THE omen AND CONSTRUCTION or A
UNIVERSAL-WAVE DOUBLE
SUPERHETERODYNE 8mm RECEIVER

THESIS FER THE DEGREE OF M. 8.
Harold joseph McCarvey
1933

37...... ‘. ,. .

...ot
....

 

S
E
m
a
U
E
T
A
T
8
m
w
M

31293 017721162

 

-_-_-‘t M

.‘--

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

DATE DUE DATE DUE DATE DUE

 

DESIGN LED CONSTRUCTION OF A UNIVERSLL'WLVE
DOUBLE SUPERHETERODYM RADIO RECEIVER

thesis for Degree of LS.
Harold JJaGervoy
1933

Aoknowledg ments

I am greatly indebted. to Mr. Herold Riggs for the
suggestion of the problem,to Professor 0.L.Snow and
Professor O.W.Chapman for helpful suggestions and the
securing of parts and materiels,and to the whole
Physics Department for advice and help whenever 6t was

needed.
Herold J.M¢Garvey

“1.023?" 3f)

Contents

Introduction

Reasons for Choosing this Project
General Theory of Operation
Obtaining Selectivity

Drawing showing Operation
Oscillators and.Harmonics

Selection of Parts

Experimental work and Construction
Diagram of Iecuum Tube voltmeter
High Frequency Intermediate

Second Intermediate Frequency Amplifier
Shot Effect

Curves of Shot Effect
FinaerdJustments andChanges
Limitations and Conclusions
Diagrams of the Receiver

List of Parts and the value of Each
Photographs of the Receiver
Bibliography

Page

~1wa

14
17
20
21
23
25
28
29
32

36
39
42

Introduction

The object of this work has been an attempt to design
and build a universal wave radio receiver which would be
stmple both in construction and operation.

This report gives the complete results of the design
and experimental work upon such a receiver, Which can cover
a range of’l5,000 K.C. to about 150 K.C. without the chang-
ing of a single circuit therein.

There are several limitations which must be recognized.
These are covered in the report, and suggestions are givm
as to the solution of some of them.

The receiver was built with the idea of using commer-
cial parts and tubes as much as possible, and all these
are Operated within their correct limits of current and
voltage. Since the completion of the receiver, several
new tubes have been made available, making it possible to
substitute one tube in place of two in a number of cases;
therefore in any future study of this project, it may be
seen that changes may be made in the types of vibes used.

Reasons For Choosing This Project

The design of any universal wave radio receiver
brings up problems.that are not involved in the design.of
a receiver for the reception of broadcast prograns only.

In the design of a receiver for the broadcast band, it is
only necessary to provide tuned circuits that can cover a
range of 540 K.C. to 1500 K.C., which is a frequency range
of about 1 to 5. This range is not large, but vhen any
attanpt is made to increase it to any extent serious dif-
ficulties are encountered.

The problem is tint radio receivers are tuned by
varying the capacity of a tuned circuit in which the fre-
quency varies inversely as the square root of the capacity.
This means that to have a frequency range of l to 3 it is
necessary to have a capacity ratio of l to 9. When the dis~.
tributed fixed capacity of the circuit is considered it
will be seen that a greater range of frequency is quite hm-
possible without also altering the inductmce of the circuit.

If the extending of the frequency of any receiver is
considered, the most obvious solution, therefore, is the
changing of the inductances or the tuning coils. This prac-
tice gives good results; but if the range is to be made
very large, it becomes necessary to have complicated switch-
es and coil combinations. Commercial receivers upon the
market at the present time give all wave reception, but it
is necessary to have at least four combinations of coils
and sometimes five to cover from 540 LC. to about 15,000
K.C.

3.

The earnercial receivers tint are called all wave
do not cover all the waves, because in most cases there is
no consideration of the 1013 waves. The long waves are
those longer than the broadcast band. These waves do not
ordinarily have any of the broadcast prgrams, but they
include almost all the shipping and most of the commercial
news as it is given from one station to another. The
weather reports can always be picked up every few minutes
on the longer waves as they are give: to airplanes. It
might be stated that the longer waves do not have very
much use in this country to tin ordinary person vho does
not know the code, but still there is enough phone work
to interest may in this type of reception.

There is some chance that in the years to come com-
mercial programs may be broadcast over the longer waves
on account of their greater range of direct covering
without fading. At the present time the radio receivers
that are shipped to Europe must cover a range of 20 to
2,000 meters. This type at receiver calls for very can-
plicated coils and switches.

‘ It can now be seen why some thought has been given
to the problen of designing a receiver which can cover
all the wave lengths and still be as simple to operate

as one built for broadcast reception.

General Theory of Operation 4"

Since in this receiver we prepose to avoid the chang-
ing of coils or condensers, a different method is re-
quired.

If a razge of from 20 meters to about 2,000 meters
is desired, the necessary range of frequency is from
16,000 K.C. to 160 K.C. For purposes of design such a
range as this might be considered to be from 15,000 K.C.
to zero.

Having decided upon such a frequency range, we must
think next of how to cover it with commercial coils am
condensers. Accordingly, we are forced to adopt some
form of high frequency circuit in which it will be pos-
sible to get a wide rarge of frequency. A high frequen-
cy oscillator, built in tune from 15,000 K.C. to 50,003
K.C. will give the required range. The problem then is
one of making proper use of such a high frequency oscil-
lator and its wide range of coverage.

The only way to make use of such an oscillator in
a radio receiver is in make it a superheterodyne, thus
making possible the use of the full range of the oscil-
later.

The superheterodyne receiver is very different from
other types of receivers. The crdirary set is tuned to
the frequency to be receired, but in the siperheterodyne,
the frequency we want is in effect changed to some other
frequency by means of an oscillator. The method by which
this frequency change is effected is as follows. If two
different frequencies are mixed together under the right
conditions, the aim of the two frequencies, the differ-

ence of the two frequencies, and also the two original

D.
frequencies will be present in the resultant. If the

receiver is designed to amplify at one frequency only,

it is possible to have an oscillator that can form
combinations of frequency such trat almost any frequency
can be received. It must be remembered furthermore that
oscillators do not always give out pure sine waves of
energy. They may radiate large number of harmonics. These
hamonics are also radio frequencies, and they too can

mix with the signal frequency and therefore it is desirable
that the oscillator fundamental and its harmonics be con-
trolled to such an extent as to avoid undesirable combi-
nations of frequency.

It is the presence of the second harmonic that makes
it impossible to have a range of oscillator frequency
greater than 1 to 2, because when the oscillator is at
the lower frequency, the second harmonic is givirg the
same result at some higher frequency as the oscillator
tuned to the higher frequency. This would cause double
tuning on some stations.

Accordingly, an oscillator is chose: that will tune
from 15,000 K.C. to 30,000 K.C. with which to cover the
desired range of frequency and yet not have a ratio of
frequency greater than 1 to 2. This range is the lowest
frequency possible which would meet all the requirements.

Now let us consider a: illustration of Just how sich
an oscillator may work in the reception of a station at
1000 K.C. If we assume 15,000 K.C. as our fixed intermedi-
ate frequency and if the oscillator is adjusted to 16,000
LC. we find that the difference between the signal and on!
local oscillator is 15,000 K.C.

6.

If we wish to receive a signal of 10,000 K.C. it is
only necessary to tune the local oscillator to 25,000 LC.
and again the difference is 15,000 K.C.

This Shows how such a receiver might work, but actual-
ly it is not so simple; because most stations are only 10
LC. apart and when a station is tuned in on' 1000 K.O. the
intermediate frequency is 15,000 K.C., but if some powerful
station is at 1010 LC. its signal will be changed to 15,000
LG. by the oscillator. This means that the intermediate
amplifier must receive 15,000 K.C. and reject 15,010 LO.
This is impossible with coils and condensers.

Obtaining Selectivity 7'

The problem of selectivity of a receiver built for
general purpose work is not an easy one. Selectivity is
wanted, but a receiver must not tune so sharply that the
side band frequencies are cut off to any extent. With
some receivers the cutting of the side bands is very pos-
sible, but receivers for broadcast programs should not cut
any of the side bands because quality is thereby impared.

The selectivity of our receiver must be such as will
give good results on all the broadcasting stations, and for
this reason an attempt hast be made to obtain what my be
called average selectivity, thus retaining most of the side
band frequencies which in turn makes average good selectiv-
ity and qmlity possible. Selectivity to wiihin 10 K.C.
cannot be obtained whm we are working with 15,000 LC. un-
less a Quartz crystal is used as a tuned circuit in a band
pass filter, which would cost too much, hence we are forced
to use mother method.

The method adopted is to effect a second change in fre-
quency and this time to a much lower, then our 10 K.C. be-
comes a much greater percent of the total frequency. In the
application of this sch one of lowering tin frequency we are
limited only by the appearance of so called image effect.

This image effect may best be explained by an illustra-
tion as follows. If we want a signal of 1000 K.C. and the
intennediate frequency is 100 K.C. it is only necessary to
have a local oscillator of 1100 LC. and the difference be-
tween the oscillatcr and the signal is the correct frequen-
cy, but if for some reason a station should be on 1200 32.0.

it can be seen that the difference in this case also is the

8.
intermediate frequency of 100 LC. The problem is to pro-

vide circuits ahead of the oscillator and mixer to prevmt
such signals from caning through. This is a problu in any
sips rheterodyne design.

If we are conddering a 100 K.O. second intermediate
frequency it will be see: at once tint a second local oscil-
lator of 14,900 LO. must be used. The difference in the
frequencies of the first intermediate and the second oscil-
lator will give the 100 K.C. for the second intermediate,
but if the first intermediate is not selective mough to
keep out signals of 14,800 LC. thw too cm come through
and combine with the oscillator frequency an! form the last
intermediate frequency, which is the imge effect.

It will “be see: that a low freqmncy last intemediate
would give very good results, for selectivity, but it is
necessary to prevent signals from coming in on the image
frequency. In the case considered above it can be seat that
the high frequency intemediate stage must be very selective
to prevent image. In fact it would have to be better than
we can build.

The solution of this problem is to use a higher frequen-
cy last intermediate am use good coils am a larger number
of circuits to obtain selectivity.

If the limit of tuned circuits is three or four and cr-
dinary coils are used we cannot allow this last intermediate
to go much above 1000 LC. because it will be impossible to
obtain the desired selectivity.

Let us now consider the selectivity at 1000 K.C. It is
the sen frequency as the broadcast stations. This means

that we can obtain the same degree of selectivity as is ob-

9.
tained in tuned radio frequency radio receivers of two or
three stages of gain, or the use of three or four tuned
circuits.

If we mnsider the probability of image trouble at
the last intermediate frequency of 1000 31.0. we find that
if the same high frequency first intennediate is used the
oscillatcr must be tuned to 14,000 32.0. This means that
the image frequency to cause trouble would be 15,000 K.C.
The difference between the 15,000 K.C. and the 15,000 K.C.
signal is a rather small percent, but from the experimen-
tal standpoint it has been found to be very satisfactory.

The 1000 K.C. last intermediate may not be the beat
frequency to use, but if the frequency is much less the
image effect would be had am if the frequency is much
higher selectivity would be poor.

The picking of the frequencies for this receiver is
the same problem as in any other receiver. It is a problem
of cut and try. When something is gained in one way there
is almost always a loss in some other.

Another problem mich must be solved is the rejection
of any signals of the acne frequency as the first intermed-
iate, because tiny would be coming in all the time through
the untuned grid circuit of the first tube. This rejection
is done by He use of a series tuned circuit between the
grid of the first tube and ground.

It has been found tint automobile ignition system are
the greatest source of trouble at this big: freqxency. The
spark discharge in the spark plug of the automobiles is a
danped oscillation am thus it is a rather effective in

cans irg trouble .

O

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/“ -
\— u
l
‘ - l ‘ ‘ L
1‘ ‘ ' ' v $ .. e_
e'. s -, '
’ l "V ‘L I ~ 1

’ ' t - - ( . Ii
L - "a. _ - .
“ ‘ ~ jCJ.
:V‘ " h; C t? L 4. ‘ L &'
I“ _ - ‘57 in“ :j, C" 1 1 : i
H - «fist uctrctor,
I ' z”’ 10 ‘LL‘L’I.

.1 \

 

 

 

 

1"
.LL‘ka

“lver.

 

1*? L '
.' H J.
_ -_.._e,
I t :1} ‘4 K,
. 1 .
1
h. o
t I. .
C
L- ‘ - \ ;(‘5 \ V 2 ‘ ‘I‘
A _ . ' ’ > ‘
.3. i :1 1' t." 1. I \, L .(L
CLLY; magi. *J. l;0 L

k1 . (.e_

ll.
Oscillators and Harmonics

Since the receiver is a double superheterodyne with
two oscillators both of which are at rather high frequen-
cies there are problsns in their design that are not ordi-
narily encountered. One of these problems requirixg care-
ful consideration is the effect of mmonic frequencies,
or multiples of the fundanmtal. frequencies of the oscilla-
tor. These harmonics can cause conbinations of frequency
which, if in the first or second intennediate stages, would
act the sane as a station except they produce only noise,
much like the noise of a carrier frequency, hence all har-
monies are to be kept out if possible. In this discussion
there is to be no reference mde to the first hannonic be-
came the fundsmmtal is considered to be the first harmonic.

The first oscillator is variable and the second is
fixed. The one fixed oscillator helps some in eliminating
harmonics because then tube voltages can be adJusted for
best results and ttnn left at that setting.

The first oscillator must tune from 15,000 K.C. to
30,000 K.O. This is rather a high frequency and a high
voltage output at such a frequency is hard to obtain. The
total number of tubes being limited, only one can be used
for each oscillator. Since the standard screen grid tube
is good at the high frequencies it is used for the first
oscillator. The voltage output of the oscillator should,

for best results be about 10 volts. This may be regarded
* This is give: as the correct voltage upon a mixer tube

in a superheterodyne by the 3-10 Technical handbook of
The Radio Corporation of America.

12
as a rather high voltage for such a high frequency, and
the only way to obtain it is to use a very good quality
tuned circuit, one having a relatively small amount of
resistance, and operating the tube at its highest safe
value of current an! voltage. The plate current is usual-
1y the limiting factor in the output of snall tubes arch
as are used in this receiver.

The hamonic output of the second oscillator is kept
down as much as possible by the use of high capacity tuned
circuits. Low capacity in the first oscillator however
is not required, because the harmonics of one oscillator
cannot cause trouble unless they can combine with lemon--
ice of mother. ,

In building mess oscillators the values of the in-
ductances were kept very low, in fact for the fixed oscil-
lator tin inductance of the coil is about .4 microhmry
and the capacity about 250 micro microfarads. Such a com-
bination of inductance anl capacity is not common, but in
our work the limit of capacity is the point where the volt-
age output of the oscillator is just enough to properly
excite the mixer tube.

Another way to prevmt the hannonics of an oscillat-
or from causing trouble is to use a pick-up coil in the
tuned circuit of the oscillator, then any voltage in this
coil will have less harmonic content.* The voltage induced
in this pick-up coil is used to excite the mixer tube.

The frequency stability of the two oscillators is an-
other problem to- be considered, if the frequency of either

* Prevention of Hamonics Radio News. April 1, 1953.

13.
oscillator varies much the set will not be stable in
operation, and because these frequencies are so high a
very small percentage change would cover quite a range of
radio stations on a basis of a 10 K.O. separation of
stations.

The adding of capacity to the tunirg circuit helps
to stabilize the frequency, because a snall charge in
capacity in the tube will not make much change in fre-
quency.* The use of a high quality circuit with a snall
amount of resistance also helps the frequency stability,
because the circuit has larger circulating currents and
is therefore not effective as much by out side changes

and inf]. uences.

* Frequency Stability, from Principles of Radio by Henney.

l4.
Selection of Parts

The selection of parts for this receiver involves
mostly the selection of tubes only, because most of the
parts are standard or else are so special that they must
be made up particularly for this receiver.

The selection of tubes is made according to the pur-
pose for which they are to be used. The oscillators which
we will consider first, must Operate at rather high fre-
quencies. This means that to get good Operation, it is
almost necessary to use screen grid tunes on account of
their greater stability in high frequency circuits. 1 There
are several types to choose from and several different
tubes were actually tried out as oscillators. The 224A
tube worked very well as also did the 267 tube. The 224A
tube being an older tube and easily obtainable at every
store, was chosen for these tubes V2 and V4,

The selection of the mixer tubes V1 and 13 was a
different problm, because practically my tube will work
in this position. Screen grid pentode tubes type 258
were however selected for this position, because they
function equally well under a great many different con-
ditions.

The next part of the set consisting of the second in-
termediate frequency amplifier tubes V5 and Y5, and the
last detector tube Y7 can also use almost any type of tube,
but here again some work better than others. A screen
grid tube is better in the amplifier so the choice was
between.several, but‘egain.au older type was used. The

final choice being between the 224A and the 235A. The

15.
224A was used because of its lower plate current and

much less shot effect.

The last detector tube V7 is a triode because there
are times When headphones might be used in a receiver of
this type and a triode is a good tube to use with.headphones,
although its gain as a detector is less than a screen grid
tube operating under the same conditions.

The last tube in this receiver is a 247 pentode power
tube. This tube does away with the audio stages and gives
very high output. Its qiality may not be as good as a tri-
ode, but by using it less tubes are required, which.is a
big consideration.

Since the building of this receiver several new radio
tubes have been brought out, sane of vhich could be used.
There is a frequency changer that could replace our pres-
ent system of oscillators and mixers. This would reduce
the total number of tubes by two. The tube is known as
number 2A7.

There are a few other parts that should have special
mention. Condensers are one of the most important itens.
The one tuning condenser must be offsolid construction
and the spacing of the plates must be large so they will
not cause trouble by vibrations, which would cause the
set to drift from one station to another.

All'the fixed condensers must have lOW'tenperature
coefficients or they“will cause drifting of the set..All
the condensers used in the tuning of the high frequency
circuits are air or mica and all the mica condmsers are
sealed in tight containers so moisture will have less

effect.

16.

The coils used in the last intermediate are canmer-
cial tuning coils, tuned with a trimming condaiser. The
coils L5 and L4 in the first intermediate are made up of
four turns of No. 14 copper wire on an inch form and are
not commercially obtainable.

The power transformer and filter chokes and filter
condmsers are all standard. The shields, bypass condens-
ers, sockets, resistances, and all small parts are stand-

ard quality parts obtained from radio supply houses.

17.
Experimental Work and Constmction.

The experimental work on this receiver was tin most
interesting, because much Of it was new and if trouble was
experienced it was an interestirg problem to overcome it.
The first experiments were upon the oscillators. The real
work with them was to increase their voltage output to a
point vhere they would sufficieitly excite the mixer tubes.
The tubes tried were the 27, 6'7, 224A, 255A, and the 6?.
The 27 womed very well up to about 10 meters, but it would
fail to go much beyond this value. The operation on longer
wave lengths was very good.

The operation of the 67 was much the same as the 27 ex-
cept that its operation was some better, but it too, was
limited at the high frequencies. The 224A worked very well
to about 4 or 5 meters. The operation of the 255A was poor
in the ease way. Its current drain was high and its Opera-
tion at the higher frequencies was much like the 27, not
very good. The Operation of the 5'! was very good, but lit-
tle different than the 224A, so the 224A was take: as the
best tube to use, then further experiments were carried on
to increase the output of the oscillator tube.

The tubes were Operated all the time airing this ex-
perimental work with meters in the plate circuits so the
safe current would not be exceeded. The final Operating
voltages for the 2243. as an Oscillator was 300 volts plate,
100 volts screen grid, 50,000 ohm grid leak, plate current
of 3.6 milliamperee, am screen grid current of 1.5 milliam-
peres.

The output of the Oscillator was checked at every

18.
change in its Operating conditions by a vacuum tube volt-

meter. This voltmeter was Of standard construction and

its diagram is given here to indicate its Operation. The
voltmeter as used in this work measured the peak voltages,
because it is usially the peak voltage that is the most im-
portant in radio wont.

Experimental work was also done on the inductance of
the oscillator. The first tests were run with coils made
of NO. 14 copper wire. The results were Just fair. The ‘
size Of the wire was increased and the voltage output was
increased, but when the wire was increased above 1/8 inch
the voltage did not increase to any extent. Good qmlity
soft cOpper tubing of 1/8 inch outside diameter was final-
ly used and the results were good.

The only type of circuit tried for the oscillator was
the electron coupled Hartley. In this circuit there is a
grid leak in the grid circuit which connects to one end of
the inductance; the other end Of the inductance is ground-
ed. A tap is made on the inductance to form the cathode
portion of the coil. The best position for the tap was
found to be 1/4 the coil length distance from the ground
and. The portion of the coil between the tap and the ground
acts just the same as a plate coil would in any other meth-
Od Of connection.

A very interesting point was brought out in this work
by the use of mica condensers. The first oscillator is tuned
by an air condaiser, but the second beirg at one fixed fre-
quency all the time, is tuned by a mica condenser. The out-
put VOItage of the circuit tint had the mica condenser was
about 75% of the output voltage or the circl it tuned with

a very good quality air condenser.

The greatest output obtained with an air condenser
and tubing inductance was 15 volts peak measured across
the inductance. Ihen the mica condenser was used the out-
put was decreased to about 12 volts peak, and at the same
time there was a slight increase in the current drain Of
the oscillator tube. This check was made at the same fre-
quency and with the sane tube am inductance.

 

 

 

LJ""‘|'H 4:;

0;} ’2 e

rVud;
ja“va.
Ci 1' Fe

581V r

till

ZC‘L.

'

 

‘ 0

~

C cter 51:1 r

83"

meter U‘ai: rLzLS

 

1 93

 

 

 

 

 

 

 

h, c

—

JI

M
B

 

 

 

‘1

 

 

C

we7(z»~1 7143i

Icr iw' L. 336
C l. l 1. (j C .7
u

a d r" (*1 I
~ IJ Ll

urod

thtl

JOY]: t3}. 8
l,JOC,L?0 OI;
ic,ooo Uh;-
1c,000 1

6535

1.0 hf.

:vL t e ULl the :wuflz‘vo;ta e to be

6 to E V its
2 volts

25 v Its

a I 1 }-—++ iii/L-

VdZTflflF777?

LC. 10.1 tic \N‘lt.t-~=r
Jc_» had until the
‘ .811 (2:5! mushrmvm voltage

.x. 9. ‘ar , '
.iicc31u , «_ a.

,7
u

of the

130 type ratio tube

@alwunoneter.

V

LLLLfi R9 uxiil tLe

:010 current. ;e volt otcr

11‘? any.

'eicr

reaswrcd.

81
The High Frequency Intermediate

The problem of the first intermediate which must
Operate at about 80 meters was a hard one,beceuse it is
new and data as to the value of any parts could not be
obtained. The first work was done with a single coil
tuned with a variable mica condenser. It was found that
a circuit could be made quite selective,if the inductance
was low and the capacity high. All the tests were made
with a plate tuned circuit and the grid Of the next tube
connected through a small condenser and a high resistance
to ground.

The capacity used in the circuit was about 175 Hiero-
microfarads; the inductance was made up Of four turns of
NO. 14 cOpper wire,eelf supported,end wound so the diameter
of the coil was one inch. The inductance was about .35
microhenry. This arrangement gave very good resulte,but
when the whole receiver was tested it was found that the
single tuned circuit would pass enough signals in the
broadcast band to cause whistles in the Operation of the
receiver.

The amount of energy going through the circuit at
broadcast frequency was too much,so the only solution was
to change to a double tuned circuit. This was tried with
a coil in the plate circuit of one tube and a coil in the
grid circuit of the next tube. The values of inductance
and capacity were kept at about the same values as were
used in the first arrangement.

The orperimental work necessary to arrange the coils
so they would not tune double was considerable,and the

only method was to cut and try. The solution was found

a- - -xe‘n ...... re. .4 -L-...|. 1 -..n '1 lo a..-u.

\ \ x w u 22 x
placed at right angles to each other and one coil offset
so that it overlapped the other only to the extent of
1/4 inch; The coupling with this arrangement was found
to be about correct. Even with this arrangement a certain
amount of energy would feed thorough,but by changing the
last intermediate frequency a little,pratically all of

this trouble was eliminated.

s s s s -. , . 23
Second Intermediate Frequency Amplifier

Since the second intermediate frequency amplifier
Operates an the broadcast band at 1000 K.C., the selection
of coils for it was not hard. The coils used have 145
turns of Lita wire equal to number 26 and the forn.is
7/8 inch diameter. The coils are placed in the plate
circuit of the tube and coupled to the next tube through
a condenser and a high resistance to ground. Each coil was
tuned with a trimming condenser which had a capacity of
80 micro mierofarads. The full capacity of the condenser
in each case was used. The inductance of each coil was
about 240 microhenry.

Tests were made at first with simple transformer
couplingibut the total gain was not enough to give
satisfactory results. The plate tuned circuit gives very
good results for gain,and in this instance the added
capacity of the plate does not cause any trouble,because
we are not trying to cover a band of frequencies.

When the set was completed it was tested with three
coils in.the second intermediate amplifier,but the gain
and selectivity was not good enough,so an extra tube was
added. The results were good,but there were too many tubes,
so an attempt was made to use only two tubes for gain.
This study of reducing the number of tubes led into
some rather interesting details. When the number of
tubes was reduced,and these pushed harder to get the
necessary gain it was found that the noise was increased
in the output Olen though the total number of tunes was
reduced. After considerable experimenting it was found

that the total noise was not increased if the number of

\. ‘h \ ‘a Va . - . 24
tubes was increased ewen more. This discovery let to

much thought as to why an increase in the number of
tubes should decrease the total noise in the output
when each tube added ordinarily adds a certain amount
Of noise to the total. After much experimenting,and
study this noise was attributed to shot effect,which is
discussed in the next part of this report.

The selectivity of this last intermediate was found
to be inadiquate,but the adding Of coils or more stages
could not be considered. The lowering of the frequency
would solve the problem,but again this would not be
possible because the first intermediate was already at
the limit Of image separation. One Of the possible
methods left to effect an improvement was some form
Of regeneration.and accordingly it was tried.

The introduction of regeneration has the effect Of
decreasing the resistance of the circuit to which it is
added. This decreasing of the resistance of the circuit
increases the selectivity. This process can be carried
on however until all the advantage is lost. In this
receiver regeneration.was added by putting a small
capacity 023 between the plates of the last amplifier
tubes V5 and 77.

. q 85
Shot Effect

The shot effect is known as noise that is generated
in the tube itself. It is caused by the unevenflow of
electrons from the cathode to the plate. The plate
current is quite regular in flow,but it is made up of
electrons and the number released from the cathode one
instant,may not be the same as the number the next instant.
This irregular flow of current is what causes the noise
called shot effect.

when the flow Of current from a cathedeis space
charge limited,it means that there is always an over
supply of electrons at the cathode,but that they cannot
get away. When the flow of current is temperature
limited,it means that all the electrons being released
at any one instant are being drawn.away and that the
only way to increase the number is to drife more out
which usually means increasing the temperature.

Experiments carried on in the past have shown that
when the plate current in temperature limited the noise
of shot effect is the greatest and the least when the
cirrent is space charge limited. '

The problem then in this receiver is,to always make
the current space charge limited. This means a high
grid biasing voltage on the tubes so that the plate
current will not be great. This shows why noise was not
so bad with more stages,because each tube was not pushed
as hard,thus the plate current was not great.

The problem then is to get gain without large plate

‘ Noise generation within radio Receivers,By R.DeGOla
Presented betore I.R.E. From Radio EKG. Aug. 1931.

26
currents.,Here the literature on the subject was consulted
and it was found that the noise level of different tubes
is very different. During the first work on the receiver
the type 245 tube was used in the first stages and accord-
ing to Rinaldo DeCola the noise level Of this tube is
much greater then the 224A. *

The 224A tube was then used in place of the 235 and
the results were better,but still not good enough.

Then after considerable study and experimenting the
volume control was taken off the last amplifier tube and
it was connected up for full gain all the time. The first
amplifier tube was the only one used for the control of
the gain.mThis method worked out very well. The total
noise was reduced,because the first tube was Operating
most of the time with rather low plate current as the
the volume control was usually not advanced to the full
extent. We might expect the shot effect in the last
tube to be high,but the signal reaching the last tube
would be strong,so the noise would not be increased to
any extent.

The manner in which this works out can be shown by
a little illustration. If the total gain necessary for
a certain station is 1000 and the last tube gives a
gain Of lOO,the first tube must give a gain Of lO,but
6f the two tubes are controlled together the gain of
each would be only about 31.5. This means that the first
tube must give three times the gain with the standard
method as when.only one tube is controlled. This reasoning
applies very well except when the maximum gain of both

* Noise generation.Within Radio Receivers, R.DeCOla.

. 2?
tubes must be used. The curve taken from.an article on
shot effect shows what shot effect amounts to in volts
for different tubes under different conditions. The
reduction of noise level in a radio receiver is a
problem in itself and was teens so to be in this

receiver.

 

rum

 

 

 

C no >
2: .5 cm. 2

61.» 33>:
muse) {wallow Q>\Q

2w diva.
or am

4|‘l

0m

 

 

 

«meow .6 a

a m . :3 wet
verdQfiflu‘ \lxuflfmkimw

Emu: 31m

.3 new
misuse
sinus

 

VOLTS

 

NOISE

‘ 89
Final Adjustments and Changes

The final adjustments of a receiver of this type is
a most important and exacting Operation. The range of
the receiver was considered tO be from zero frequency
to 15,000 K.C. This range should be extended to include
stations below 20 meters,and so in the final adjustment
the range was extended down to 18 meters.

The first adjustment was for the purpose Of getting
the correct range for the first oscillator. The final
range was 8 to 16 meters or 17,500 K.c. to 18,750 K.G.
The next adjustment was the tuning Of the high frequency
intermediate to 18,750 K.C. The last intermediate frequency
is 1000 K.C. so the last oscillaor was adjusted to 17,750
K.C. This then means that the highest frequency that the
receiver can receive is 17,750 K.c.,because when this
frequency is reached it will pick up its own oscillator.
The reason the second oscillator is at a lower frequency
than the first intermediate is that if it was higher the
first oscillator would tune at one time to the same
frequency as the second oscillator and there would be
trouble as the shielding is not good enough to permit
two oscillators to tune to the same frequency and not
make beats which would be picked up. The two oscillators
would also tend to fall into step when their frequencies
were close.

The design and construction of these“ high frequency
circuits is rather critical and cannot be made very
accurately so when the set is put into service it is
necessary to make several adjustments to get good

Operation. The fact that the last intermediate frequency

g 30
is in the broadcast frequency band was also a source of
trouble. It was necessary to make some rather close
adjustments in this part of the receiver so that it would
not tune to some powerful station and be picking up
signals from it all the time. Even with the best of
shielding it was found that the set would pick up
enough voltage from some stations to cause trouble part
of the time.

The adjustment of the first intermediate for the
correct amount of coupling was an experimental problem.
The only way found was to adjust the circuits and coupl-
ing so as to make each coil Lgand L4 have a single
point Of resonance,then increases the coupling until
this condiion was lost,and finally decrease the coupling
a little until the tuning was best.

The adjustments of the filter condenser 01at the
input was rather simple. Some station was tuned in,
then some local noise or frequency was noticed and thai
the condenser was adjusted until this local disturbance
was decreased to a minimum. When the correct setting has
been reached the filter acts as a low resistance path to
ground for any noise that was of the same frequency as
the first intermediate.

Some changes were made in the receiver after it
was completed. The 235 tubes were replacedby 224A tubes.
This reduced the noise. The number of turns on the
Oscillator pick up coils were reduced to provide a good
output voltage at all frequencies. mixing of the signals
was tried on the suppressor grids Of the 258 tubes,but

this failed. There were several little changes made in

51
the voltages to different parts that were not recorded,

but the final values of all voltages are given in the

table Of constants for the receiver.

. . . . 38
Limitations of The Receiver

In.the building of something a little different
there are always new limitations that may be placed
upon it. The building of this receiver has brought up
limitations that are very new to commercial receivers.
A different kind Of trouble in this receiver has been
encountered, caused by the heat of the set changing the
temperature Of the condensers and their dielectrics so
that their capacity is altered. When the receiver has
been in Operation for an hour or more this trouble
disapears,but it is quite seriour during the first
30 minutes or more. A drift of 40 K.G. has been Observed
between the time the receiver was first turned on
and when.it became stable. The solution to this problem
is a study Of the condensers to find out where changes
might be made to overcome the trouble. It iSa difficulty
that has been avoided commercially, where the best of
construction has been followed,and such high frequency
circuits are not attempted.

The receiver has been found to be about average
in selectivity. It could be improved by the use of a
lower frequency last intermediate. This would mean that
the selectivity of the first intermediate would have
to be increased. This however could be done by the
addition Of another stage at the high frequency.

The harmonics Of the two oscillators have been a
source of trouble even though an effort was made to
eliminate them. However harmonics cause trouble only
on four or five points over the whole range,so they
are not very serious.

The receiver is rather noisy in Operation. This

. . . . 33
noise could be reduced by the addition of another

high frequency stage. More work on the oscillators with
the idea of increasing the voltage output would also
help in the reduction of noise.

One undesirable feature of this receiver,but one
which does not mean so much now days is the large
number of tubes necessary for its Operation. The parts
that go with the tubes are however simple and inexpensive.

A study of the receiver by an unbiased person
would of course bring out different points of advantage
and disadvantage and limitations and Operating character-
istics. The above list would doubtless be enlarged,if
a complete study was made by such a person.

The work so far has shown very well that such a
receiver will operate and that the coverage of all the
wavelengths is possible. Whether improvements can be
made so the receiver will be thoroughly satisfactory for
home Operation can only be told as more experimental
work is done upon it,incorporating the suggestions

indicated.

 

_/ l.

 

t»? sh KQ Etmwmxa

M:

 

 

 

 

 

GU

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

WOLF.
s .T .ifl
«I
lfllL P
. m as .
L V
+.Om. - a
H t J
+3 % Huh
v! r“ .J 4 ”ya-II—

 

 

 

 

 

 

 

 

 

,,7|||||II‘||I1

4
2Q

\VQ

NC

\E: l

Ql

>>

a: Q“

as L

3Q.

 

 

.. o E
_

p
:5:
.u.

u H
uflfluuuudd,
Judge
33. T
mod flaw

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ozyce
C4, 06
05,07

08

Parts and Constants 36
.000025 microfarad condenser used to tune the
tune the filter circuit. A standard trimmer.
.000125 microfarad condenser..d general Radio
double spaced transmitting condenser.

1/10 microfarad by pass condenser.

.000 25 microfarad mica grid condenser.

.000175 microfarad mica condenser,constructed so
its value can be cahnged. A DeJnr Amsco Corp.
product.

.0003 microfarad mica condenser with a small

pilot air condenser in parallel.

010.013, 031. .00008 microfarad mica trimmer.

011,014,015.0008 microfarad bypass condenser.

012.018
O15

017

619.020

1/10 microfarad by pass condensers.

8 microfarad filter condenser,with a .25microfarad
by pass condenser in parallel.

16 microfarad filter condenser.

.0005 microfarad mica condenser use for tone
tone correction and radio frequency bypass.

25 microfarad by pass condenser.
A loop of wire between the two platewire to
produce feed back.

12 turns of no.26 wire on a 1 and 1/8 inch form.
5 turns of 1/8 inch cOpper tubing on a 1 andl/4
inch diameter. Self supported,and a tap one turn

from the ground end.

. . 37
L3,L4 4 turns Of No.14 cOpper wire self supported

diameter land 1/8 inches.
L5 4 turns Of 1/8 inch copper tubing self supported
tap one turn from ground end, and l and 1/8 inch

diameter.
L5,L9,L10 150 turns of Lita on 7/8 inch form.
L7 1 andl/Z turns of No.30 wire on a land 1/8 inch

form. Placed very close to the main coil.

19 8 and 1/2 turns Of No.30 wire on a 1 indh fern
placed vary close to the main coil.

L11,L13 Filter chokes.

T1 Audio transformer.
T2 Jefferson power transformer.
31 50,000 Ohms.

22,34 2,000 Ohms.
R3,85 50,000 Ohms.
Ra'neiaio 1,000,000 Ohms

R7 500 Ohms.

39 0-7,500 Ohms.

311 30,000 Ohms.

R12 500 Ohms.

313 18,000 Ohms tapped voltage divider.

'1,Vg 858 tubes.

v2; v4,v5,vg 2241 tubes.
v7 27 tube .

'8 247 tubes.

'9 280 tube;

 

 

 

 

 

 

Bibliography

Henney, :iKeith

Henneyg Keith

Ramsey, R.R.

De Cola, Rinaldo

DOW. JeBe

Smith; ‘eWe

R.C.A.

Messing, Edgar

Hirsch: Charles

Radio Engineering handbook.
Principles of Radio.
Fundamentals Of Radio.

Noise generation.Within Radio Receivers
Presented before I.R.E.
From.tugust, Radio Engineering; 1931.

Temperature Effects on Oscillators.
Q.S.T. Mayil933.

Electrical Measurements in Theory and
application.

Technical Handbook.R~10.

.Acoustic feedback in Superheterodyne
Receivers. hem March,l93z
Radio Engineering.

Design of R.F. Band pass filters.
From, Radio Engineering, Sept. 1929.

 

   

“@1111? W; ”Mini!" Mini iii “

 

1162