction temperature" derived from Fig. 3. (Continued on page 297)
TABLE 1

Distance, in inches, of pressure lines in Fig. 2 above datum 1050 mb

.

Pressure
Millibars 0 IO 20 30 40 50 60 70 80 90
700 11.02 10.63 10.25 9.88 9.51 9.14 8.78 8.43 8.08 7.73
800 7.39 7.05 6.72 6.39 6.06 5.74 5.42 5.11 4.80 4.49
900 4.19 3.89 3.59 3.30 3.01 2.72 2.43 2.15 1.87 1.60
1000 1.33 1.06 0.79 0.52 0.26 0

TABLE II
Values of Tr

Dew Point (°C)

Abso-
lute 20 -IO o +5 +10 +15 +20 +25
zero +30 +35

Air
Tern-
+40
+30
+40
+30
+35
+25
+ 29
+18
+16
+ 5
+
--19 6 --21
7 -26
-40
-50
-67
-83 -124 -178
pera-
ture
+20
+10
+20
+10
+14
+
+
- 7 - 7
-19
6
-34 -55 -83
-101 -145
-31 -48
(°C)
-10
0
-10
0 --17 4
6
4
15
26
-31

296
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 The second term of the equation can be obtained
by a graph relating e and T and the position in Fig.
200

'
"
2 denoting total M' value then found by moving,
from the position given by the first term, through 150
50 F,
the corresponding number cf units to the right. Since °o
/4>

the expression for the second term,

379200.e
does 100
pip/ c
.J

la
+20
not contain P this horizontal distance is constant for Tr(D EG. C)
all pressure levels. It can, therefore, be expressed -
-' ,'--
+10
in units of the parallel temperature lines. Hence
50

the final position denoting M' can be found by the - o -io

-20
intersection of the appropriate pressure line, and a - 273
" temperature " line obtained by adding to the o

observed temperature a certain number of degrees

determined separately by a graph of T and e. This
leads to the concept of an " effective refraction +40 +30 +20 +10
temperature " Tr which gives directly the final AIR TEMPERATURE (DEG.C)
sloping temperature line required.
The graph of Fig. 3 combines the above steps to Fig. 3. Graph from which "refraction temperature" can
be derived from observed values of dew point and tem-
give directly the refraction temperature correspond- perature at different pressure levels.
ing to each combination of air temperature and dew
point likely to be experienced in normal work. The TABLE III
process of deriving the position of radio refractive Chart error (M' units)
index value is reduced, therefore, to ascertaining Tr
from Fig. 3 and plotting a point on Fig. 2 at the Pressure Temperature ( °C)
intersection of the corresponding sloping Tr line (mb)
with the appropriate pressure line. +30 +20 +10 0 -10
Numerical Values. -In most cases the structure
given by a series of points so determined wilt be 700... -12 -9 -6 -3
all that is required. When, however, the approxi- 800... -- 5
2
-3
-I
-I
0
+1
+1
mate height of a significant point is required it can 900...
0 0 0
1000... 0
be directly measured by placing a straight edge,
graduated to the basic height scale adopted, between
the pressure levels of the surface and of the point
concerned. Values of radio refractive index for a When much work is contemplated the chart is
given occasion are conveniently obtained by draw- best prepared on tracing linen and then duplicated
ing a vertical line representing the value M'=300 as required.
through the intersection of the -10 °C temperature Table II gives data for reproducing the Tr graph
line and the 1000 mb pressure line and then measur- of Fig. 3.
ing with an appropriate scale the horizontal dis- Accuracy.-Considering Fig. 1 it will be seen that
tances from this line to the series of plotted points. replacement of the converging curved temperature
The scale is graduated in M' values and when lines with a set of parallel straight temperature lines
placed in registration with the line gives a direct introduces an error in M' which varies with position
reading of M'. in the diagram. The slope selected for the straight
Specifications.-The results of soundings of the lines has been chosen to minimize this error in the
atmosphere made twice daily at stations of the normal area of use. The magnitude of the error is
Meteorological Office are published in the Daily given in Table III. It will be found that errors in
Aerological Record*. The information is in the excess of three units are extremely rare.
form of temperature and dew point at pressure levels There is, of course, additional uncertainty arising
selected individually for each sounding so as to from errors in reading from the graph of Fig. 3
delineate the observed structure. On 1st January, and from plotting.
1956, the scale of temperature was changed from The height at which a point is plotted will be in
Fahrenheit to Centigrade, and the chart and graph error by a factor related to the departure of the mean
are designed for use with the current temperature temperature of the air column (between surface and
scale. the level) from the mean temperature (283 °A)
For the plotting chart (Fig. 2) it is recommended assumed in Fig. 2. Correction is at the rate of 7%
that one inch represent 1,000 feet and 50 M-units. for every ten degrees of departure. On nearly all
It is then possible to use, a rule graduated in inches occasions in the vicinity of the British Isles the
to read both M' and height. Table I gives the dis- correcting factor is less than 3 %.
tance from the lower edge at which the pressure lines In making use of the radio refractive index struc-
are drawn. ture obtained from radiosonde data it is assumed
The Tr lines are drawn, five to the inch, with a that the given values are representative of an area
slope of -1.9 inches in ten inches of height. The sufficiently large for the purpose in hand. It is
given separation arises from the convenient fact that, necessary, therefore, to pay attention to the effect
between -10 °C and +30 °C, one degree Centigrade both of instrumental uncertainty and of atmospheric
is closely equivalent to one M -unit. inhomogeneity. Consideration of these effects leads
to ,the conclusion that the standard deviation of a
* Obtainable on application to the Director, The Meteorological spot value of radio refractive index from the British
Office, Dunstable, Beds. Single copies 4d, 1 month 9s, 3 months radiosonde is about 10 M-units, while the uncer-
24s, 1 year 95s.
297
WIRELESS WORLD. TUNE 1956

www.americanradiohistory.com
 tainty of the height of a given point is represented continuity which appears on both results, although
by a standard deviation of 217 feet. It is concluded, at different heights. The discontinuity extended as
therefore, that on most occasions the approximations a layer over south -eastern England from the morning
of Fig. 2 are acceptable and that the method is of December 3rd and then rose and drifted away
sufficiently accurate to depict meteorological sound- early the following day. On the evening in question
ing data in this form. a very strong signal, 40 dB above normal, was
Example,-The radio refractive index structures received in Sussex on a frequency of 180.4 Mc /s
derived from the results of two successive soundings from Sutton Coldfield, while communication on
over Sussex in December, 1954, are shown in Fig. 2. 145 Mc /s was achieved between southern England
A marked zone of discontinuity can be seen, between and Germany. From the results of similar analyses
A and B on the earlier and between C and D on the carried out daily over nearly a year it is known that
later sounding. From these, and the results from the two effects, extended propagation and refractive
adjacent areas, it is apparent that it is the same dis- index discontinuity, are very closely associated.

BOOKS RECEIVED
Department of Scientific and Industrial Research, ferent elements of equipment are followed by direc-
Annual Report, 1955. Summary of work of all research tories of manufacturers. Pp. 136, with numerous
establishments of the department. Includes notes on illustrations. Price 8s 6d. Miles Henslow Publications,
the investigation of tropospheric propagation and scat- Ltd., 99, Mortimer Street, London, W.1.
tering, direction-finding problems and noise in semi-
conductor devices by the Radio Research organization. Television and Radar Encyclopedia. Edited by W.
Pp. 321. Price 7s 6d. Her Majesty's Stationery Office, MacLanachan. Revised second edition of an illustrated
York House, Kingsway, London, W.C.2. glossary of terms which includes signed contributions
from a number of recognized authorities on specialized
Germanium Diodes, by S. D. Doon. Monograph in subjects. Pp. 216; Figs. 224. Price 30s. George
the Philips Technical Library popular series on the Newnes, Ltd., Southampton Street, London, W.C.2.
history, characteristics and applications of crystal diodes.
Pp. 85+ viii; Figs. 72. Price 9s 6d. Cleaver Hume Radio Servicing Pocket Book, Edited by E. Molloy
Press, Ltd., 31, Wrights Lane, London, W.8. and J. P. Hawker. Condensed information on test
equipment and its use for fault-finding in sound broad-
High Fidelity: The Why and How for Amateurs, by cast receivers, including v.h.f. circuits for f.m. recep-
G. A. Briggs. Beginners' guide to the science and art tion. Lists the valve sequence and intermediate
of sound reproduction, covering the equipment required frequency of popular post-war receivers and includes
and its handling to produce the best results in the valve data and valve and battery equivalents. Pp. 200;
home. The text is enlivened by the author's personal Figs. and tables 188. Price lOs 6d. George Newnes,
experiences which included the giving of large -scale Ltd., Southampton Street, London, W.C.2.
lecture-demonstrations of high -quality sound on both
sides of the Atlantic. Pp. 188; Figs. 65. Price 12s 6d. Radio Receiver Circuits Handbook, by E. M. Squire.
Wharfedale Wireless Works, Ltd., Idle, Bradford, Yorks. Descriptive analysis, from the practical point of view,
of the function of the different stages of sound receivers
Hi-Fi Year Book (1956). Edited by Miles Henslow. and amplifiers with a chapter on f.m. discriminators
Survey of current practice in high -quality sound re-

E
and their associated v.h.f. circuits. Pp. 156; Figs. 122.
production from discs, magnetic tape and radio. Co- Price 15s. Sir Isaac Pitman and Sons, Ltd., Parker
ordinating chapters by acknowledged experts on dif- Street, London, W.C.2.

SHORT -WAVE CONDITIONS

Predictions for June
ME/s
40

30

20
MONTREAL
MIIIMIEMNIEMMEMIM
BUENOS

--____-_-
AIRES

- OMM JOHANNESBURG HONGKONG M Is

40

30

20

si
15
15
10 Own
8 10
8
6
111111111111111111MINE
4
5
111111111111 111=111111111MMII 111=11111111111111111111111 11111111111=1111111111111
6

30
IIMMIIMERIEMI 1111111MMIIMIMIM 4
4 8 12 16 20 0 0 4 8 12 16 20 0 0 4 8 12 16 20 0 0 4 8 12 16 20 0 3
G.M.T. G.M.T. G.M.T. G.M.T.

THE full-line curves given here indicate the highest FREQUENCY BELOW WHICH COMMUNICATION SHOULD
frequencies likely to be usable at any time of the day or BE POSSIBLE FOR 257. OF THE TOTAL TIME
night for reliable communications over four ling- distance
PREDICTED AVERAGE MAXIMUM USABLE FREQUENCY
paths from this country during June.
Broken -line curves give the highest frequencies that FREQUENCY BELOW WHICH COMMUNICATION SHOULD
will sustain a partial service throughout the same period.
BE POSSIBLE ON ALL UNDISTURBED DAYS

298
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
Live and Recorded Movie
on Electrostatic
-And Speakers
Views
G. A. BRIGGS' third " adventure in ditions and overload. Nobody knows
sound " at the Royal Festival Hall last the answer yet. Look at pickups.
month was described as a concert The early models were moving iron-
instead of a lecture- demonstration. and crystal with very crude per
His commentary, in consequence, was formance. Then came moving coils
shorter than on previous occasions, and ribbons and frequency- modula-
but his audience found it just as fully tion types, giving far superior results;
loaded with wit and wisdom. but moving irons and crystals were
This year the London Mozart improved and their position to -day is
Players under Harry Blech, Denis as strong as ever. In fact, they have
Matthews (piano), Leon Goosens already knocked out some of their
(oboe) and Campoli (violin) col- more fragile opponents. The sim-
laborated in the comparisons of plest system always wins in the long
" live" and recorded sound and run. A similar position applies to
demonstrated that realism can be microphones."
achieved in single -channel as well as Responsibility for the amplifier
Some chain and the control of balance and
stereophonic reproduction. was once again in the
single- channel sound effects included sound levels ears) of P. J. Walker,
a recording of a helicopter, which hands (and a short talk after the
caused many in the audience in- who also gave
voluntarily to look upwards. interval. Using a one -pound note as
circulating two
Moving-coil loudspeakers were a diaphragm,on and either side of the
R E48 :
used exclusively for this demonstra- coins (one improvise a source of
(Above). Heavy -duty Reflex
tion, and, as if in answer to the un- paper) tonoise of uniform intensity,
Horn speaker. Wide -area speech/
spoken thoughts of some of his " white " music diffusion; suitable for parks,
listeners, Mr. Briggs had this to say: (it is practically impossible to vary swimming galas, all large P.A.
" In view of the tumult and the the level by rubbing
harder or softer) systems.
of
shouting created by the new [elec- he showed, with the collaboration the
trostatic] speakers -or rather by the orchestra, how, by moving G7822:
those who make them and listen to source towards the ear until it could (Below). Ribbon mic. of excep-
them -I cannot let the occasion go just be heard during a fortissimo
by without a brief reference to passage, a standard of volume level
tional response and quality. For
table or stand use on all indoor
them. . . The wide response and could be carried home in the pocket locations. Satin -chrome finish and
freedom from distortion are not in and used again to set the level of the slim appearance.
dispute ... , but we are still volume control for realistic orchestral
waiting for the fanfare to die down reproduction.
and the battle to commence. This It was intimated that this might
is the position as I see it. Electro- be the last of Mr. Briggs' Festival
static speakers are coming into emonstrations. Let us hope
Hall demonstrations.
but the extent to which they will Hat de will be persuaded to give
replace moving coils will depend not many more, even at the risk of
on perfection in performance but on becoming -like
facility of manufacture and reliability
the -a
soprano
in use under various climatic con- " celebrated
farewellist."

SOUND
EQUIPMENT

THE TRIX ELECTRICAL Co. LTD.

RD.
MAPLE PLACE, TOTTENHAM COURT
LONDON, W.I

Tel. Mita. 5817. Grams. Tnzadro. Wesdo. London

Pocket "white noise generator (non -electronic.)

299
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 RANDOM RÄID)IATIONS
By " DIALLIST "

" Bournemouth Eject" in chilly weather those who gaze at

A ROMFORD reader sends me an its screen sit as close as they can E.H.T. Regulation
interesting account of a wireless to the fire, irrespective of their THE e.h.t. regulation in some
oddity, which he has christened the distance from the set. Careful television receivers is by no means
"Bournemouth Effect!" Some months investigations convince me that the as good as it might be. In some of
ordinary viewer is becoming -if them this can produce an effect that
ago he made a v.h.f. receiver for his
parents, then living on the second
floor of a seven -storey block of flats
indeed he has not already become - may be rather puzzling if you
line-unconscious. He accepts a liny haven't come across it before. The
in Bournemouth. Excellent reception television picture and doesn't notice height and width of the image are
was obtained with a dipole- plus -re- the lines any more than he notices correctly adjusted on Test Card C,
flector on the hoof. Later, when he such little trifles as sound -on- vision, with the black and white borders
was not there, his parents moved to violent ringing, or tall people who just fitting into the mask; but when
a ground -floor flat in another part of
grow short and thin people who a studio programme starts, you're
the building. This meant adding 70 grow fat as they move from one part surprised to find that the picture is
feet to the feeder; bringing it up to of the scene to another. too small and has black margins to
some 150 feet in all. Low-loss cable all four edges. The reason is this.
wasn't used. When he telephoned to Hi -Fi TV The Test Card contains a good deal
enquire how the set was working of white as well as large areas of pale
Many viewers, if not indeed the grey. Reproduction
under the new conditions, he was majority of them, don't worry over- heavy of these makes
surprised by the reply: "Oh, quite much about the kind of picture they and demands on the e.h.t. supply
well, except that you can hear every when regulation is poor there's
get, so long as they get a picture. a
train leaving the Central Station." drop (which may be quite con-
The average eye seems to be just siderable)
And so, he found, you can indeed! about as accommodating as the aver- result in the e.h.t. voltage. The
Whenever a train starts to pull out age ear and as ready to accept of that is an over -large image.
in the daytime the set " huffs and it imperfections in reproduction. But If this is fitted into the mask by
means
puffs" like a small locomotive. The more discriminating eyes and ears trols, of the height and width con-
a normal studio picture, mak-
effect is not observed at night, when do notice such shortcomings and are ing
the signal is stronger. My corre- smaller demands on the e.h.t.
offended by them. I believe that if supply,
spondent suggests that each cloud of it were more generally realized how same
won't be enlarged in the
steam sent aloft by the engines while good a 405 -line picture can be, small, way; the picture will be too
starting acts for a moment as a screen, there'd be as big a boom in high - other with black surrounds. An-
reducing signal strength to a level evil effect of poor e.h.t. regula-
fidelity television sets as there has
which puts the limiter out of action, tion is defocusing on whites. If
been in high -quality sound gear.
and so allowing a "puff " of noise to
the regulation is really bad this may
be heard. He asks whether anyone
else has experienced this effect and
invites confirmation or refutation of
his explanation. I don't think he's
ASSOCIATED
ILIFFE
' WIRELESS WORLD " PUBLICATIONS
far out, myself, but can you think ECHNICAL BOCKS
Net By
of a better one? Price Post
SECOND THOUGHTS ON RADIO
of Wireless Warn
H TEO.
RY " Cathode Ray
25/-
Line-unconsciousness? THE OSCILLOSCOPE AT WORK. A. Haas and R. W. Hallows,
M.A. (Cantab.), M.I.E.E.
26/2
... ...
AT one time I began to think that ... 15/- 15/10
WORLD F.M. TUNER. S. W. Amos, B..Sc.(H.ons.),
the insistence by the man in the A.M.I.E.E., and G. G. Johnstone, B.Sc.(Hons.)
2/- 2/3
street and his wife on bigger and RADIO LABORATORY HANDBOOK. M. G. ScroggieWIRELESS
Scroggie, B.Sc.,
bigger television screens might mean M.I.E.E. 6th Edition ... ... 2S/- 26/5
STUDIO ENGINEERING FOR SOUND BROADCASTING.
that we should eventually have to B.B.C. Engineering Training Manual by members
abandon our 405 -line system in of
B.B.C. Engineering Division. General Editor J. W. Godfrey.the
25/- 25/11
favour of either the French 819 lines INTRODUCTION TO VALVES. R. W. Hallows, M.A. (Cantab.),
or the 625 used by the other M.I.E.E., and H. K. Mllward, B.Sc., (Loud.), A.M.I.E.E. ...
RADIO INTERFERENCE SUPPRESSION: As Applied 8/6 9/2
European countries. The rooms in Radio and Television Reception. G. L. Stephens, A.M.I.E.E, to
the more recently built homes of 10/6 11 /1
to -day tend to be on the small side SHORT -WAVE RADIO AND THE IONOSPHERE.
T. W. Bennington, 2nd Edition
and I found it difficult to see how, ADVANCED THEORY OF WAVEGUIDES. L. Lewin
... 10/6 11/2
in winter time, one could sit near FOUNDATIONS OF WIRELESS.
... ... 30/- 30/10
enough to the fire to keep warm and M.I.E.E. 5th Edition M. G. Scroggie, B.Sc.,
... ... ... ... 12/6
yet far enough from a 21 -inch view- 13/6
ing screen to avoid lininess. The complete list of books is available on application.
A
answer is that in countless homes Obtainable from all leading booksellers or from
you can't. The 17 -inch or 21 -inch ILIFFE & SONS LTD., Dorset House, Stamford Street,
London, S.E,1.
television set is there all right and

300
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
extend to all the lighter parts of the
picture. But don't jump to the con-
clusion that there is anything more
amiss than a faulty e.h.t. rectifier.

The Battle of S.E. London

VERY soon both the B.B.C.'s
Crystal Palace station and that of the
I.T.A. at Croydon are due to increase
their e.r.p. to 120 kW. My heart
bleeds not only for viewers in Nor-
wood, Sydenham, Croydon and parts
adjacent, but also for harassed ser-
vicemen, who will be putting in some
pretty work on attenuators and such-
like. As soon as the new transmitter
at the Crystal Palace opened up, List No. 13.650 Colour.
though on low power, things became
more than somewhat hectic. One not
uncommon complaint by users of
convertors was that when they turned
to Band III they got the B.B.C. pic-
ture superimposed on that of I.T.A.
Just what will happen within short
Mains Voltage
range of a pair of 120 -kW stations is,
as I write, anybody's guess. Things
will, no doubt, sort themselves out
fairly quickly; and this experience will
give dealers and servicemen some
SIGNAL-LAMP
I5 WATT PIGMY -SIGN OR INDICATOR -NEON LAMPS
idea of what to expect when the per-
manent aerial tower at the Crystal * special designs to inter -services'
Palace comes into use and the e.r.p.
of both stations is raised to 200 kilo- BAYONET -CAP, MAINS- VOLTAGE specifications and specialized uses
Wide -angle brilliant glass lens are also manufactured. The
watts. Bulgin patent Multi- aspect Signal
ensures adequate viewing from all
directions. Strongly constructed, Lamp which shows a choice of
Television Interference black -crackle finished metal body three different colours through the
with louvred light trapped vents. one lens aixi the Surface- mount-
IN a recent issue of W.W. I wrote Highly polished black moulded ing model for festoon bulbs are
that I'd never seen any reference, at screw-on front bezel and 3 -hole- very popular. Amongst the
fixing bush. Solid -plunger type newest additions is the Pillar'
`
any rate in the past few years, to line base -board illuminating fitting and
B.C. bakelite lampholder with
timebase interference with sound or terminals shroud. Transparent the all Moulded range of lens in
television reception in any American lens colours:- Red, Amber, brilliant Polystyrene.
Green, Water -Clear; also Frosted -
paper, technical or lay. An East-
bourne reader tells me that he re- white. List No. D.650/Colour. * * *
members seeing a while ago an
advertisement in the Saturday Even- * * * For full technical information
on all the Bulgin components,
send for a copy of the latest cata-
ing Post of a television receiver, logue (Ref. 1951WW) price 1/-
which was described as a " Good WIDE RANGE OF SIGNAL LAMPS post free. Listed are:- Crocodile
Neighbor Set" because it was in- The Bulgin range of Signal - Clips, Fuses and Holders, Con-
capable of causing interference with lamp fittings is vast, and includes trol- knobs, Jacks and Plugs,
types to accept all lamp -caps, Mains and low voltage connectors,
other people's viewing. Well, that's including the tiny Lilliput- Edison- Resistors, Pilot Lampholders, Sig-
the sole example I've had so far. Screw type. Enclosed or open - nal Lamps, Micro Switches,
Another reader reminds me, how- bracket models with moulded Terminals, Switches, Test Prods,
ever, that the real reason for the plastic or metal bodies and bezels, Tag Strips and Group Boards, etc.
absence of line timebase interference
in the U.S.A. is that the Americans
have no long-wave broadcasting
band. Complaints of such interfer-
ence in this country come chiefly
from listeners to Droitwich on
200 kc /s; the interference is due to
* BULGIN *
the 20th harmonic of the line time -
base which falls at 202.5 kc /s. In BARKING
OF
the U.S.A. the lowest frequency
broadcasting stations are about
550 kc /s and the American line time -
base frequency is 15.75 kc /s, so the A. F. 1I1'1Ä:1\ .t CO. LTD.. HAWKING. ESSEX
lowest -order harmonic which can lines)
Rippleway 5588 (5
cause interference is about the 35th,
which is much weaker than the 20th.
301
WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com
 UNBIASE By FREE GRID
hullabaloo for more lines for both
panchromatic and monochromatic
television. The one thing I have to
criticize is his statement " The
question of more lines would prob-
Prudence, not Parsimony fore the B.B.C. followed suit in 1936. ably not be raised if we made the
There is one other point which I lines invisible by spot wobbling...."
IN the April issue I suggested that wish to clear up. Many of my Let me tell Mr. Cooper that spot
those living in remote parts where correspondents seek to draw a dis- wobbling is not everybody's cup of
they cannot get a useful signal from tinction between the meanings of tea. According to my observation,
any of the B.B.C. stations might the words " wireless " and " radio." the switch controlling the wobbler is,
solve their problem by getting the They argue that " wireless " is an as often as not, set at " off." Spot
local council to erect a fly -power all-embracing word which includes wobble has the same effect as the
relay station linked by landline to the every method of communication soft -focus lens which some profes-
B.B.C. network. without linking wires, including in- sional photographers use to get rid of
A reader writes from Aberdeen- ductive methods, while " radio "
shire to say that listeners in his implies the use of electromagnetic
village have overcome their difficul- waves. Needless to say, this is quite
ties by private enterprise. As others wrong. By common usage -an
may be able to solve their listening im-
portant thing in law-both words
problems in a somewhat similar way, mean the same.
I quote his words:-
" . . we are in a blind area so
far as the Aberdeen v.h.f. station is Diamond Jubilee
concerned and we are, therefore, I WONDER how many of you
unable to receive the signals direct. realize that the diamond jubilee of
On a nearby hill between us and wireless occurs this month. It was
Aberdeen a signal of some millivolts on June 2nd, 1896 -just 60 years
is obtainable on a simple dipole.
ago -that Marconi filed his first
We have, therefore, erected an array patent. Thirty years later Baird gave
with aerials on the Aberdeen side his first television demonstration.
which give a gain of about 14 dB. If asked in which half of this 60-
The reflecting area is of rabbit net- year period the most progress had
ting, and on our side of the reflector been made in radio communication,
we have aerials which produce usable I think many of the younger readers Suppressing the lines
signals in our village. These aerials would say the second because
are fed from those on the Aberdeen the apparatus of 1926 must seem the lines on the faces of their women
side.... " very crude to them. But to us " old sitters. This lens produces a soft
My correspondent from Aberdeen- contemptibles " of 1926 the appara- fuzzy effect, as it is intended to do.
shire seems to think I have basely tus of 1896 seemed even cruder.
impugned the financial habits of the Strangely enough, the birth of the Information Wanted
Scots by suggesting in the April famous Everyman Four " receiver
issue that co- operation on a basis of in Wireless World in 1926 seemed to RECENTLY I have been busily
voluntary subscription to defray the mark the turning point between the engaged in searching through learned
cost of this sort of thing would not stone -age and the modern periods of text books to find the answer to a
pay north of the border. It would wireless. puzzling little technical problem and
ill become me to foul my own nest, have even consulted officials of the
as the great R. L. S. once put it, as I
I, myself, would unhesitatingly say
that the greater progress was made Central Electricity Board. I have,
have Scottish blood in my veins, in the first 30 years. After all, in however, drawn a complete blank.
having once received a blood trans- the wider world of general science The problem is this.
fusion in a Glasgow hospital. no single modern invention -not If you have been living in a d.c.
Prudence, and not parsimony, was even the harnessing of nuclear area and using a universal receiver,
what I was trying to imply. My true energy -has been so great in its and the supply is changed over to
opinion of the Scottish people was effect on the progress of mankind as a.c., it won't be very long before you
given in the July, 1955, issue. the invention of the lever and the need new valves as the heaters soon
wheel. I wonder what readers of die. The phenomenon is far more
Historical Heresies Wireless World 60 years hence marked in ordinary domestic electric
will think of us and our much - lamps but whether this is due to the
MY QUERY in the April issue, fact that they are run at a higher
asking when radiotelephony was vaunted technical knowledge? Not
much, I expect. temperature I do not know. If,
invented, brought me in a large when the change -over is made, your
number of replies. Many of them valves or lamps are fairly new you
wrongly placed the invention before Lèse - Majesté will have no trouble, but if they
1888 when Hertz first demonstrated have been in use for some time and
the electromagnetic waves that had IN venturing to criticize the words have therefore become thoroughly
been mathematically predicted by of a writer on which the Editor has
Clerk Maxwell. set the seal of his approval, I feel
saturated with d.c. -if I may so
express it -then they will burn out
These historical heresies are due much as Nehemiah did when he within a few days or even hours.
solely to the habit of applying the went into the presence of his lord This phenomenon is not an im-
term radiotelephony to systems and master, King Artaxerxes, with aginary one due to faulty observa-
which employed magnetic or electro- " dangerous thoughts " troubling his tion on my part. It is freely
static induction. Such systems had mind. Nehemiah records that he felt admitted by lighting engineers but
a very limited range and were never " very sore afraid " but nevertheless nobody seems to know the
capable of development much fur- by his boldness he won the king over technical reason for it and that is
ther. They were blind -alley systems to his viewpoint and maybe I shall why I am following the example of
and therefore analogous in this res- do the same. St. Paul by appealing to the highest
pect to mechanical methods of tele- Let me say at once I am fully in authority which, in this case, is not
vision scanning which condemned agreement with what V. J. Cooper Casar but the learned technologists
in these columns over yp rs pe- said in the April issue against the who read W.W. regularly.
302 WIRELESS WORLD, JUNE 1956

www.americanradiohistory.com