I ITITI III ‘I TIT
TI TITIIIII‘I .II‘
TI T T'

 

ELECTRON MULTIPLI‘ER INVESTIGATIONS
THESIS FOR THE DEGREE OF ELECTRICAL ENGINEER
WARREN HERBERT BLESS
I 9 3 9

Inlot V|rc ‘(.11u.l ,- rps.\ LEI

.. no .Ivu.~uq:r. N fluid?

wk {11

I;

.1 1:3:(u D u .....

 

 

.IA-l¢£I.tl.L‘q.; I... ..u .1

A...»

» f‘vfiLIJ. “rut v.
T

ELECTRON MULTIPLIER INVESTIGATIONS

A Thesis Submitted To

The Graduate Council of

Michigan State College
of

Agriculture and Applied Science

by

Warren Herbert Bliss
~

Candidate for the Degree
of

Electrical Engineer

June, 1939.

 

' .J
T
7» up sum“.
3.

IHESTS

 

Electron Multinlier Investigations

 

Table of Contents

References Sheet 1
Pages of subject Matter Sheets 2 to 21 inclusive
Figures Sheets A to L inclusive

Photographs .; R-318
R—319
3-320
{-332
3-333
3-334
3-344

121475

References

 

l.

2.

3.

The Secondary Emission Multiplier - A New Electron
Device. V. K. Zworykin, G. A. Morton, and L. Halter.
Tarch 1936. Issue of Proceedings of the Institute of
Padio Engineers.

secondary Emission Electron multiplier. An article
in the November, 1935 Issue of Electronics.

Electrostatic Focussing in Secondary Emission nul—
tipliers. Engineering Report TP-369, RCA Victor Div-
ision, by J. Hajchman and E. L. Pike.

l - Svnqps1s

1.1 - Purpose of the Investigation

 

The investigation which is outlined in this report
was carried out with the following threefold object as a
guide: ~
(A) To obtain a working knovledge of light controlled,
secondary emission, electron multiplier tubes.

(B) To determine the practicability of vacuum tubes
of this type then employed with suitable circuits and
auxiliary efiuipment to generate high Speed facsimile
test waveforms.

(C) To consider a few possible scanner arrangements
and briefly outline for reduction to practice those

that are found to be suitable. ’

1.? - General Procedure

 

A ten stage, photoelectric type, secondary emission
electron multiplier tube was obtained from the tadiotron
Division at Harrison and a test "tea—Wagon" was built up
to house the eeuipmcnt which included a television receiver
type power pack obtained frcm the Victor Division at Gam-
den and a 1500 volt rectifier. A special light source,
lens system, and chopper disc assembly was constructed for
supplying light to the tube.

Static and dynamic characteristics were determined,
the latter tests being made prircipally at 60 and 12,000
cycles per second.

A few tests were made to determine the feasibility
of introducing an a-c carrier frequency into the electron
multiplier so that a light modulated a—c wave mould re—

sult in the output. This was tried electrostatically
and magnetically.

The electron multiplier output was also used in a
balanced modulator to modulate a 75 kc carrier. In con—
nection with this a special compensating or counter emf
circuit was used to improve the linearity of the modu—
lator. The re-occurring phenomena generated by the
Specially prepared chOpper disc was used successfully
to test facsimile terminal facilities at keying speeds
up to 12,000 cycles per second.

2.

_ 3 _

1.3 - General Statement of Findings

 

On the basis of information obtained to date the electron
multiplier has a very definite promise of providing at least
one scheme for satisfactorily scanning subject matter at
"page a minute" speed. The apDaratus arrangement to be de—
scribed provides an extremely flexible tool for use in the
investigation of high Speed facsimile terminal and radio cir-
cuit requirements.

Recommendations

 

Experience with the electron multiplier test assembly de—
scribed in this report indicates that it can be successfully em—
ployed to generate accurately controlled high speed facsimile test
phenomena. It is accordingly recommended for studying terminal
and radio circuit requirements.

It is also recommended that the study of the general per-
formance characteristics of various electron multiplier modifica—
tions be continued in cooperation with the Victor and Radiotron
divisions of RCA—Manufacturing.

Also, as time permits, that detailed consideration be
given the following:

(a) Introduction of carrier ac into the multiplier by the
electrostatic method.

(b) Determination of the minimum number of multiplier stages
or minimum total voltage necessary to give sufficient output for
scanner applications.

(0) Further deveIOpment of a suitable modulator, probably a
balanced type.

(d) If theoretical considerations and preliminary experimental
work indicate the desirability, the construction of a high speed
scanner using a suitable electron multiplier tube as a pick—up unit.

Detailed Discussion of Investigation
3.1 - Statement of Problem

For the study of terminal and radio circuit reouirements
for the successful handling and transmission of high speed
facsimile signals it was desirable to have a test apparatus
or device which would produce such signals or waveforms of
voltage and current. Ordinary vacuum tube oscillators do not
readily produce these and hig~ speed scanners were still in
the process of deveIOpment.

3.2 -

- 4 _

Furthermore, there was still the desirability of determining
a suitable scanner to be used in a facsimile system capable of
handling a letter size sheet per minute. In present speeds of
facsimile operation there is barely enough output from the
photocell to operate circuits of minimum acceptability. Higher
speed operation requires improved reaponse, especially a
linear frequency characteristic covering a wider band. Higher
outputs are also desirable, since, with these, some of the well
known types of modulators could be controlled directly.

The photoelectric type, secondary emission electron multi—
plier appeared to be a unit which would supply hoth of the needs
outlined above. Hence, it was decided to make a preliminary in—
vestigation of the characteristics and some possible applications
of multipliers.

Resume of Electron Multiplier Theory

 

The construction,.theory and operation of secondary emission
multipliers has been described at length in Reference 1. A brief
treatment will be given here to familiarize the reader with the
functioning of the magnetic, photoelectric type.

Fig. l of Sheet A shows the arrangement of electrodes and
resistors inside the tube and also the usual external connections
for a 10 stage experimental tube.

The electrodes are divided into two general groups ——————
accelerating plates and emitters. Electrodes b, d, f, h, j, l,
n, p, r and t belong to the former group and a, c, e, g, i, k, m,
o, q and s to the latter. v is the collector and u, the screen.
In the tube used in this investigation the first 6 accelerating
electrodes were made of screen so that light could be focused on
any one of the first 6 emitters.

The multiplying action in the tube is a result of the ratio
of secondary emission electrons to bombarding electrons on the
emitters being greater than unity. Light, falling on the first
emitter a, releases electrons by photoemission (the emitters are
capable of photoelectric emission as well as secondary emission).
These electrons are attracted by and accelerated toward electrode
b which is at a higher potential than a. Except for the presence
of a magnetic field whose lines of force are at right angles to
the plane of the sketch in Fig. 1, sheet A, the electrons would
go to plate b. However, the combined influence of the electro-
static and magnetic fields causes the electron path to be curved
as shown in the sketch so that the electrons arrive at plate c
which is at the same potential as b. The electron velocity at
this bombarding point is sufficient to cause appreciable second-
ary emission.

A ratio of secondary electrons to primary electrons of 3
per stage can readily be obtained in a 10 stage multiplier. The
increased number of electrons emitted by plate c is acted on by
the next accelerating electrode and the magnetic field and then
bombards plate 3. In this manner the electron streams are
amplified as they pass down the tube until the last plate v collects
the final emission from s.

The distribution of voltage on the numerous electrodes is
accomplished by means of a tapped resistance bleeder. To avoid
having too many terminals on the tube the bleeder resistors for
the first few stages are Inounted inside the tube envelope. The
output current may be passed through a resistor PL where an emf
proportional to the light will be develOped.

3.3 - The Physical Setup
3.31 - General

Fig. l of Photograph P-319 shows the general arrange-
ment of the electron multiplier test setup. This shows
the top compartment of the "tea-wagon" with the light
source and chOpper disc assembly. This compartment is
26 inches wide by 17 inches deep by 15 inches high. The
front and top, which are on hinges so they may be turned
up and back out of the way, are equipped with interlocking
switches so that the power supplies (1500 volts and 6300
volts) cannot be turned tn while the compartment is open.
The tOp, front and right hand and are made of transparent
material so that meters may be observed from outside. The
bleeder or voltage divider resistor bank is mounted in
the back of the COMpartment and the output meter and inner
tube compartment, toward the front.

Fig. 2 of Photograph D~320 is an inside view of the
inner tube compartment taken from above. The field
magnet is shown pulled back to bring the tube into view.
A rotatable base is controlled by an external knob so that
the tube may be turned about its axis to bring it into
proper relation with the magnetic field. Slots through
both compartments provide an unobstructed path for the
light beam.

PhotOgraph R-344 and Fig. l'of Photograph P—32O show
the light source assembly. This unit was made to simulate
the reflected light produced in an actual scanner. It was
designed to comply with this in three respects:

3.32 -

-6-

lst, same aperture distortion, 2nd, same light quantity,
and 3rd, same top frequency of light fluctuation for page
a minute speed. The assembly consists of a lamp house,
10 volt lamp, a lens barrel, and a motor driven chopper
disc.

Details of Light Source Assembly.

 

Fig. 2 of Sheet A gives the details of the Optical
system. The light is supplied by a 50 watt, 10 volt
lamp 1 which is operated at 8.5 volts. The first unit in
the lens barrel is a condenser lens 2. From this the
light passes through two screens 3 and A. The first is
one of fixed density which was chosen so that the light
output of the syssem felling an plate 11 would be the same
as that reflected from a conventional scanner drum fir '
white subject matter. This value is about 6.4 x 10-
lumen. Screen 4 is a stepped, density wedge slide, the
tOp of which may be seen in Photograph R-344 projecting
upward to the right of the lamp house.

The next important part of the optical system is the
aperture 5. Two of these were used; the first one was
0.025 inches square and the second, 0.012 inches high by
0.025 inches long. These values, as will be explained
later gave the desired aperture distortion.

Lens 6 focuses the image of aperture 5 on the chopper
disc 8. A mask 7, having a small round hole through its
center, cuts out most of the stray light arriving at this
section. Lens 10 picks up the light as it comes through
the chOpper disc and brings it to a 5 millimeter spot on
the photoelectric cathode ll of the tube which is to re-
ceive the light.

A close inspection of Fig. 1, Photograph R—320 reveals
the details of the chopper disc which is driven at 1800
RPM by a 60 cycle, synchronous motor. Two alternate 90
degree segments of the disc are solid. The other two
alternate 90 degree segments are divided radially into six
different bands.‘ The outer or first band is cut away so
that when this position is used 60 cycles per second square
wave chOpping is done. The second band has 100 teeth and
100 slots, each 0.032 inches wide, in each of the two 90
degree segments. This band, then, produces 12,000 cycles
per second chapping. The next four bands produce, reSpecte
ively, A impulses, 3 impulses, 2 impulses, and 1 impulse
of 12,000 cycles. The disc drive motor is mounted on a

- 7 _

sliding base so that any one of the six bands may be
brought into the light beam at the proper point.

Chomper Frequency Considerations and Aperture Distortion
For a 10 inch page per minute scanning the

Screw Advance 3 .12 3 0.1667 inches/sec.
60

For standard line advance of 120 lines ner inch the
Screw Advance = 0.1667 x 120 = 20 lines/sec.

For a standard drum circumference of 9.22 inches the
Linear Scanning Speed = 20 x 9.22 = 184.4 inches/sec.

The stems and bars of letters of 8 point type are
0.0082 inches wide. Taxing twice this value or 0.0164
inches as the length of a cycle in scanning such letters,
the following is obtained.

Maximum Fundsmental Scanning Frequency for Page a
Minute Speed

= l84.4/0.0164 = 11,244 cycl s/second.

Hence, the chOpner disc was designed to simulate 12,000
cycle scanning, which is a slightly greater requirement
than that actually expected.

In the standard AA-914 scanner the light sDot as it
strikes the drum surface is 0.006 inches by 0.008 inches.
The first dimension is that measured parallel to the
direction of motion of the drum surface at the scanning
point. Hence, the period of aperture distortion in
scanning a letter bar cycle ( = 0.0164 inches) is given as

Aperture distortion 3 0.006 = 0.366 cycle
0.0164
2 0.366 x l 3 32.5 x 10'6 sec.

32.5 microseconds
for page a minute scanning.

In designing the chOpper disc a diameter of 8 inches
was selected. Too large a disc would require a large

motor and too small a disc would require teeth and slots
too small to be made accurately. with the 8 inch disc

the teeth were made witha 1/32 inch cutter and spaced
about 1/16 inch apart (center to center). This made

the finished teeth and slots each practically 0.032 inches
wide.

For the same aperture distortion as discussed above
the light spot should have a

Width = 0.366 x l = 0.0228 inch
16

The aperture 5 (Fig. 2, sheet A) was actually made
0.025 inches souare 1.hich gave slightly more distortion.

The oscillograms shown on Photograph sheet R—332 were
made with this aperture. Since the duration of the flat top
Of the wave was very short another aperture having a width
Of 0.012 inch was made. The oscillographic results of this
are shown on Photogra1h sheet 3-333.

3.4 - Static Characteristics

Figs. 1 and 2 on sheet B are sketches Of the circuits used
injeternining some Of the static characteristics of the electron
multiplier and on sheets C and D are the curves resulting from
these tests.

The double, heavy voltage divider system shown in the
circuit diagrams was used to ins e sood voltage regulation at
all the multiplier terminals. Practically all the tests were made
with a total potential of 1500 volts which was diviied into 130
volts per stage and 200 volts for the collector or anode. With this
condition and zero resistance in the anode circuit the data for the
resonance and linearity characteristics shown on sheet C were taken.

Fig. 1 shows the relationship betwe een the anode current and the
magnetic field strength for the major resonance peak n. This curve is
similar to one shown in reference 1 and shows the nece:3sitv Of holding
the field coil current constant. It was observed that operation of
the tube in either side of the resonance pea: caused appreciable
noise in the output res ulting from small irregularities in the high
voltage or field supplies. The most stable Operation, as well as the
maximum output, was Obtained when the conditions were 8 dch that the
tube "as performing at the tOp of the resznance hump. The amount of
light used in this test was about 5 x 10 lumen.

 

 

_ 9 _

Fig. 2 of sheet 0 is a linearity check on the multiplier
for small quantities of light. A light density wedge was cali-
brated by means of a standard type 913 photocell with a compar—
atively large amount of light and then 'he wedge was used to
check the multiplier with about 5 x 10- lumen maximun. No data
was taken for large anounts of light since the saturation point

was believed to be well above one millianpere of output current.

Figs. 1 and 2 on sheets B and D show the test circuits and
results of variation of load resistance. Fig. l on each sheet
is the case for normal operation with the load resistor in the
anode circuit. with this condition the output terminal at ground
potential is the plus lead. The circuit was also arranged as in
Fig. 2, sheet B, so that output voltage of reversed
polarity with respect to ground could be obtained. In this case
the terminal at ground potential is the negative lead. Both
of these schemes were tried because either polarity, with respect
to ground, mar be desirable depending on the type of modulation
to be produced. It should be noted that there may be objections
to the second scheme since the variation in IR drop across .
the load resistor alters the potential of the last emitter and
the accelerating electrode connected to it. However, the curves
for output voltage are similar in shape, and in practice the
load resistance may be kept low enough to limit the output to
a few volts so that this alteration of electrode potentials may

not be Objectionable.

The two tests discussed above were conducted with 6.4 x 10'4
lumen of light, which value, as mentioned previously, is equivalent
to that reflected from white subject matter on a standard facsimile
drum. This va ue was determined by measuring the current in a type
918 phototube when receiving the reflected scanner light. The
light from the light chopper lens system was then directed into
the same cell and a light screen selected to pronuce the same cell
current. It was assumed that discrepancies due to color differences
in the two sources could be neglected. The numerical value given
above results from the rated sensitivity of a 918 cell (110 microamp./
lumen) and the measured cell current (0.07 microamps.)

-3-5 - Dynamic Characteristics (Experimental)

Photographs R-332 and R—333 are the results of a series
01? tests on the electron multiplier when operating with "chOpped"
itight. The circuit of Fig. 1, sheet B, was used to Obtain all of
tilese. For the oscillograms en R—332 the aperture distortion was
322 microseconds for the 60 and 12,000 cycle tracings and 43 micro—
Senzonds for the 8000 cycles while the quantity of light for all
W843 about 10"3 lumen. For those on R-333 the aperture distortion
was 16 microseconds and the uuantity of light, 6.4 x 10-4 lumen.
Thrase two groups of oscillograms were obtained by contact ex—
lasc screen of the RCA type THV-l22-D

L)

3x>£fiure prints on the g
°S<3fillosc0pe, whose vertical sensitivity was 90 volts per inch.

-10...

In considering sheet H—332 the effect of load resistance and
capacitance on the wave form is not very evident except for the 60
cycle souare waves. The influence of these factors on magnitude
is quite prominent, however. Referring to Figs. 1 and 5 or to
Figs. 2 and 6, the output voltage is seen to be preportional to
the load resistance as would be expected. In Fig. 5 the output
voltage oeaks are 70 volts. A comparison of Figs. 2 and 4 shows
a decrease in magnitude due to the current shunting effect of a
small condenser in parallel with the load resistance. The aperture
distortion was so great for these cases that the waves appear to
be almost sinusoidal.

The 8000 cycle oscillograms were obtained by driving the
chopper motor from a 40 cycle mechanical fork supply. The ten—
dency for flat tOp on the waves is evident in Fig. 1. The theor-
etical duration of flat top in this case is 1/8 of a cycle. In
spite of the efforts to reduce the stray capacitance to an ab-
solute minimum the corners of the waves are still slightly rounded.

Fig. 7 shows the quality of wave form for 60 cycle square
wave with an aperture distortion of 32 microseconds. This wave is
perfect for most practical purposes. Fig. 8 shows that the shunt
or stray capacitance can be as great as 3.005 mfd. before this
waveform is altered appreciably. I

‘ The waveforms on sheet R-333 give a better indication of the
dynamic performance of the multiplier. With the aperture distortion
reduced to 16 microseconds the flat—tOpped nature of the waves is
very pronounced. Figs. 9, 11 and 13 show the effect of varying the
load resistance; With values of 100,000 ohms or higher the flat top
is distorted. For a 50,000 ohm load the waveform is good and the
peak value of voltage is 22 volts; this is from 6.4 x 10‘4 lumen of
light. Reducing the load resistance below 50,000 ohms improves the
wave shape slightly but decreases the output voltage directly.

Figs. 10, 12 and 14 show the influence of shunt capacitance.
The first and last of these were taken with a load circuit time
constant of 10 microseconds which distorts the waveshape appre—
ciably. The wave of Fig. 12, which was made with a time constant of
5 microseconds, might be regarded as the limit of allowable distortion
but even in this case there is a time lag which does not show up in
the oscillogram. See section 3.6 of this report for explanation of
this lag.

Figs. 9 and 15 were made with identical circuit conditions but
the latter was produced with the light juantity reduced considerably.
This test was carried out to show that the waveform (except for
change in magnitude) depends solely on the circuit constants. A
close examination of these oscillograms will show that this is
true. The magnitudes are different but the forms are the same.

3.6 -

Fig. 16 was made to show the effect of light saturation on
the output waveform. The peak value of current in this case was
about 5 milliamperes and the light, several times normal.

Dynamic Characteristics (lheoreticall

The response of the load circuit of the multiplier to the
output current wave caused by incident light fluctuation can
readily be determined mathematically as follows:

Except for the case where the load circuit contains series
inductance the output current from an electron multiplier or a
photocell can.be expected to follow the incident light fluctuation
almost precisely. This is true only as long as these devices have
sufficient anode potential to be well above the saturation point.
The output current is then independent of the applied voltage.

It is the purpose of this treatment to show how much the
voltage across the load resistance deviates from following the
incident light variations when leakage or stray shunt capacitance
exists across this load resistance. Two cases will be considered:
First, the response to a single rectangular impulse, and second,
the reSponse to a single synmetrical trapezoidal impulse.

Case L

Fig. 1, sheet F, shows the circuit and notation that are
to be used. For Case I i = f(t) will be defined by

/':f(C)=0 fir 1(0)
i:f(t)31 for O<Z<C )
and i=f(t):0 for t>t,.

Applying Kirchhoff's First Law to the Circuit gives

fear-1,21.
Applying Kirchhoff's Second Law gives
. _ I '
R ’3 “ET/)6 (It. (3)
Combining (2) and (3) produces '
. . l ,
I=IC+RC/’c dr- (4)

IHeaviside's operational notation and method may readily be
8tpplied to (4).

- 12 -
(”Fa—6%?“ ‘ (5)

. / P_
or Ic=(P+§-Lc—)I7

 

The solution of (6) is
. t/
.. ‘ RC

Since the load voltage, eL, is desired, it is obtained thus,

6L: 3’5 =R(I-I;) =RI(/~ 6‘”? (a)

This is a well recognized form. It is similar to the ex-
pression for the rise in current in an BL circuit when unit
function voltage is applied. Fig. 1, sheet E, shows the rec—
tangular impulse as plotted for the case of R = 50,000 ohms,

C I 100 mmfd. and I = 5 x 10’4 amp., up to the point where
t = t1 = 30 microseconds. '

Beyond where t - t1 1 - o and iC = -i . If the variable t

is now measured from t1 as a new zero point the condenser dis-
charges through the resistor according to

. ..C
IR: Inc? ”'3 (9)

where Ir is the initial value of the discharge current.

The load voltage in this case is

8L: RI, 5%" (10)

This has also been plotted on the same graph.

The complete curve shows the characteristic "rounding of the
corner" effect that is frequently noticed when square wave is applied
to amplifiers, filters and other circuits. The smaller the value of
BC (the time constant), the more nearly the response approaches the
actual applied waveform.

Case II

In facsimile scanning the light variation is never a true
rectangular waveform due to aperture distortion. The theoretical

 

I not}
Wen-“4.5

 

 

_ 13 -

variation simulates a trapezoid as illustrated in Fig. 2 of
sheet E. For this case i = f(t) may be defined by a

i: (717:0 for t< a,
i: {[19:52 for 0<Z<t,,
[Bf/t): I ‘ for t,<t<t2, (11)
;.-.f(z)=(lra+I)-Irt f" tz‘ktv
4nd i:f/t)=o for 2‘22}.

There are now four periods to consider.

 

'

Period A (0 42‘ < L)

Equation (6) may be altered to apply in this case as

follows:
. 6 P
I: .1... Kt; p
c +86) 2 (L)

since I now becomes Kt.

The solution of this equation can best be accomplished by
application of the Superposition Theorem which is usually written
in the following form taken from page 56 of ”Operational Circuit.
Analysis" by'Bush.

/;= eMMWf/Ajéh 06(14):!) . (13)

In this, e(t) is usually the applied voltage function but will be
Kt, the applied current, in our case. A(t is the indicial ad-
mittance which, by equation (7), is 6‘ RC . We accordingly
have the following values to substitute into equation (13):

e(O)= 0 t/
Mere“ ”6 ,
A (t—2)= 6‘("‘V”° (1")

e‘OD-‘Jf

Substitution of (14) in (13) gives

t
I; =1'é-{tvv/m/fiM .-.- KE~VRZEMRCJA. (l5)
Integrating and simplifying (15) produces
I; .-. KRC (I‘EJIRC). (16)
The load voltage will accordingly be
3‘: Ring Raft-4;): KRt- KR:C("€VR°) (17)

The curve resulting from this equation for period A is
plotted in Figure 2 sheet E, for K = 31.2 amps./sec. and the other
values as previously given.

An interesting and important fact is brought out by the
above develOpment and the resulting curve. The voltage rise
across the load resistance lags behind the rise in light value
and this lag cuickly approaches a fixed value which can be de-
termined as follows:

Let the voltage at time tl as given by equation (17) be V'
(represented by point b on the curve). Hence

V'=KRt,‘KR2C (’"E-IMRC). (18)

In the case of zero shunt capacitance the voltage V' would
'be reached in t0 seconds (point a) when

V': Kflto- (19)

Equating the right band members of (18) and (l?) and simpli—
inng gives

fl,- to = RC (FEVR‘). (20)

 

 

 

 

-t/Rc
As in the numerical example illustrated above, (5
rapidly approaches zero as a limit as t is increased and (20)
reduces to

t'~ tor}. RC for Z, >>RC. (21)

The load voltage in practical cases soon lags the ideal wave
by an amount equal to the time constant of the load circuit.

Period B (z. < t < 12)

During this interval 1 - f(t) a I. Using i and iC again as
the current symbols and measuring time from t = t1 as the new zero
point, the application of Kirchhoff's laws gives

and Riflzfcf-EI-fl; dt, (23)

where Ec is the initial condenser voltage.

Solving (22) and (23) gives
° . ~VRc
Q=IE’ , ‘ Wu
in which I '=: (I- £57

From this

. . c “(/86
8‘: RIR:R(I-Ic)=RI‘RIE , (25)

This function is also plotted on the graph'between the points
marked t and t . For the numerical values used eL practically
reaches the steady state value at the latter time, t2.

Period C (t2 < t4 :3)

Measuring t from t? as a new zero our fundamental equations
become

grézI-KZ (%)

INN v.

:3 “4...“. .

-16-

and RiflzEc'I'El—flé JZ‘. (2'7)

’;=(P*;%)[2I~%)IK€]7° (3:)
Using the Superposition Theorem again gives
4; = (I‘ §¢_)E-C/Rc_/yé~(t~A)/Rcd A (29)
a
or I;=(I- %—)€VR€ KRC(I~ «SJ/"‘7.

From these

 

(30)

Then

eLsRI +KR2c~th +(Ec-RI~ MOE?“ (31)

The plot of this curve between t and t on the graph shows a lag
on the voltage decay similar to that on the build-up. This, of
course, is what was anticipated.

Period D (t >t3)

Equation (10) will apply in this case
.t/Rc

6":56 (32)

t being measured from t as a new zero and E0 is the value of eL
oat t3 from equation (31;.

This completes the theoretical reSponse curve of the load
voltage.

Two facts have been brought out by the above considerations.

I. The original wave is changed in form by the
presence of shunt capacitance.

II. The response for a trapezoidal wave lags the
original wave by an amount which Quickly approaches
the time constant of the circuit as a limit.

-17..

The oscil lograms of Photograph 3—333 check the above theory
in respec ct to the first fact but do not shoe the lag very well. A
Specia he t techni«_ue would be re uired f'or this The same numer-
ical values of the circuit factors were used to cbt1§n +he oscillo—
gr-m of Fig. 12, sheet R-333, as were sed for the theoretical case
of Fig. 2, sheet E. The shape of these two curves is the same.

3.7 .Modulation Schenes

 

In facsimile systems for picture transmission via radio
the light fluctuation picked uo by the scanner is usually used to
modulate an ac carrier. Several schemes were tried in connection
with the electron multiplier as exnlained bel w.

3.71 - Carrier Annlied Electrostatically to the Electron hultiolier

 

In order to determine the feasibility of introducing the
carrier emf in series with one of the multiplier electrodes
the static characteristic shown on sheet G was obtained by
means of the circuit of Ei_g. 2, sheet F. The potential f
nllmber 3 ele ectrode (this is actually the joint connection
of emitter o and accelerating electrode n, see Fig. l, shee
A) was varied over a wide range and the correswonding var—
iation in load current noted. If the electrode is onerated
at a dc potential of + 40 volts with respect to the preceding
electrode, a cabrier emf having a peak value of 60 volts
may be introduced. The output current should then.vary
directly with this carrier since the curve is linear over
the range covered.

This scheme was tried with a 60 cycle s uare nave li;' ht

variation and an 310 cycle carrier. Fir. l on Photograph
—334 is the resulting oscillogram of the output. This os—
cillogram was reproduced from a nencil tracing. The modula-
tion enveloge has two serious disadvantages. First, 100 per-
cent modulation without distortion was impossible, and

second the envelooe has a varia.ble dc component. There is
no system knovn to the writer which will separs te out the
dc component of this type of 1ave iithout distorting the
have envelOpe. Unless his fundamental difficulty is over-
come the electrostatic methoc of introdlcin ng the carrier does
not appear promising.

3.72 — Carrier Aeolied Ha.oet1c 11" to the Electron EulH ilisr

An attemp, was made to introduce the carrier by nodu—
lating the magnet'c ii eLl of the multidlier. The results
were less pron Tsi.g th“.u for the electrostatic method. Wot
only did the mo<M1 ation envelone contain the undesirable ac
conionent but con= ider ble mower was necessary from the
carrier sunnly in order to vary the magnetic field a

sufficient amount.

f—o

.J

3.73

13

of a DTish-Pull Todulator Vith the r'lectron hultialier.

Because of The fWIHd mental eifficulties mentioned above
(section 3. 71) in trying to introcuce tne ca~rier into the
multiril -ier itself The next logical schrie seemed to be to
take the normal nultinlier output ani drive a suit=ble Type
nodzllt ator. A pus‘—pull modul;tor using a double triode was
set up so that the grids were excited 1310 out of nhase with
the carrier and in phase (or in parallel) with the niltinlier
output .

A linearity check was made en the push—pull unit by
varying the common bias voltage and noting the variation of
ac output. The circuit of Fig. 1, best H, was used for this
test and Fig. 1, sheet I, is the chw -ctcr1,c1c obt:ined.
This curve shows linear rela tionshi ? over most of its ranee
but is not linear for :m ll v lues of output. This is in—
pcrtcrt bcceu;e a.hijh r t'o of output for shit; to blac
is desired in picture scanning.
is 7, sheet H, Chen; a ciTCiit Those ICJWJCL
is linear down to zero. The fundamental ices oi
n is h8t ttxe carrir: is bel need cu+ c? The output

('A'L

this sy-
for The ltwmn‘znon—linc331iurrt of The;ruxnnzl Ch'r cteri.sti c. A
counter emf cir nit composed of transformer T3, resistors H1
and Hq, and condenser C supply a voltage of the proper phase
and mtgnitude to balhnce out all the output repre-rnfv: by
point a on Fig. 1, sheet I. Hence, point A on this curve br—
comes the Virtijr'.l reint of zero output. Fig. ? is the actual
Eu racteristic Thigh results. This is linear dovn
tiors inoicetcd that

modulation C
to the zero paint and oscilloCCOpe observ:
the uaveforn was also good over this range.

The next tCst as an overall linearity check of the
electron nultiul ier nd moch stoL“ chitin tior. Fig.1, sheet
E3, shrfinr'the nutttiflifu? circui1311sev, True 01. wit oi‘-.irict
applied to the modulator circuit oi Fig. 2, sheet H, in ser
with lead s—B. Fig. sheet J, shows The result. This curve
has a "'TT‘C slone beccuse oi The multiplier connection used
which :ave an output emf 1ncrea.3n~ in the urcuiive cirecH on

s the light increered. The overall characteris ic, Which was
made wi_th a 5 kc carrier, is very good as indice te; by the
small deviation of points from the curve which was drawn Wf.h a
straijht edge. The arount of lieht used new about ten percent
of that obtained in a standard rno or53_ o scanner. (The latter
was metsured sometime after T‘ elinearity test had been nude.)

('1‘

Ho
(1:)

C‘
-L.

g_.|
so

The system described above was oneretcd with 69 cycle
Square have fluctustion of lirht and the resul,s were good as
observed on the OscilloSCOpe exceut that difficulty rat had in
'balancing the modulator to keep the dc transient ":ick" of the
inultiplier outp t from appearing in the modula tion envelope.
This difficulty has now been overcome as will be mentTQDed later

in this renort.

3. 8

Jtrt

S

In order to determine the performance of the electron nultinlier
with a moiulator at freguencies in the general range of rage—a—minute
speed a "brevC—board" setLIp of a 75 kc oscillator and moiilator were
made. Thi.s ner1101lar v:1ue of cmuiriei freguency V88 chosen well
above t1w:niiimum value necessgry for page-e—ninute speed in orjer
that there tcul} be several cycles of carrier per cycle of light
fluctia.ti on. Hence, the roiulation envelope ceild be 1ore accurate-
ly obserVee.

Fig. 2 of Photograyh 9-319 is a vie* of the 75 KC oscillz+011
and modulator units and fig. 1 of Sheet K is the combinefl circuit.
of the multinlier and uonnletor. In +his case the la ter u1it ets

one of the
multinlier
no output
of inversely

its fixed bias voltage from the IE drcp across par'
bleeoer 1eszis1 018 in the mu_lti1lier circuit. .M1 0 1h
circuit is arranged to orive the m01h11ator so th:t it
is directly oroyortional to the incicent light inste
as .as t}1e case of previous tests.

. O
U) (D ch‘l-

W
A
,. J

t sroph F.- 33 4, (toys a tracing of an oscillo—
T] I

115 W35 made

grapii observe ion of the Loculaten 75 kc carrier.

vi 1 3 19 kc light iml ctuaticn and an aperture sistorticn of 3? micro—

seconds. A p935VV9119 of 1C' volts ec~oss the 670 ohm load was easily
his is good evioence thr:t 1n t1 additional development

obtained. T
the multiplie er v.0ulc be suitable f'or high speed scanning.

One of the problems to be solved in this connection is that of
finding a sui+eble ta lanced modulator. In all the tests so far dee
scribed co siderable difficulty was had, as mentioned 1wr Mi o11sly,
in balancing Ehe no1u‘“t1nr emf entirely out of the out out. 11 is
problem vas oiscu ssed titi engineers of the Electronic Research bab—
oratory at Camden ano they su*"est“c using a push—null 1.001m11tor
with tetro des. The carriei could then be applied to one set of grids
O

0 degrees out of phase and the modulation, to the other set in
para allel.

Hf

A brief check made on this scheme recently inflicz tes t.hat it
s

is sati factory. Two type 6L7 tubes with the carrier anplied +3 No.
1 grifs and the monulation apnlied to No. 3 grins urooucec the os—
cillogram of PL .3, Pnoio'r'wq R-334. This was made with 60 cycle

ssuare wave lL htv vari ation anti 3 3240 cycle carrier. The balance
was orecticelly per rfec t.

11cellgggpus Qbservations and Comments

One disadVantage in using the mag. 11etic tyne nultinlier as a fac-

imile scanner is that the unit as a wh 'ole is quite bulky due to the

mac ssity oi' the macnetic field sup lJ. The electrostatic type of
multi-lie1 whicl1 is being ceveloped by the Electronic Research Labora-
tory at Ca.moen will larg ely overcome this difi’ioulty.

I“
— ‘il -

A small amount of dark current (from 5 to 10 microanps.
at times) was observed in the particular tube tested. Thi
creased several-fold when the ambient temperature of the tube
compartment was EllOted to rise 15 or 20 degrees centigrade above
room temperature. This dark current is believed to be cue largely
to thermionic emission. Hence, if the tube is kept at normal
room temperature the amount of dark current rill be small comoared
to the norm-l output.

The output of the multinlier rises rauidly with the total
annlied voltage or rather, the voltage per stage. Noweve:, for a
facsimile scanner head, flexible leads are usually necessary and
space requirements are at a minimum so high voltage is undesirable.
Conseiuently, if the 10 stage tube gives more than enough outout to
drive a suitable modulator then either the voltage per stage or the
number of stages may be decreased.

4 - Summrrv of Posults

 

4.1 -

Y

4.6 -

The maximum amount of li;ht available at the oho ccell in a

a
standard 5A-914 scanner or one of similar tyne is in the order
of 6.4 X 10"4 lumen.

Eith this a ount of lijht the dc outout current of the 10 stage
multiolier used in this investigation is in the order of 5 x 10‘
ampere for 130 volts per stage.

With a light chonoer fre<uency of 12,000 cycles her second the
maximum load resistance that Can be used without appreciable dis-
tortion of output waveform is 50,000 ohms.

With 6.4 X 10"4 lumen of light (chooped) the peak value of
voltage develooed across the 50,000 ohm load is 25 volts.

Additional stray or shunt capacitance in parallel with the load
resistance alters the waveform as shown on Photograph 8—333, Figs.
10, 12 and 14. If Fig. 12 is the case of maximum allowable dis-
tortion then the time constant (EC) of the load circuit should not
exceed 5 microseconds. This same amount of distortion is also
given by the curve of Pie. 2, sheet E which mas computed from
theoretical considerations.

The theoretical deveIOpment shows that in addition to the change
in waveform caused by excessive stray capacitance there 's also
produced a lag in resuonse. This lagoin the voltage or current
wave is almost eoual td the time constant of the circuit.

Excessive dark current results if the multiolier tegperature is
allowed to rise more than a few degrees Centigrade.

4,8 — A carrier freiuency can be successfully introduced into the
multiplier by electrostatic men s, but the resulting modulation
contains a large dc comnonent which cannot be removed without
increasing the build up time.

5 - Conclusions

 

The electron multiplier should made a successful high speed
facsimile scanner unit for the folloving reasons:

1, It has a large output of undistorted voltage (5.x 10"]+
amp. through 50,000 ohms ecuals 25 volts) even with the small
amount of light available in scanner systems.

2. Its frequency response is linear OVer the entire range
covered by page—a—minute Speed.

3. The electrostatic type of multiplier will not be too
bulky to mount in a movable scanner head and re uires no magnetic
field structure.

4. although twelve to fifteen hundred volts are needed
this could be supplied by a rectifier and be properly insulated.

, 5. The output of the multiplier is sufficient to drive a
modulator.

The electron multiplier and the associated equipment developed in
this investigation are proving to be valuable laboratory tools for

0

making circuit tests in connection with high speed facsimile retuirements.
As a result of the work completed to date "n this study, the cir-
cuit of Fig.1, sheet L, is suggested as the most promising scanner
circuit in which an electron multiplier may be used. Althoujh this cir—
cuit as a whole has not as yet been tried, all of the elements which
comprise it have been found to be entirely satisfactory. The multiplier
output (which is arranged to rise in plus polarity as the light increases)
is applied to one set of grids of a balanced modulator unit. The carrier
is applied on a second set of grids. A counter emf circuit is also
proposed to improve the linearity of the modulator for low values of
output. A final tube is used to combine the normal modulator output
and the counter emf and give additional amplification.

 

 

 

 

l

 

 

 

SHEET/1

 

:F
K m o «y su

 

 

 

 

 

 

 

 

 

 

 

 

4L R7 R8 R9 R0 RH

#500 Volt:
Fiji Io Stage E lecfron Muir/plier and Circuit

 

 

 

 

 

 

3“. ,. I
..- [-

 

 

 

 

 

L

Fig.2 Chopper Disc and Oprical System.

N MU mum AND ASSOCIATED
EL ECTRO LEQU’PMENT

a u .v
”’4 u414...” ...__.
V V"

 

P'
RI: .JL.

 

 

 

 

A

“*VID. on.

‘w ‘ " ."m‘ "5-" no . votes I I-ucc 1”

A'-‘ .9..— ..~—— .r'.

- 4m’fihtm-.~ ._
.f .

 

 

1r '1 '2

 

 

0
~IJ'00 Volt:

Fig. I

(It, 1r’2 ((83

 

 

[0K 4-

 

1' If

I ['3 [u l J‘ l 6 . __

60K-4- IOKA IOKA- [OK-fl l4K-“-

 

4vwwwwy
36 K4»

Multiplier Circuit for Negative Output

 

 

is

- I200 Vo/f:

 

 

SHEET 8 j,

  

 

 

JWWVWV
M K“-

 

 

 

 

 

10 K»- Io K-e . ' /0 KA Io K-e 10 K4» "'
we ' -4 wwww——-o
36 K4» =1.— ’0 K“ i»
7 +300 Volts

Fig.2 Multiplier Circuit for Positive Output

El. E6 TRON Muz T/PLIER CIRCUIT

 

 

 

 

“.7...‘

 

It-

J1! , it!“

-.-_

 

,-L.._.---.-.\.-..-.... _. ._..

 

‘x
\
|
'\$_

’.4'_..._ ‘_ _

{Cr

L

n
g

 

. . . u ,. . .. ...... Y.

. . . . . . . . .h

I . - 'i-"'c ..Y‘I‘ll-l|1vqn I 01!; u-‘ .‘v.lAv.l LO-
. . 4 fl 0.... . . . - “- u.

-. , . . . . U ., o .. . ..u.

. . e u . .....-nv .. .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a i.-

.. #- H _ - . . w . - - 7...; “guy
if}... . h... T - i .- if. i ”In... .. .. -. L -f T- -.. - .-..-H--..LH--..:-:.-....-
.H.._. m _ a M _- _ - H . i. W w .. m; .. ...MM
.10! . L L i “5.”- -. e .H _ . ;. L. m...”
L .- m. M _ . l- ; _ s. - l . . .m a ...-m-.-...
7!”! - --.w l- N - _ .. f. .-- x! a- .4- - -- Al - - a. . H T... t . {ti-“1.-....t. at. 3.1..
l»; l . .w _ M .- . ..... . a ..L . v.3. i. s...-
..“-.-m- fl I l H: . - T- H. - I. ...r. w .H. ..T... S . L - 4 i -... t ...4-.-; - ratings I...” .- $2-4.“
Til... «L . LTLL 4 L4 .. . . L . H

. - . . _ l . .. _ - m. . .....
_ . .4. - . . H - -4. ...... - ----

p. L LT. L”: LL #1114???

 

‘ 1
l

 

. .419
amp
i
L...
in
L
L

K .4
i -

7 ..-

L

 

l.
.‘l ..
an -

 

 

 

4}

 

.pr

 

 

 

 

 

 

‘i
L
.!
Y‘LL'
49.5.”:
L—

i .4
unre
..La.

 

C

 

 

 

 

 

 

 

L
I-
:L:
”F

Y

 

 

 

' 9 gL‘dLITehj"
twrr-t‘
L
L
l -1. -
I
L

i
444L444?
i
L

o.
I

 

-..“L....
x
|

 

 

 

 

 

~4
._ J
'_'uf utpu

-.._-o.. o

 

 

 

 

 

 

 

 

 

 

  

.Jgfl

I
n...—-
. ‘. '.
.l A F. I.
g t
a .. .. D-
. - l . -.
.I .
no ~¢ ..
. .1 '

 

' ”Li

Y 3 I

I __

sew-«mew
l
1 .
l
uh" I
L 11+!

1......

4'
3E
.’
#-
3—434
2°
:

o y
’5
Q.

9 .
‘34.;
“:5.“
Le.
22

;

 

 

 

 

17L: IT. ...LT- -H ...,-.,-..¢ IV... .2 4.. w. -.
-* A. -

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ELEC TRON MULTIPUER STATIC CHARACTERISTICS

 

 

 

 

cccccccc

 

 

 

 

 

 

 

 

 

 

 

 

 

 

We
1.
L-

'. “ML-l " "I.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I
r
L
3
4. LT
‘5

U
I

 

 

 

: j :"*”[ :-
. .- - 4 .-‘.1
L 4' ‘ '

 

 

 

I544.
transit

 

 

e‘. 1
6'0

 

L
‘- rad-"T
_ ... 4L“...

 

Ohm! x _

Ew‘mmt
.._.'__.__. .....

30

‘-

 

 

 

 

..-
». -.. ...-

‘1?-

 

 

To

Reflutenqe f—

M
.j__.i..i
L

1W
_..T

I..."
L
_. We... .- _-
L
L L
7 4p . :
Rectu‘ramcc 4-0/1”" '¥ [0

 

f— T L
L

-.- L -.-
“—4-
L
I

I
'— "7
L

in
J

T
L
Load;

Load

 

rLLL L. . 0L2
-- ho hrLc to no WLQ e6 .

.uQEEE‘t- guess.» 30.30 angst - . l
.L: . L- ama 9! cum on 3. o. . t ..d c0-..--_LLLL

H - +6 .L «.6 ~
l, - use; ES 2.323 1e§§l§hfi 3&3. E... as:

- .WL. .LLL... - .LLL- LL ..LLL .
. - . l _ my a ca
- . . L. 4 . T... w .. _J. owe?) bum“ nuaEHauq ~

h NL_Ll7.--4L-...-- .L_...- -- . -

   

 

7
g-.
i

 

 

 

 

 

      

 

 

 

I...L.1L-...L. -.L- ..ILLL L?

p-- -rL

a- v . _ . Teal.-.
_ _ a .. - .

L...”- ;.L

 

 

6
...l. _..... ..._

 

 

 

 

- .
T _.. _...J . . L14--. -.I-L
- -

 

 

, _ . - . .
LL...1_.L...L
l.

a...» 5-. .fl -..4 l i.-.

. ..L.-.\»\L\\I|L. -u .

 

 

 

 

 

£C£¢TR0~ MULTIPLIER STATIC CHARACTERISTICS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L
Rd:

 

 

L

L
.L.
-L- .

Ll

 

 

 

 

 

- L L UM ._
.
A . fl» .

 

 

 

....o to ..w

L

La

3
. ... h9g0flfi$gf 5 L559.»

JD

Liz.--

L
I
771777: Lin I‘Lficrog'cra"
VLLA
_L--;-------

 

 

Law an a.
..L ”unfit

 

 

 

 

 

 

W M M... H.

S

L L+L LILLI +LL LL-

. w L... _ .
§.b§0u . .

 

 

 

 

 

 

 

 

L~:e-i
J

 

 

 

.
J p L LLL - L IL

. M L L . .

. . H W . .

LD -- ML ._L.LL LL..- .LLLLL-L L-_.- L. -
M -. H ..L -
U . Mr
# Ota ., . I O"... I. A
S m . . « . 2

. .. .

W . _ .
.pstzerLfLLL . LLLLLLLL;:LLL-
I M . . - n

. _

 

 

 

 

 

 

 

 

M32 XQEEEQ S ucuMEG

66 «..d-L- _..o . . -

 

'o'naL-L- 4' (7’3" :

.. ..t-.....-
L

L

In: La Nkrb:

L L
L
....51.
L
EL
‘ ' L

TIL

'L
'L
IL
..-]-
ll
:l‘
TTY

 

. .
t--- 7".-
I

...W.. L” .H fl... Law .m... .97. 3.: ...-L . 3.. THE: .... LtL. x r . WW 7.1.: ...“...tiwmum-m.
..W..;. . .r g u r .H- x ..r. g... .. ... ” .... h. .:.. ..Fr. .....- .. .. -. -- H . ”W in”... “m
1'90 thIn L6;\Io .v.~§ LL I ...: ... --lLYUOlIl' . OI... 'Div .00.. .0- a .1 .IL .‘Allll‘tAa-Oa . -|.v'||n .LI .9! .300 . tn $0.: 0 OI... IVA I ‘lv...Y.oLIO I t
E- “ .-. L ... 4 . . .....HLL .. -..- .M. ...H. ......“L” 7*; .- . . . .. .. ..T:-.. 7 u». g 1... ._
..u. . . “ ... . .. . a .. . M. r N ._~ . . .. o. .. .. -. .H. . . .. .Mo
. ._ . _ Fl L.. L“L .L1. U PW L... . .. .u.. ..... 4.L .. ..LL.- . .1 L.. . .Ol' L LW.. 0 M .sb.
E .n . as. n . . u .. . .. .. . . 4 .. .u.- ....o ......h- .... .. u”. ... . a. .a .....Jcofi
fiEIL . . ....m. . .M 5 W. . o o n . W» V .w... .. — ¢ .41. r u“. fi-.-.— .4”
.L ..L: .. L L. -. - .- .. ...v-L L --.r 3... ... . H. . .- . .LLL.L.LM..-.. . ... L ... .9. -. . ... L LfiL .L...-....»L.
E . . W. . H g . n. W .... .- . M .. ...; «MW-um
.. . . L L L .. .. H .. .fl . $323
L L 4 ill] L
W . . . -. 7 . . ... . ... 7... ...... ...H ....m
... a...LL.fiL L LL. . .4 u- .. 214......LL L.L. ..L ....LHHCTYLYLL L L L .-.M L-.L . - L L: an..nh.L..mLi. 19L: .-
. .o . . . _ . ... . . . . . . ”...... .fi.......- .. I u - . . “.fiuhuq‘
- .- . . . _ _p - - .H ...M.L .... ». ”-...-“...... H. 4-H .. . 4.. ..fiKLJq
. . L M . LL. ”LI L. A. . .. ...L .4. . -. . J J. .. A ...4- ...
. .. u . . . .. . .. ... ¢. . .. . o n . . .A4
.. ... __ . - .- ; . .. L p. . . . : pm..- . . _.. .L L , - .. . .. ;...-.-.-
... 7 . L -. .. . -. L. ML...J..T M.. .. .LML 4.4L; 2+..L .... r . .w L LN." ........ W .L.L..
” w .r L.L . M . L .. .. . LLL. .. -. M.” .. . mL L L. n .W.....
.LL . .L . D . . . .. . Lilfi ‘L . 4. .11LL
. . . . - M . ML .. .. _.. . _..“ .- .. - . :..”...
. -. . W . . ... - .. . .. “...-W... W. .. .W
. . u o . ... . . Ht 1 L -1401 ..... . I A .II U. u. .1. 7-. LOIQLJOl-rll 0‘1. A I fl ‘1'. ‘‘‘‘‘‘ I 1......lk
. ._ L . L . . .. ...-...... ...-.... .. .5 .... ..
. LML . Lip-LL!- L LLLLrLLL. - . - -
M 4 -W .
F L . .

u. .
1 - Io-|l10}
. <- 5
.o . .
O 1.4
.. .
L
L L {L
o

 

 

C

' L
L
L

 

M
.'
O

 

"L”LJIV

‘L

TRLLIREzom/LII.

- ....

;.-_

 

L
L
L

".L'" ....T.
L -. L
2 ' .
' L' L .
. ’- ..-. .
L -

 

O
m-4L-—O‘- —o-§'----o.
. Q

L
L
i .
I
L

 

 

 

 

 

 

 

 

 

. -w......-T.-.u...n -LL.
. ......z :0.

 

 

 

ELECTRON MULTIPLIER LCAD . CHARACTERISTICS

 

 

 

 

 

-._-.—_-—_...—-._— .— _

 

 

 

——)>+ . .
l, ”c ’R
i=flfl c _T_ TL R
_ a i

 

 

Fig. I Electron Multiplier Land Circuit

SHEET F

 

 

 

,“I '2 ‘3 , ’4
1 60m» tou« ioxa ioKA
‘ K
20:“ “A
r-‘WVWW—
:oxa- .
2.an

Fig.2 Circuit for Obtaining Characteristic
Shown on Sheet 6

ELECTRON MULTIPLIER CIRCUITS

 

 

 

"Tn—Lu:

 

 

 

 

 

; 4—4
L
*1

 

 

 

 

 

3.5. i 33..

 

Lima; 6
1.--.+ «.1
1 I
* f‘“

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

--+.~.-'~s
I

 

 

}.-

 

 

 

L. . . 1 . L. . H
.iN. OWN- .Q N1 W . 92. . h. .mmr.umu1...u.
T q 1 .. 1 . .
.. . w . _ . .. . . .. muffin
. w: . +1! . L 11 1 1 _.-. ltwfl- .. ....H. -.1. --.111M3H1w11ffiwmwu
_m 1. 1.1L L _ Liv . 1 . ”g .1 L. ._ LL”. LL
L.....“ . : L L .V -. - n .. “...1.”-.- ... Wit
L” m .. -. . . w . W2“. fin .. I’m 1...
L.. .U L L... a
11” 11 H.” L..1..“1....11_r-:..:..._0 111,41 _1111
T .. 1 w .. _J ... ...h .” ._r.>
1. 7 _ _m L.. 11
$1..“ 1 1.111111. ..1.. 1_ L. 1910.71.11 11 £1 1:11 1 .1. . _.....1....r.1w11111 11 1 1 110L111
..-L.. . -.. . _wm. L.. .1.. L. 1 .. .. .. LL. 4&1
1 1.1-1.1.. A . ‘T11 1 1 “1.1- w ifiiWflr. UNLIWLJJ .....1; 1 11.11711. ..rwm
3. -11... . :11 _ .1 1 F 1M- 1 11111.11.-. ..
11.10%. 11 — . 171.11%... 1111. .11 3.1.7,“!!! ’1 1M ..H k ..

 

 

t (L ..
-..-.... -...

a v C

I . O

 

 

V

I z . |
.0—
1 I
-—*-t—
I n 1 v
. a 1‘ . O
. 1 1 - 1 I ~
1 .. 1 u 1 I
-4;.- .. ... . 1. . 4 .
. 9‘
. a
I 1 - I 1
'Cd

‘11.:—
0
I
9
. L .
7T—

1
.1,
l
‘
a

..

9.1449994. .
'L; ' ' '

WNW” Nina?“

 

 

 

a u:
1 qt ‘
. 1101104
I

' fagw’ up“;

 

- ...191?

o- —.
1.
u
1 . .
WW
I
I1 .
1 1
-

 

l

 

 

 

..9P44439112

O

2

"i". t-.
L. .
9

f.

. .
....
rw

a
l

 

 

 

 

 

..
1‘
I
Y
0
0
.g .
' I
1 1. .
a
" V
.

  

' 1L '1' '.
7.;

 

 

' 41-1}{4'4391‘9101

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MUL Tl PLIE R
CHAR/46‘ TE R I.“ TIC

ElECTRDN
STA TIC

.1020 brain. .5. F111 .Nnaldflll". .. my!!!“
.

 

 

—_.. -1—._._
E...

 

 

 

1.7000.“ .
C 4 trier _ ; _

 

6“; GAG

 

SHEET H T

 

17m
@139

 

lI
*1F*

 

 

 

 

 

11+

 

-C‘ircuit for Date rmim'nj Balanced Modulator

 

0 2E0 VoIT:

 

    

 

 

 

 

 

 

 

 

 

 

 

9%?

v

 

 

 

 

 

 

 

' Fig.1 . _
Chdrccterutic
6A6 ’ 6A6
*“2 i 3
1170009- A H n n n
Cdrricr U f
T 1L L
[C /%I
F v =1:- .
0- .- .1
19°11
O
D (:5. § V.

L

 

. F132 Being-cod Modulator with Counter Emf.

CIRCUITS FOR BALANCED
MODULA TOR CONS/D E RA TIONS

 

+25; Volf: .

 

 

 

........

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

  

; -3?

   

' l

3.»‘

.-
..... ..

W ... . . ...._ . f. a. m in. .. .. M... .. - .7 a ... ..... ...
, . w - u .. -H. . .. . .-.. . . L: .1.. .+3., 3333* 33.3.-. .....- ...W . .3.-ka ... - . £1333.
_ . u w . m d, . M g ._ ....m. . t. . u .. .,
- f. 3 -334333... .313. . , 3T ..M ...r. .....-J. ...... .3; . . U 3.3 .3 M . . .
. ‘ w . _ n . H. .L ... » ... . u . an _w
_ .... . ,-.. ... ... . .u. -. w. ..... :-- ;..-: 33?. ... ..- :_x. . ... “1+”. -
. _ . : ; .. . . .
_ . ____ ... . 3 .m. ”7..
A . . M . . t. 1.. _H -2. -.. _ _...mr. a -
.... , . ....... .. ..
. q r. . .... , r, .. H
F 3. t . . . .n a.
. .. .. . _
fl . .2: h. .... .N. . a H.
3”. .. - .. -..}...T... - 3t3L...o..-- .. . m.
. _. ,
. ;.... .1. ., , . 1»...
A 0. .l . . . Y a .In
. pf 3 . *3. v F ..
. _. -. .13 .fl _ A...
.. ...- 1,..- 3.. ..- .n: “..95.m
.. M 3 ‘ 1

O .
. .

 

 

‘- . . .
;.-..i...,-....,....- ... --

c . -

3

- -.....Lw..-
O a

l
3
I ' ’ _
l l
_‘fi-w-W- -.......-
3

 

 

      

1"

 

 

a J 3 . .7
Munro war

 

 

 

 

 

 

 

 

o
”5-“... _ ......“ --__ _,
' I
' ' I

 

 

 

 

 

 

 

 

M00009 TOR CHAR/90 TE R I 5‘ THIS

PUSH-PULL

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

« a 3 .H. . fl .1 . .. . ..... 4 ..u3 A «I .a . a. .4 ..N. .« I u. .n ..u. ....Huo.4
. . I . .. . . o c u . . a . ....... . .oo‘lo~_
._ _ ....-. . n ‘ .. . H.. . . n c“. . I w. ..“n. fl ..—I 3.“. Inv..a.n.I.oO
. . . . . . . I. ... . . . . ... . ... ..
It I II» l 33‘ ‘ 5‘ v. 03“... .3! II ... . 0- 3...... Illr'llo— _ I a’I.‘voI9-m- 'A—lnznt. IIIw ‘JufldLjvqu éflLrI‘IJIo'Ioll l‘a-HIu‘HllHH‘F-Q‘Lj.06u.’ad.]qb'4
. . . . . a . . . 4 u . ... v... a A - ... . . . I a ....... .... ...I
. b . A . .. .a. .. .u.... . ”I ...- . . u . . u M . .. . o.“
v . .. I t I ..... - . no I. - Id
. -.., w . w. _. n w . y 3 3F 3.......1 . a . . .p.3..
. .. . . ..... . . I.. . ..
H . . . . . , ... n . .“ .... ......t... . .....:..T...
; . * w . Tl H L w ... 5.. .- oh.... 1 o ...+ 0. fi. . ... c...
ta 0 6 In I o o IIIv I l; I’I‘o . 0 III I 5 I Q 0". IO 0'. ..II. .oI I'LOIOIIIO.‘ I. Q I.-. .00“ .00 9 IQ. 000. 4344
. _ . . . . . . . .. . .43.. .- .
. n _. .. . ... . . . .. ..L 4 L.. “a. ...fin.
— . . .. a. a . .o a. o . o A .... . . . .-.? a} I L
a — H. . . . c o ”I o v I . *3 I . r A I IQ. I.’ IW‘I
3-33 I. 3 H 3 H . . ..... .3 . .......
m _ _ H - .. . . t . A . . .. . . t. . .. .
H u .I ' n .. H u . a . .0 c . ~t ...-N O - a
.I TO. 0 VII I A III. .o. . '0. .I I? I.‘ .0 010-1. QI‘II I 9.0 '3 III. - ..- PI ... II V 00' no: IolI'IkV II lix-"lo‘h
fi . . J. . I . o . . a «A.- I... 5”
| IA. .I V. . ...
m a . .. .. I . I .o . no a....... .w ..
V33 3 . F ...L.. . 3 .. .. 3.. . H H 3 3.3!. - 3.3.. .- . F
4 fl . H . .. . 3. . - . :....-
. a. . I . O «0“..-
a . _ * pm H . . . . a . . . . ......
. . .... . o P... . u
..... W1 * o o I. O . r36 . l I III-5|... "3 II 1‘ '1‘! I _I o to... O. Q IO!O.¥OO I“lt,-
. _ . . g 4 . _. c. . ..nn
. . . . ..

 

 

 

 

 

. . _ a
. . ..... I a:
. . . . . j
. ..l . A. . . fi . . .. . . . z o .. a o. - n. .
I.... .v ....... V. I 0.0 ... col...I II. b IIII 90.3. .lll-.. ... .I I! .0 I.IO...oIr..II13.-VII .. 5' IIVI I . 5. Ifi. ....... 1. 003.1 . v ..1- .v - .o'II'. . I- .0 O . on -..-conblto.
. o . q . . . . . I a. u a
. . o — . . 5 . I

 

 

 

 

 

 

.0--. IIIOIu . 0....OI0I

 

 

 

 

 

 

 

 

 

 

 

3 .
i
I
3
I
_..._.L..
- 3
b—o—Ov-v-
3 .- -.
“ \‘i‘
3- '
_- :.-,1
12
-3.
3 a
i-
l
I
-0 R
f
3
i

 

 

 

 

 

If

 

  

 

 

 

 

 

 

!

 

 

 

M»
3.
A 38
1.
33-

 

 

hbfilq; _
. “.5

 

 

:
T
3 -_

 

 

 

 

”4...-
I. .'
.i .

3 3
3

 

c
O
...
I

 

 

 

 

L..—*— l— _-.—....
C t
T--WI3~ -...f
9
O
|
v

 

 

 

. j :W
MEML

 

 

 

 

ELECTRON MUZ TIPlIER AND

 

 

 

MODULATUR SYSTEM CHARACTERISTIC

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. H. . U
_ . o o . I u
I 3 .P . . . n .. .
q . a w ._ _ . ... .. . u
. . . a c . . . n
.. o 5‘ . ..... I I .
004‘ - I ‘ 3 I 3!. IIIII 0 :..?»3 c -oIOII AI4-(lo I.I.OL 3 «I. u out... . OI LVIOIIL33|I~V A I III I.h.0.l-
_ . _ . 4 .J ... .- .. u .o.
. _ . . . . ... n . . . n
. . . L .. . . . fl .. . . -
. . . . b . I . .....
3"lY'lllllA].I A *lw 31. 3 “I3 .c 30333 4‘
n by.“ Q .fi .9: . . . 1 .ha . “ ;.... ..u . H . r
fivIrIuo ‘ I 03I35L .II.I3I A III?! IT..- old I II! I I TD .L. I- Q ,|0 Av q I ”.03. ¢ 3 03. 3A 9-3-1: '00-...‘00 IOU“OFI00A-IHI II o. :MIJA IA
. .. ... . . _ .. :...wa S N; Q... 83:32 .I - .
Y .. II- . ¢ 5 u . I . n
3.. I F. P 3V 'I... I L by II . La .3 3M
A | I..m . . . o a u . . . . o F. n 0 . Q o
hIll’IILYIL! Ila . 3 r ..I III I I. # .3. I . . A I o I I I v u I VI III I I #III'IFJIA II T II 3w . .. II- IA I331 JII IIIOVQI# I .33 IA
. w . . ._ H k. . . .. h... : H.
I . . . . . r u 0. I m.. .
3h 3 W .3 .. . O .1. .~ . .
. 1 _ . 3 . J34 .1 ”3.“. .
. . I.. I n .
v _ .— -.u I. o c I «- o .. b ..
fin I.|.~ O IIILI I Q In 3 v I I. I I t ...-:03 I LVVIIIIIt.I.OlerOII§IIIIIJv O'HII‘IIA IIIOIIQI. AWI .... Tlfiiu.v‘.‘., VLH‘” .
. _ , - _ - . . I .u ,. .. .~
4 . . . . .. .... . . ..
3 3 03 3 . L . ....— 9 g .. I
... . I 0 [I33 31 {IL
. . o . . . . .

v . _ . . h . .. .a .. . . . ~
- -3 .L - ... - - .- I. - --.... - . -I -- I *1 .. ... ... H . .
TI 3+ . _ f # a . YI o v I v . a . -It....I ....IOIIIIIv . L
r I. .. . c . . .IIUOII ...-.... ...... . ... _
..va _. . . , b 3 . _ . .3 ~ . .3: a. g . 7..

3. I . ...o ... . .

 

 

 

 

 

 

 

SHEET K ‘3'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

=36»-
__3 4h- __
‘1 IA"
3—" ,r.
zona-
GOK“ IOKA I mm 10K“ 14K“
WW . 462% MIL-3W. .
' JIZW
. 3.5K“- . 2.JK4* 13 “if/aka
3 ‘ .W‘v‘ *-—-’W\NVV‘-—-<’
* '--’VWM—-1> 3
‘1200 V0,?! SKA =3: . +300 ‘VOITJ‘
76': A
75;“: *IJE - “En J ' .2
03¢. “‘3-—-" arrans U?“ 3— :: Load
*ZJ'OVoIr:

‘53.: ELECTRON MULTIPLIER DRIVING A
BALANCED MODULATOR

E [56‘ THO” MULTIPLIER CIRCUIT

 

UN

 

SHEE T L

Lghf

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

;
r

ng.’ Circuit for Preposed flecrron Multiplier
' Scanner System

ELECTRON MU]. T/Pl. IE R CIRCUIT

A

 

 

3F] [-4 . i ' '
(3 f
4:3 , E I 033—— 3% our‘pur
3333 , 3e * '
‘ .. 1 3 ..
Carrier ~'
Input

1

 

Winn ...-...... ...“ .. . ‘3'

“le11§q414§1:11§<1l4§H 3% ‘ 4 .I .1 14.1. 4| 1} nl E 14‘ . ‘ ‘ \
lunar n rfiéuxbflrks .. .L. - b. artisa;

n... «.1 ..m
r.» f:

.o mwhm

 

 

91229
:2 - "4 z

 

 

 

 

m‘~ m9m1*1'-H'JNE'V*'W '— I ~ gm“. ‘0

 

F19. L

III-I'llwfluh'.“ ..m .Ilv
i. n . ‘
.1. I.

 

 

Fig. I F =8000v- Fig.2 172/2000"
R = 25000n C . 0 3'25000» C '0

 
  

   

;- - ‘
..-.“-

Fig.3 Faaooo - Fe). 4 ' Fa Izooo ..

R’ZJOOOA 6:0.000‘5/14f "'25000“ 020.0005uf

 

A
Fig.5 F-aoao - Fig.6 F . 12000 *~
R =sooooa c to R =Soooo~ c :0
— — - —.

D -’ - - d -
n9. 7 Peso .. Fig.3 Few-a
R250000A C'O R’500004‘- . C-‘QOOJLJ

mama/v mummm OUTPUT WAVEFORMS R9223;

A

Fig.9 R amo, 0004-
6:0

Fig.“ 0,000-"-

12:0
641.

 

w,

Fly. 13 R-2.£000-‘-
C = 0
.. : \\\\\“ w- -

F19. I: R=I04000v~ (“=0
Lyhr Reduced

ELECTRON MULTIPL/ER

L:—

Fig. [0 R2 ’00, 000‘“
0: 0.000 Lat

A..-

R=n5'0,000“-
6:0.000M1S

W

F13. I4 R=50.000""

——-.—- ...

sz9./6 R=I6,000- 0:0
Lighr Increased

OUTPUT WA VEFORMS 95226.

R-333

ELECTRON MU! T/PL/ER MODULA TED
’ WA V5 F OR M8

 

W

F {9. I 60 "‘ Chopper
8/0m Carrier

“fl

wf H n}: W

Fig.2 [2,000m Chopper
7.5 KC. Carrier

 

 

 

 

 

 

 

 

‘1
‘

Fig. 3 ‘60 " Chapper

.3240M Carrier 95225..

R-33‘fi

 

. . 1. . . . A
V _ . ... , A K, . ...».t. P. .r. . I .

I v ”2: .c» .l .. . .. ‘ ‘ ... . u

 

: v Hi?

 

 

 

 

 

 

MICHIGAN STATE UNIVER ITY

l n H i’um ,Ililjmfimn'“

3 1193 03058 1