.
o
.
. . . I -
'- . - . ' '.
. ~ . . , , .
. . - - - c
.. _ ‘ -. . - ,.
. . . . . -
. . ‘ 1 x
‘ u . ‘ ' . ‘ ' u
u ‘ I . . ‘ I
." - I ‘ ~— . ":
. ‘ -. -
¢ .
. . . . ‘V I ' .u ‘ .‘ ' I—.. .

 

 

 

 

 

 

   

 

 

 

 

   

   

 

 

 

   

. O . ‘ '. .
. ' . ' ‘ u
o‘ . .
_ . - o .
. o . . -_
_ - . . ‘ ~
‘ ‘ . ' . o n. . '
I ' In . ‘
' . o . . - ~ I . .
— . -
. - . . . s .-
o . . a _ - . -
. - _ I . ' - ' ‘ . .
- . o "' .
. "v I . O. n I
. . . . .. ~ - -. .. .
‘ - . n ' ' I ‘- '
I h I t
I . .i . .-' . -
t . n a. ' ' ~ - ‘1
- ' 'I‘ .
. _ _ .' . . .I . _ .-
. ~: ‘ u u. . “
. . . . . . u. ..
: ._ , . a . ° . - ‘ ‘ ~
. . . - . . . . .
. . . o -
v . . ‘ . . ~ ' . ' ' ' - '
. . ' 0 “ ~ ' . .- ~ ‘- .
. ‘ . . -_ . . ..x
x . . . 0, '. ‘ ‘ - . a
" . . . I w ‘ ,.. -V - . .
L' ., . . ‘ . . ' ' ' . ‘ °‘ ‘
u . . . I . . .
. I n . ' ' ‘ -‘ . .n ._ . . .
.- o - - ' _ 0 ' '. - o . I '
. .
- . . - ., ‘ .. ‘
.. . . . . . ~oI .
- . . ' . ‘ - ~ . ~
. . . ' a ~ ‘ ‘ .' .
. - l - ‘ .’ ' '
. o . - -. .
. , l - . ‘..
. . . a H .. -. , ‘
o . . - . . ' c .‘. - “
Q v I ‘ . . ‘ . .. 0- o _
. . ‘ " . ..- , _ . .. ' -
-- . I - ' " - .
. ‘ . .
L _ - . , ~ . . ' ‘ '
. ' . . . _ . o ‘. V
. ' I . ’ " ' I
I. I . . . -
. I . .
. C . ‘ ' . ' . . . - .<
' . _ . . - .. - _ .-.° I o .
o ' I _ . _ . . y.’ . .
- .. - ‘ - ' I . ~ ' . .
. . ' .‘ - ‘ - ‘
a- . _ 4
. . I ~ .. . .
. I ‘ I .
.‘ , . . -‘ . .
A . ' . I- v ‘
. . . .
. , , . . _v _ . - -
- _ . . ' , ‘ : a _ - ' _ .-
a . ' ’ ' 'o . ' ‘I ‘ - ‘ ' ' ‘
- ‘ O _ . . . '
. . ' o ' V ‘
, a
n - .’ . - .o . 1» . . .
. . ~ . . ,_ V
n I ~ ~
_ c . ~ . ‘ .. . --. . -
o , . a n ‘ ‘ . " ' . ‘. " ‘- . - .'
I .. ~ 0 .
. . n _ ' ’ o
' - . . ‘ ‘ ‘ .‘ A ‘ . -. ‘ ~- . .. . .
; . I - . . . ’ ’
. . . ' w . ' o . . I . '
. _ '4 '_ ' ' ‘ v - ‘ ' '
. n. - - . _
‘ ' ‘ . . .0 . . . . . . I _ _ .
' ' O ‘. . . . . -- — ‘--. .
. ‘ ' - . ‘ v . o ‘ ‘ ‘ '- . h
.0 . , - . .‘ ~ _~. - . . .._ . _ -
v . . v ‘ . . ..,
. , . . o . .. - . ..
. _ , _ . . . I. . .. I
- . . u ' v . I ' ' .‘. . ‘ '
; . . ' ‘ - .. . . - . ‘ .
. ‘ v- . ' ' 0' ' I. ' .
‘ 1 . .
‘ . . A ’. 0 . .. _ _ -- - .
,'. . . . . ‘ w . . . - .- -~ ~o . . ° '
. . . . . ., - . ‘.' '
‘ ‘ t . ' ' ‘ .o —‘ . a ‘ '
. ~ . I ‘ - ‘ . . - . _ a
'- .. . ' . n' i ’
_ - . . I - . . . ~ . . _ . I - .
! I ' c - ‘ . ' ' . ~ - h ' ‘ - ‘ - “ I . ‘
_ . . . . . . . ' - ~
I ' . - .
. . . . . - . a
_ _ , _ . . ' 9 '
n . . ' ‘ .- --
. ' . . ~ 0 .' ' . . ‘-
. . , . . :' - '.-
. .-
. 4 I o a - , ' .. . .
. . I . I > . -
. l . - . ~- .
I . . . . - . . ‘ - ' ‘
’ ' I _ I .. . . . . I . I
. . . _ . . v . . . . '.
. . ... ._
. . - . ‘ l t '
__ ~ I ' . . - . 4
. _ . - ' - . . ' ' '
. . .‘ . Q. ~ - " c . . I I - '
n . i . . ' u- _ _
> . I . a. ‘ i' . - . -
_ . . ‘~ ‘
. . ‘. _ I. I - . . .‘ . .
. . . ~ 0 . _
. ' '_ . II . . ..
. I . . . V _
‘ - - ' . ' ~ .. . .g- . ‘
- u. ’ 4 - . ‘ . n 0
__ _ . . . . ~ . . . ' -
' . - a . . 0 '
. ' . ~
_I . . . . - ‘ . .
I. . g o , ‘ , O .n. ' ' .'
- - 9.. , ' a . o
A ‘ ' , . . . - I . -
, ‘ . . _
- .
,.‘ » . . .
. . . ' ‘- . . -.. .
. . . . _ . . .
. -'- - . ‘ ‘ . . .
I n ‘ o ' - . . . . . .
. ‘ . ‘ n . ' ' . . - “ ' '-
. - ‘ I >
. . o o ’
.0 4- ' . . . I . .
‘ I . " v ‘ ' . _. . ' A
I . n . . - ’ ‘ . ‘ . '
I . '. . . . _ .' _ - ... .
. . . - - . o '
.. . ~ . . , .. - . . . . .
3. x a . . E - . .
v- ' ‘ - .' ‘ - -‘ ‘ — ' ' .‘
‘ . I a . A - . .
o ' V , n ' - c c‘ Q
. . a r
- ' v 1
. o . . - . . . , I _
o
. ' ~
‘ . . .. . .
. . o
- . . . . ' 0 ' . . C on
. O . ' . .
." I ' I . ' ‘ a . ' . , '
I . < . _ - ._ '
. - . ‘ . * ~ ’.
. ' I . .
7 a . . .
c -- . - . o . _
. ' v . h . - ‘ . .‘ '
. . . I . , . . - . - 3
; I . - - . .‘ . ‘ I . . K.
‘ I ‘ . . a
. . . '
° ' ' .' . ' ‘ . '
- O n o . . . .
n . . ' . . ~
. ' . v . 1
u . ‘ ' V
. I . . . ' - . . .'
. . ' .
aI - . . . o . - a
‘ I v ‘ I _ I‘ . ' -
. 3r - . a ' ' ‘ 9 "
0.. ‘ . - . . -
. ._. , -o - ‘ ‘ ~ .
. _ _ ¢
. , . .
I. . . . .
- .,‘ . - . ' . ' ' - ‘ .‘
. . ~ ‘0 . -
' . \ . ' . I . . a -
. '. ‘-. . _ . -
- . . u
, , I . , .’ ' . ‘ -
. .
I -
_ I u . - . I I -
n r G ' .
.- v ‘. .
. . v o . ‘ _ .

 

 

 

 

WIlfllllllllllllflllllllIlllHlillllllilflllHllllHllliHUI
3 01627 9725

3129

 

LIBRARY
Michigan Stat; 4

Universviy

 

 

ELIMINATION OF INTERFERENCE FROM RADIO RECEPTION BY
THE DIFFERENTIAL OR BALANCING OUT METHOD

A Thesis Submitted to
The Faculty of

MICHIGAN STATE COLLEGE
of
AGRICULTURE AND APPLIED SCIENCE

BY

Stuart W. Seeley Fred Pacholke

Candidates for the Degree of

Bachelor of Science
in

Electrical Engineering

June 1925

TH ESXS

ELIMINATION OF INTERFERENCE FROM RADIO RECEPTION BY
THE DIFFERENTIAL OR BALANCING OUT METHOD.

Recent and marked improvements in the art of radio and
particularly radio telephone broadcasting have perfected the
transmission of signals to such an extent that about the only
room for improvement today is the elimination of interference
from the received signals. Work has been done in this connect-
ion but no decided results have as yet been attained. Evidence
of endeavor along this line is given by numerous and varied
patents, files of Which have been obtained for reference in
connection with this work.

The methods employed for the elimination of static and
other interfering impulses as described herein, While entirely
original with us, were found, after much of the work had been
done, to have been at least partially covered by'a patent
filed in July, 1920 and issued to Mr. Lester K. Jones of
New York City. A brief description of the underlying prin-
ciples of this method together with the history of its in-
ceptioh.may well be given here.

Any separation.of one type of Object or thing from another
or other types of things must of necessity depend upon some
fundamental difference of that type. When a child is told to
sort out the blue marbles from.an assortment which he is given
he is able to distinguish them by their color which is common
to them alone. When a farmer threshes his grain.he takes ad-
vantage of the fundamental difference in density of the chaff
and the kernels and is thus able to separate the two. Likewise

the elimination or separation of static impulses from signal

10: 770

‘9

'0“

v
on:
”.41

.
..'I

I
I.-

hr,

0-.

R

(I)
I

‘

II)

(I)

-2-
impulses must take advantage of some difference in.the two

or some characteristic in.the one which does not exist in the
other. The static impulses which interfere with a received
signal travel via the same mdeium, are of the same form, have
the same wave length and may possibly come from the same dir-
ection as the signal impulses with which they interfere. It
is true that they are not grouped in such a way as to form
intelligible communication but it is also obvious that the
only device capable of distinguishing this difference would
be the human.ear which is not what we are after.

Certain authors, notable Elmer E. Bucher in "PRACTICAL
WIRELESS TELEGRAPHY“, maintain that at least a portion of the
so-called static heard in a radio receiver is caused by the
passing to ground, through the circuits of the receiver, of
static charges of electricity induced upon.the aerial by the
potential gradient of the atmosphere. If this were the case
the effect of these charges upon.the receiver could be elimin-
ated by supplying a grounding circuit for the aerial which
would include a radio frequency filter. Such a circuit would
carry off the induced aerial charges only but would not
affect the reception of signals. The fallacy of the electro-
static origin of interference can, however, be easily proven
by merely inserting a set of high resistance telephone re-
ceivers or phones in series with.an serial and ground at some
time when.the interference is known.to be bad. If the poten-
tial gradient of the atmosphere were changing rapidly enough
to cause surges in.the aerial circuit sufficient to interfere

with received radio signals,they most certainly would affect

-3-
the phones and cause audible sounds. The most that can
ever be heard under such conditions is a slight 60 cycle hum
due to potentials induced on the aerial by nearby power lines.
It is probably from this outworn conception of the cause of
interference that it was termed "static".

Several months ago, before this investigation.was
started, while listening to two radio receivers, simultaneous-
ly, which were tuned to different wave lengths, it was dis-
covered that the static impulses from‘both were, as near as
the car could detect, identical. This suggested the possi-
bility of eliminating the interference by the use of a differ-
ential transformer having three windings: one connected to
a receiver detecting both the signal and the interfering im-
pulses, one connected to a receiver detecting the interfer-
ence alone and the third connected to a reproducing or ampli-
fying device. This, in short, is what we have termed "The
Differential Method of Static Elimination”, and is the sub-
ject of our investigation. Our attempts to put this theory
into practice are described herewith.

Rather crude attempts along this line had repeatedly
been.unavailing before this research was started and, there-
fore, we tried to formulate reasons for failure and guard
against them, even before we started. We have applied certain
principles and theories as follows:

The fundamental difference between signal and interfer-
ing impulses lies in the fact that the fbrmer’may'be detected
with a receiver tuned to but one wave length or frequency
While the interference apparently contains waves of several

different lengths simultaneously. Thus two receiving mechanisms

-4-
tuned to intercept two different frequencies will both be
actuated by a single interference transient. Further it
is believed that the time-intensity contour of the potenr
tials produced in the rectifying circuits of both receivers
will be nearly identical.

Consider now for a moment the probable origin of an
interfering impulse. Let us say a stroke of lightning
takes place at a point 100 miles from.an Observer with
two receivers tuned as described above. Modern research
workers maintain.that a lightning discharge is uni-direc~
tional and could not, therefore, of itself produce radio
frequency waves. However in fihe neighborhood of the dis-
turbance there most certainly will be oscillatory surges
set up in trees etc., which will be highly damped. These
surges might be likened to a rubber ball drOpped from.a
high place onto a smooth surface. If it were drOpped from
a height of fifty feet its period of bound would at first
be about four seconds but as the height to which it bounded
grew less and less, the period would grow shorter. Likewise
the surges cussed by the electrical discharge as they are
damped out emit shorter and shorter waves. Thus it is
evident that waves that will affect both of the observers'
receivers may be emitted from a single highly damped
oscillator Which in turn is actuated by a unidirectional
electrical discharge. 0n.the other hand, waves emitted by
a transmitting device from.a free oscillator, with little or
no damping, all have practically the same wave length and

frequency of succession and can only affect one of the ob-

server's mechanisms- that one which is.tuned sharply to the

-5-
transmitted frequency. It now only remains for the observer
to connect the output of the two receivers to two windings
of a differential transformer in such a way that equal im-
pulses in'both windings produce no effect in a third, at
the same time any unbalance in the surges in the two re-
produces itself accurately.

Obviously, for a device of this type to Operate success-
fully the potentials produced in the balanced windings must
be identical and must produce fluz.changes in the core of
the transformer exactly 180 degrees out of'phase with each
other. For this to be the case it is necessary that both
receiving mechanisms be identical in their electrical
characteristics and that the transients affect both simul-
taneously. Return fer a moment to the analogy of the rubber
ball. It is oovious that the shorter periods of vibration
follow the longer ones by a certain definite interval of
time, depending 0n.the difference in.period length and the
damping of the oscillations. The greater the difference in
period length, the greater the time interval and the greater
the damping the less the time interval.

It would thus seem that if our hypothesis concerning
the origin of interfering transients was correct there would
always be a certain phase displacement between the potentials
they produced in the rectifying circuits of the two receivers,
those in the circuit tuned to the lower frequencies taking the
lead. Our results showed that this condition was not serious
enough to cause trouble. It is conceivable that enough peri-
odic circuits of a natural origin in.the neighborhood could

be set in oscillation so that waves of almost any frequency

-6- _
would be emitted simultaneously with the discharge.

Our first laboratory experiment was to determine how
nearly the same two separate and identical amplifying
devices could be made to reproduce in amplified form.the
same applied impulse.

For this experiment two identical two-step vacuum tube
amplifiers were built. The wiring and the position of the
instruments were duplicated exactly and the correSponding
vacuum tubes in.the two were of as near’the same character-
istics as could be obtained. The input for both amplifiers
was supplied from a vacuum tube detector which in turn
received its energy from an outside antenna and one-step of
vacuum.tube radio frequency amplification. The primary
windings of the two first-step amplifying transformers were
connected in parallel rather than in series to secure a
more exact duplication, a series connection resulting in the
capacitance of one amplifying transformer winding being
shunted across the other.

For this test signals were picked up from Station WEAR,
the local College Broadcasting station. The filament temp-
erature 0n.the radio frequency amplifying vacuum tube was
decreased to a point such that the detector and amplifier
tubes would not be overloaded. The output of each amplifier
was impressed on an element of a Westinghouse Portable
Oscillograph. Step-down transformers with a ratio of 20 to
1 were used between the plate circuits of the amplifiers and
the oxcillograph elements. It is interesting to note here
that at the time this experiment_was run, station.WKAR was
broadcastihg a talk from the college gymnasium by Edgar Guest,

 

 

Q b\¥.\\fisxv\mck. w NQKRQ

 

 

 

 

 

mlhl'lahlthlh—
k A
3 L

W'I'I' WWW—-

h: E

 

 

Wk kaxwxutx QKUIMQE UMQ

 

 

 

 

«at out w%
“text «Stag.

 

bk 35330 L

 

'Mul'N—t

 

th M

J

Wuhldthhhlh
WM"

 

 

 

k.\_

 

\eK QM

 

 

‘IKS

 

W

in!

 

 

 

Em

 

0H

 

 

-8-

 

Osoillogram recording the output of the two duplicate

uplifiars o

-9-
.Michigan's well known poet, and the oscillograms on.page 8
are of his voice.

A pair of telephone receivers connected to the output
of either amplifier indicated, as near as the ear could
detect, equal amplitude and equal clarity and no difference
in tonal qualities was apparent and yet an examination.of
the two traces shows decided differences in wave form; so
much so, in.fact, that it is hard to conceive of’the ear as
interpreting them alike. A close examination shows a pre-
ponderance of the lower frequencies in the upper trace.

A difference of this kind could only be due to a difference
in the inductance and capacity balances in.the two ampli-
fiers and probably came about through slight differences in
the several transformers. Attempts to counteract the
differences by the addition of either inductance or capacity
at the output of either transformer only made matters worse.
The error would probably have to be corrected at its source.

The fact that two amplifiers gave such dissimiliar
results after having been constructed as near'alike as we
possibly knew how, discouraged further attempts at neutral-
ization after the impulses had been.amplified. The elements
of the osoillograph were not sensitive enough to register
the output of the detectors without amplificatioh.and it was
therefore hoped that the amplifiers could be used so that
the process of neutralization could.be recorded on an
Oscillogram, one element recording the potentials in each
of'the three windings of the differential transformer. Such
an oscillogram would have registered any difference in.the

phase relationship of the static impulses on the two different

H-:
-

-10-

1wave lengths. However, since this method proved so un-
reliable it had to be abandoned with the h0pe that some

time, with more accurate equipment it might still be success-
ful.

Our next experiment was with the circuit shown.on the
following page. Two identical detector tubes were connected
so as to receive their energy from the same antenna at
different wave lengths. The output of each tube was connected
to one winding of a differential transformer as shown and the
output of the transfbrmer amplified by'a two-step vacuum tube
amplifier. Here, again, there was a chance that the impulses
might be distorted by the detector tubes in such a way that
they would not be similar in the windings of the transformer
even though they were alike before they entered the detector
circuit. However, there was less chance for dis—similar dis-
tortion here than in.the two amplifiers. It was heped that
if phase displacement existed it might be possible to correct
it, in.a measure, by inserting the proper value of resistance
in one of the plate circuits to act as a phase shifter in
combination with the inductance of the transformer.

Several types of transformers were used and the results
compared quantitatively by means of a pair of head phones
connected at the output of the amplifier. A Western Electric
Repeater Coil such as is used in phantom telephone circuits
seemed to give the best results. Static elimination was
never absolute,however, and, furthermore, the interval of
time required to change from.ane transformer to another made
it difficult to remember about what the average intensity

of interference was. If the static had been constant in

 

 

K “\th m< m.
Vdemckc 3% L at «bxanKRxT
.23» as.“ shim wQ Vb QQQQQ

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

t. Swish“. curs \mkaficd

 

1

+ I
__._._____I

 

 

 

 

 

 

KS

 

 

 

 

 

 

 

 

-12-

intensity and continuous it could have been measured on
the ground glass of the oscillograph and comparison made
in this way.

While the detector vacuum tubes were both connected
in a manner to allow regeneration, coupling of the plate and
grid circuits was kept at a low value. This was because of
the well known fact that excessive regeneration causes dis-
tortion of signals.

The two oscillograms 0n.the following page were made
with this circuit. The duration of each is a trifle over
one-half of one second as may be determined from.the 60
cycle sine wave at the top. The bottom trace was made by
the element actuated by the output of the amplifier. During
the instant that the oscillograph Shutter was open the
filament and grid return circuit of one of the detectors
was either Opened or closed in such a way that one part of
the picture would show the interference, unbalanced, and
the other with it balanced out. Just how difficult it is to
open or close a circuit exactly at the middle of'a one-half
second interval, by hand, may be imagined. Neverbthe-less,
portions of balanced and unbalanced interference appear on
these two traces sufficiently to show the results cbtained.
In.the upper oscillogram the interference appears on the
left hand one-quarter of the trace and in.the lower one it
is at the extreme right. The remainder of'the two traces
shows the effect of‘balancing out the interference. We can
truthfully say that our efforts in this experiment were as
successful as could possibly be h0ped for.

The two oscillograms on the fellowing page were made

-13-

 

 

Oscillograms with part of trace recording interference

unbalanced and part with neutralization taking place.

 

 

 

 

 

-14-
with the same circuit but at a much greater speed. The

60 cycle sine wave shows about two complete cycles or one
thirtieth of a second. Artificial interference was set up
by means of a small induction coil and battery. The upper
trace shows the interference unbalanced While the lower
trace was taken.with the balancing mechanism in.Operation.
Careful examination of the upper trace shows that the vibra-
tor rate was about 300 per second and that the natural spark
tone rate was about 1400 cycles per second. The latter is
probably the natural period of the secondary winding of the

induction coil. w

 

These two traces do not clearly illustrate the marked
difference noticed in the amount of interference When listen-
ing with head phones both.with it balanced and unbalanced. It
is possible that when these traces were taken the filament
rheostat of one of the tubes had been disturbed slightly and
neutralization was not as complete as it might have been.
Neverathe-less, the lower trace shows quite a decided decrease
in interference over the upper trace.

In summing up; first with regard to the theory'of static
elimination. We believe it to be fundamentally a problem in
the separation of two very nearly identical things. It appears
that there is but one basic difference in these two on which
to work and that is the difference in.damping or decrement
. of the useful and interfering wave trains. Granting this,
there is but one way to be rid of static and that is by detect-
ing it on a slightly different wave length than that used for
communication, and applying it in the form of Opposing
potentials, either directly or through the medium of magnetic

 

 

 

I -,__.

Oscillogram of artificial interference.

 

 

 

 

 

Oscillogram with artificial interference partially neutralized.

 

4..., «1...... .2. ., .n

.

 

-15-

flux in the core of a transformer, to counteract those
potentials due to static which were detected 0n.the communi-
cation frequency.

We have found that it is advantageous and even.im-
perative, when.modern commercial apparatus is used, to produce
neutralization in the first circuits in.which it could possibly
take place in order to avoid distortion of the impulses. This
will be the power circuit of the rectifying or detecting de-
vice in most connections.

Our results have shown that this method of eliminating
interference is entirely practical and is not in.the Beast
complicated.

Previous failures have undoubtedly been due to lack of,
absolute balance in the duplicate circuits or to too much
complication about them. The Latter is an affliction of
Mr. Jones' patent ( No. 1,471,165 ) which was mentioned above.

f'lICHIGQN STQTE UNIV LIBRQPI ISE

I”! H HHHI

293016279/25