THE v" IG1., c0R..T..UC’£IOH, MID CALIBHAIIG

OF A G1.1«auATI1«G VOLIMEIEB

FOR A WO MILLIUN VOLT ELLCIRCMTATIC .Z’TGEI„ RAIOR

V

“v

Robert 1.. Bow/Ian, Jr.

Thesis su\·m1tt•d to the Graduate Faculty ot the

Virginia Polyteohnic Inatitut• V

-2,

TADLD OF uoßranfs

Page

List of Figures .................„.„„...••„.......„•...„..3

Introduction ............•.„„....„„•..•„.„••.•..„........„k

Review of Literature ...............„....•••...........,..6

Design...•.„„•„„•„...••.••••„„„•„•,.„„„..„..„•••........11

rrinciple of Opcratiom .............................11

Rotor und Jtator Assembly ...............„..........15

Rotor Drivc and iressure Housing ....,............,.18

Hectificr and Vacuum Tube Voltmctcr ...,..........,.19

Construction........„.•.„„•„•„„.••..„...„•.„„•.•........25

Rotor and Atator Assenbly..„.„„„.•••.•.••••.„„••...25

Rotor Drive und nrcssure Housing ..•.„.•„..•...„....25

Aectifier ana Vacuum Tube Voltnctor ................28

Assembly ....„..„.....••••„......•......„...........31

Calibration .„•......„„..„...................•••....•....32

Operation of Accc1c·ator .....„„.„.•.„....„.........32

Counting A;>pa.ratus .„............„„„„............33

Calibrating Fxpc·ircnt ....•„„•.•••.•.........„„•...3k

Results .„.......„„.•.„.....••„•„••„„.....„•........36

summary .•....„•„...........................•............NO

Acknowledgcscnta ...............................•..•„....k1

Bibliography......„.„.•„•„••-„•..•„„•••.•••••••„•••„„...h2

Vito

LIST Oi;' FI GUHES

Figure Page

1. Jitlplified Generating Voltmeter •.•••••••••••••••.••.• 14

2. Rotor and Jtator Design ···········•••••••·•••••••••·~16

3. Gim·.:lified Generating Voltmeter of Present Design •••• 17

4. Vol t:r:1eter Asset1bly Design ••••.••••.•••.•••••....•..•• 22

5. nectifier Design ····••••••••••······~~·····•••·•·····23 6. Vacuum 1ilbe Voltnetcr Design .••••..•.•.....••••.•.... 24

7, Rotor and Btator in Tank Reca!;s •.••..•.••.•••. , .•. ! •• 26

8. Helat1ve Position of I~otor and ;3tator with Terminal . . 27

9. Generating Voltmeter Bolted to Tank •••••••..•.••••••. 29

10. Generating Voltoeter Panel ..••..•..•.•..•...•...•.... 30

11 . Gamma Yield from Fluorine ....••••••••••.••••••.••... "3e

12. Calibration of Voltmeter .••••.••••••.••••••.••••••••. 39

-1+..

LHTRODUCTIOH

Becent ccmpletion ot the Virginia Polyteehnlc Institute

alectroatatlc accelerator makes possible the production of

a been ot protens with a maximum energy ot two million elec-

tron volts. It ls of particular convenienee to he able to

read directly the voltage with which the protons are accel-

erated. This ot course is also an indicatlon ct the energy

ot the baum. Iieasurasant ot electrostatic potentlals can be

accomplished with a gancratlug voltmeter. The purpose ot

this thesls is to deeeribc the design, construction, and

calihration ot a genereting voltneter used with the alco-

trostatic generator. The final value ot energy ie deter-

mlned by a nagnetic analyser.

The generatlng voltacter is a rotating grounded shutter

type developed tor electrostatlc gencrntora operating under

pressure. It essentially conslste ot two sectored vanes

located ln the electrostatic field. One ot the vancs is

grounded. It is rotated, alternately shieldlng the other

vane, generating an alternating charging current which is

proporticnal to the potential ot the high voltage source

creating the electric field. This current passes through

a rectitler clrcuit and the voltage developed across a re-

sistance is neasurcd hy a vacuua tube voltmeter. The netcr

was callhrated ln place utilislug encrgies ot knmn nuclear

-5-

ruunaucsa obtaimd by ganra-ray yiclris from bombardmant

ot tluorino with protonm

-6-

MEVILH UF LITKnAlU.m

Many generating voltneters have been described in the

literature. The original generatlng voltmeter was developed

independently by Gunnö and by Kirkpatrick and H1yake.15,16

It was developed in order to have a rugged portable instru-

ment for measurizg eloctrostatic fields. This type of in-

strunent proved very valuable in measuring electrostatic po-

t ntlals which were too high or where meusurcnent by other

m<ans was irpractical.

Uith the advent of thc high Voltage elrctrostatic acoel-

erators the probler of obtainirg a convcnicnt and reliable

mezsurenent of this high potential received considerable

attentl¤.2h•25 several methods co ld have possibly been

used. Axong these arc the methods of high resistance, at-

tr cted disk, proton range, generating voltmcter, and mag-

netic end electrostatic defleetion of ions. The high resis-

tanoe method reeulres a resistence difficult to construct and

ehield fror corona, ant probably results in non-linear rea-

ponse because of the size of the apparatus and the magnitude

of the Voltage involved. An attracted disk is subject to large

calibrati g errors because of the squared scale and the rela-

tivcly low available calibrating voltages. The nagnctic and

eloctrostatic deflection and proton range were felt to be more

conplicated than at first thought to be necessary,

-7-

Because it had been used with good results for a variety

of purposes6*9*15 ant becnuse of the advantxges offered by

its relative sirplicity and its independence of the ion beam,

the nethod utilizing a generating voltmeter was tried with

good Censequently, voltmeters of linear scales

and aecuracies ranging from 5 percent to 0.2 pereent have•-

volved for use xdth oloctrostatic gcnerators. The generating

voltneter has remained the primary means of measuring the

voltage on many ef the renerators for several years. With

the addition of concentric electrodes on the electrostatie

generators some other means of obtainlng the absolute voltage

have ha= to be devised siner the generating voltmeter would

only neasure the potential of the outer electrode. Also, the

desire for more accurate energy measurcments dictated the use

of other methods. Accur te measuroments are made today by

electrostatic analyzers er calibreted magnetic analyzers.

The geaeratlng voltmeter remains a verk useful tool with the

electrestatic generators, especially during initial focusing,

u

because of its continuous scale reading.

Because of its usefulness, it is well to review the

type of generating voltneter devcle~ed with respect to choice

cf parameters, d1ffie·lties found, methods ef calibrntien,

and the advant·ges found from the use of this type of meter.

The review is by no means cenplete, but contains references

to some typical voltmeters.

-8-

The parameters lnvolved in generating voltmeters es will

be shown in the next section include the size und shape of

the vanes and the speed of rotation of the grounding vane.

An early attuzpt by Van Atta, Northrup, Van Atta, and Van de

Graaffzs vas to shape the vanes es to generate a sine wave

current. However, the most expedient shaped vanes have been

sector shaped. The University of Kentuclqm and The Call-

fornla Institute of Technologyw electrostatio sccelerator

groups built voltmeters consisting of sem circular pole piece!.

Herb, Parkingson, and Ke:-et$•“•‘9 ef Wisconsin and Williams,

Hubevsh, um 1·at•2'I er Minnesota mut voltmeters er wo

pole sectored vanes. The c rrent from these voltmetcrs was

ccmrrutated by a sjmchronous meehanlcal rectifier and displayed

on a sensitive gelvanometer. Trump, Safford, and Van de

Gmaf!22 reported s four pole voltmeter the output of which

was reetified by s full-wave-bridge vacuum tube rectifler

and displayed on a mleroammeter. An in ovation of the shut-

ter type voltncter is the cue developed by Haxby, Shoupp,

Stephens, and Wel1s8•26 using the DTL ciple of compeasating

voltages described by Harnwell and Van Voorhis.7 In this

voltmeter, the generatlng voltmeter itself is used onl;· to

detect the absrnce of lnhonogencitles in the field behind a

voltege plate. Voltage is apwliod to this plate until it co-

incidcs with an equipotential surface of the field frnn the

high potential elec‘rode. This condition is detected b· the

•9-

abscaoa ot plokpup in tha ganarating voltnatar part. Tha

high voltaga la than proportional to thn voltaga on thn

voltagn plata and is mcaaurad by nnans of prnoislon r•a1s•

tore and a mllliannntar. Tha alu ot thn vanas of th•s• vas~•

loua voltuatnrs rangns h·¤¤ about twnlvn lnahas to thran

inchns depandlng on thn sinn of thn generator. In nost

••s•s thn rotating wann ia npnn hy althar an 1800 r.p.a. or

a 3600 r.p.¤• synohronous motor.

Dltticnltlns arislng with thosa voltaatars ars those

usually assoolatad with location. Ditticulty has bnan notad

man thn voltmatar ams looatad so that it night"s••"

a

chargod lnsulatad anrtaoa or mara thn tlald ie large enough

to panit •park¤v•¤·.25 Sonn ooamaxats hava han aadn on thn

nttnot ot oorona ouzrant. Tramp, Sattord, and Van dn Graatfzz

ahownd that thn artnet ot corona onrrnnt ln thalr dnslgn us

nagllgihla tor thn prassurn lnenlatad generator. Herb, Park-

lngson, and I•rst“ toamd that tha suaaltlvity o! thn ult-

aator changad radloally it the vanas warn ohangad alightly

with rnspaot to thn tank wall.

Tha most axpadisnt way ot calibrating thn voltmntar la

to apply to thn high voltagn alactroda a potential mich aan

ha maaaurad hy soma othor aaana. 81nc• thn voltmntar in

linaar a ralatlvnly small voltags can ba appliad and nau-

urnd by a proviously oallhratnd high rnsistanon. From this

-10-

a voltage sensitivity car bc est·blished. This is the method

used by Trump, Jnfford, and Van de Graaff•22 Another standard

method is to cvmpare the veltmeter reading with the value of

the terminal voltare at known nuclear rosonxnces. This method

has ben used by a nunbwr of grou;e.?•11•17•27 The linearity

has also been checked by utilizing the double and triplc

energies needed for diatomic and triatomlc ions to undergo

the sans reaetions as monatcmic lens.I

The rain advantage: of thc renerating voltveter have been

found te be as follows. The voltnetar is inve endent of the

ion beaa. The voltretrr Lrains no current from the source of

high voltage. It has a contlnuous scale reading. It is rela-

tively ruggcd. The indicating meter is at or near ground

potential and can be located at a aistanee from the generator.

•11•

DESIGN

The generatlng veltueter coneiste easentially ot an

lneulated atatcr member mounted in the ground plane ot the

tank racing the high Voltage terminal fron mich it 1:

periodlcally ahielded by the rotation et conetent speed ot

a grmmded sectored disk. The stator-to-terminal capaci-

tance 1: thus caused to Vary periedically and the induced

etator current, mich 1: a aeasure of the terminal voltage,

paesea through a rectlfler giving rise to a Voltage acroee

a resletance B. The d.o. potential developed across thia

reelstance ie measured by a special vacuu tube voltneter.

A einplified circuit tor a generating vcltneter ls given

In Figure 1. In this figure and in the analyele below V

ie the ccnetant potential to be meaeuredg G is the period-

ically varying cepacitance between the high Voltage terminal

end the stationary etatcr, the naximun Variation ct which 1:

G'.

The cycle ct operation 1: aa followe. A: G increase:

Iron its nininum value, the etetor ie charged through the

rectirler T, by the influence of the potential V until G

reachee lt: naximun Value. Then ae G decreaees, the etatcr

dicahargea through the reetitier T2 and ma p•e1,;zg¤¤• 3

until G again reaches ite ninlnun value, whcreupon the cycle

-13-

is ropeatcd„ It is evident that the current ln R le uui-

directlonel and is independent of the polarity cf V. It

is of interest to note that the total charge transferred

over one cycle is zero. Theretore, the voltmoter draws no

current frca the high voltage source, but rather the power

used for the cyclic transfer of charge is supplied by the

nechenical force varying the capacitance G.

The circultal relation during the last half of the

cycle ie

V ¤ E•

LB,G

wheree q ¤ the charge at any time on G,

1 ¤ the instantaneoua current,

V•

the high voltage potential,

C ¤ the etator•to-terminal capacitance, and

R ¤ the voltage dropping reeistanoe,

aeeumlng negligible rectitier voltage drope„ This eqnatlon

ie difficult to eolve ainoe G is a function of tue. If,

however, it is noted that LB is very srall in comparison with

V (a departare of about two parts in one million for the

present voltneter) the relation may be written with good

approxlnatlon ae

V e ,

-13•

D1rferent1at1ng with renpect to t1ne,

2. g 5- V,

dt dt

°’•

1 dt ¤ GC V.

1ntegret1¤g ever cne halt period, 1/2, the •quat1o¤ becceen

‘!0

31 dt • jd!} V.

1/2 a,

the left eide cf thin equation 1n einply the total charge,

Q, wh1ch flcwn through the r•n1ntence K dur1ng thin helf

cycle. Therefore,

Q ¤ •¤,V.

hat beceune the c1r·cu1t 1n ennentielly e he1f•weve rect1f1er,

thin 1e eleo the tctel charge through I d¤r1ng e ccuplete

cycle cf cpereticm Then d1v1d1ng by 1,

Q ¤vV

'{ '°T‘

Since Q/1 1n the current, I, in B during ue eycle end

I/T ¤ fc, the verietion frequency of the cepecitence ¢, the

voltege ecrosn the ren1stnnce B 1n

IH•

•l£°6'V„

‘!h1n reletion nhcvee that the vcltege ecronn l 1e propcrticnal

to fo end tor ccnetent frequency in proport1enal tc the potte-

@

-15..

tial V. By reading this Voltage dlrectly with an appropriate

Veen: tnhe Voltaeter, the high Voltage potential V can be

read directly.

Hotor and Stator Alseahl!

A

ASuce the lnduced statcr current ls proportional tc the

frequency cf the variatlon ct C, lt is desirahle to dlvlde

the atator menher into several peles so se to lncreaee the

nagnitude cf the lnduced current: tor a fixed speed ef the

roter. The period of Variation ot the capacitance should of

course hc large in comparison with the relaxatlen tue cf

the circult. he stator cf the present Voltaeter ls dlvided

inte light

‘•5°

secters. Alternate stator sectcrs are elec-

trically connected to fern two statcr groups cf four polel

180 electrical dogrees cut of phase with each other. The

tour pole roter ls e conplenent to each of these state!

groups. Figure 2 shows a detail drawing cf the roter and

etator. It may he noted that the induced stator corrnt

will he eseerntlally a perlodlc roctangular wave of 180•1•••

trlcal degrees. With each etator gro-ap provided with an in-

dependent rectlfier systen as ahown in Figure 3, e s¤hstan•

tially constant rectitied current flows in the reslstanoe B.

. since the frequency ot Variation of the cepaeitance ls ese•n•

tially twice that ct the previous analysis the current output

ls doohled.

@

~

'

'

HIGH VOLTAGE ELECTRODE

STATOR NO. 2

I R STATOR NO. I

SIMPLIFIED GENERATING VOLTMETER OF PRESENT DESIGN FIG. 3

•18•

It is also desirable to make the etator•to-terminal

capaaurnee as large as possible to increase the magnitmde

of the. current generated. The capacitance 1:: determined

prinarily by the size (area) ot the stator and its dlstance

from the hlgh Voltage terminal. These two paraneters, how-

ever, were fixed by the geometry of the pres ure tank and

high veltage terminal. The distance tc the stator from the

high voltage terminal was determined by the relative die-

tance between the accelerator electrode am the ground plane

of the apex of the tank. The size of the stator was 11: ited

by the *+.9 inch elameter recese aporture in which the stator

is located.It

has been noted that the generating voltneter output

ls proportional to the high voltage potential only if the

frequency ef the periedic st¤to1·•to-terminal capaeitance ia

constant. This is accomplished by spinning the roter with

a synchronous motor. For thc present voltmeter a Bodine,

1/20 h.p. 1800 r.p.m. synchronous motor is used. Slnce a

motor cf this capacity will not fit into a three inch wald-

ing well opening to the pressure tank, it was necessary tc

design a rotor drive in which the motor is located outside

ef this well. In order to oomgletely pressurize thc entire

geaerating voltmeter, the motor is located in a flange and

•19•

plpe pressure chanber which bolts onto the {lange of the

weldlng well. he rotor ls atfixcd to the end of e double

ballbearlng shett arrangement which is located down the

length ot the well. The shaft ls driven by the synchronuue

motor, being coupled by a flexible coupling. A cross-see-

tional detail drawlng of the voltmeter is shown in Figure B.

Complete detail drnwings of the generatlng voltmeter are

lnserted in a pocket Ln the beck cover•

Since lt ls deslrable to take advantage of the linear

scale ot a direct current meter, it is necessary to rectify

the generated sigel {rom the voltmeter. Referring again to

Figure 3, it can be seen the rectlfier circuit is sie-

ply a t*¤ll•wave-bridge type. A t*ul1•wave•bridg• reetitier

was designed utilising the high reverse resistance of e

vaeuun tube diode. ho twin 6% diodes were used with the

heater current limited to about one half rated value to in-

hibit the {low ef emission current d··1·1ng the non•conducting

part et the cycle. The filement supply to the rectltier ls

controlled by the Vacuum tube voltmeter power switch. Fig-

ure 5 shows the diagram tor the rectitier.

The vacu m tube voltmetor shown ln Figure 6 is based

on e difference anpllficr circuit. The voltncter power sup-

ply ia a conventiohal full-wave VH tube reg elatod supply.

•20•

The output is read on a meter brtweon the pl tcs of the tue

trlodes ot a 6SH7. The 5000 ohm variable resistaace ln the

plate circult serves to balance the plrte voltage ot the

two tubes so that ne current is indicated by the meter.

The rcctltled output signal from the geaeratlng voltmcter

ls developer across a 3.1k negohm reslstance. Thls resis-

tance is connected to a switching arrangement so that either

all, one half, or one fourth ot the total voltage drop across

the rcslstance ls a plled to the Frid ot one trlode glvlag

relative voltage raages ot one, two, and tour resgectlvoly.

An ldutlcal reslstance switchlxg arrangement ls ln the other

grld clrcuit so that ia swltchlng fror one range to another

the zero dritt ls mlnimlzed. The apgllcutlon of the signal

to the grid causcs an unbalance between the plates and the

dlttereatiel plate current ls given by

ߤi'

1 =———————"”——"“““

1

Hz + rp(2 + Rn/R1)

where: p = gala ot vacuua twbe,

e = voltage ay lied te grld,

Rh = reslstance ot meter,

rp = plate rcslstance of tube, and

R1 = loan reslstance.

This shows that unter preyor operating conditions the output

of the vacuum tube veltreter cerresroats directly te

'c”

and

heace by previous analysis ta the potential of the hlgh volt-

-21•

age el•ot1·ode•

The meter between the two plotes ls e 0•100 nicroas-

neter movement shunted so the output ranges correspond

closely to ne, two, and tour nillicn volts• Further call-

bration is achdeved with the adjustable resistanees marked

"Gal C" and 'Cel I". Since lt is not known which setting

corresponds to one, two, and tour million volts full seele,

the meter clreuit contains a swltching arrangement so that

two different calibratlng resietances nay be pleced in the

cix·cuit„ Iith the 'Scele Selector" switch in the "Ce1" pc-

sltlm, the resistence 'Cal ¢" ls adjustcd until the output

is thought to be such that a full scale reading will corres•

pcnd to one alllian volts. It is in this position that the

voltueter ls ce.l1brated• Atter celibratlon, the 'Scele

Selector" is ewltched to the "lull" position end the resis-

tance "<2el F" is adjusted until mll scale deflection cor-

reepuads to one nillicn vo1ts•

It was noted that the rectitler developed a contact

potential end that e eonstant signal across R was always

present. Ihis seall current ls coupenseted hy with the

nero edjustnent„ However, this makes lt impossible to have

a eonventional nero check bv simply lnterrupting the signal

tree the generatlng voltnetcr. A sero check is provided lv

e switch in the voltage supply to the synehronous motor.

v l/ / l PRESSURE I HOUSING TANK WALL /

STATOR \\

WELDING WELL I cCJVL.

, ' I

\

I

I I

SYNCHRONOUS MOTOR

" ' '\_,

"

ROTOR DRIVE

' 0-RING GASKET r/.' '~ \ I \

BUSHING

VOLTMETER ASSEMBLY DESIGN FIG. 4

0-RING GASKET I

I I\) I\) I

STATOR NO. I

•

VTVM I I

:r=i-11 -=-1 I

.J L Ir. __ Ir ---------11 11---11 11

JONES PLUG

6.3 VAC

RECTIFIER DESIGN FIG. 5

----il

:~IL_ ___ _ 11 , , , ,

f

STATOR NO. 2

I f\)

l,v I

5A (

115 VAC

5V 3A

120MA 350-0-350 V

.3V FIL

6H6 1S

JONES PLU G

2A

TO SYN MOTOR

NORMALLY CLOSED 11ZERO CHECK"

1.57M 1%

.787M 1%

.787M 1%

IOH

160 MA

8}J F 18UF 600V 600V

VR 105 VR 105

5 K

' 18 K-2 "ZERO ADJ°' 18 K-2

37n

IOK 11CAL1 F

IOOµA

' /

" /

' I

GEN VM

I I _c-i I r~ - II --r--

1 I 1

I I I I

1.57M, 1%

\ ----/ \_ , / 6SN7 \ -; / / I "--- / '.._ __/ I

.787M 1%

L ------------------ ___ _J .OlµF .01 µF

(M) 220D. (M) _ .787M ~ 1%

VACUUM TUBE VOLTMETER DESIGN FIG. 6

•

-25-

roter and etater are made er 1/16-lneh elunlmn.

The elght neben of the eteter are fustened te e 1/2-inch

bakelite lnsuleter by two screve each. Theee sorewe serve

also es electrloal contacts vlth every other uenber cen-

nected and tvo leads away from the stator. The bekellte

lneulater ls eeeured te the front bearlng housing ot the

roter drive by four aerews. The roter is eftlxed to the

shatt h a renevable oollar. The roter le adjueted until

tuereterie|ll¢tlybeh1ndthegrexmdpl.aneetthetenk

wall. The spaelng between the etetor and roter ls about

0.1 inch. The vhele roter end stator aseebly la reeessel

lnanepelng ln theapexot the tank. Because et thlsre-

eese the roter and steter nust be meunted Iren inelde the

tank. Figure 7 shoes the roter and stator ln this r•eeu•

Figure 8 shows the relative pesltlon ot the stetor end the

high voltage terminal.

A V

The roter drive end preseure housing ae were all the

other parts ot the voltmeter were naehined in the Virginia

Pelyteehnle Institute Physles Department shop. The rote!

drlvlng ehett and bearlng heusings are nade ef steel. The

bearing heuslngs are eeparatedhy a thin vall bress eyl£n¤•r•

~

ROTOR AND STATOR IN TANK RECESS FIG. 7

I I\) 0\ I

~

I I\) -...J I

RELATIVE POSITION OF ROTOR AND STATOR WITH TERMINAL FIG. 8

•2F-

The synchronous notor is oncloscd by a welded steel pressure

chambcr. The charbcr is fastened to the tank by four 3/N-

inch ca; screws. Pressure seals arc schievcd with *0* ring

pressure gaskets. Figure 9 shows the voltreter assenbled

and holte to thr tank. Four clectrioal connections are

brought out through the front ·late of the pressure chamber.

These bushings were soldered into place after welcing by

preheatiug the complete plate. Access to the bushings is

ma.e by 1/¥·inch pipe plugs.

Rectifier and Vacuum Tube Vsltxeter

The rectifier is nounte= on a senicircular chassis lo-

cated oa the pressure chamber. The signal from the restl-

fier is carrioe to the vacuum tubc voltretcr bv a low loss

coaxial cable. The power supply for the synchronous motor

und filament supplt for the rectifier are brought from the

Vacuum tube voltmeter loc‘te„ on the remote control console.

A view of the rectlficr may bc seen in Figure 9.

The vacuu tube voltneter is located on a control con-

sole so that the meter may F< reae easlli by the operator.

It is of conventional _anel and chassis construction and is

rack nounted. The range selector, scale selector, zero avjumt,

and zero check are located on the front of the panel for eas

use. Figure 10 shous a view of the console panel of the gen-

erating voltneter. The calibrating rzsixters are located on

the chassis and can be adjusted from the top of the rack.

GENERATING VOLTMETER BOLTED TO TANK FIG. 9

I f'\)

'° I

' ®

®

@

-@

---•

_. \

#"'"

- 0-.

,~ -

a II • .... - .. ' •' (.. (

1 •• _ _,,. -. -- - OCMI IIUCTOO

GENERATING VOLTMETER PANEL FIG. 10

.

•

..

I vJ 0 I

-31-

Assembly

Because of the unique construction of the penertting

voltnetcr the following assenbly procedure is used. The

motor is provided with a clarying ring and is end nounted.

The motor is mounted first, making sure the two lends to

the stator are brought out the front of the chamber ·ad

the two leads to the motor are brought back around the motor.

The flexible c0u;ling uring this time should bc secured to

the shaft of tho motor. The roar bearing housing is next

fastcned to the chambcr, shaft inserted into the counling,

and the cylinder und front bearing housing secured in place.

The cou ling nov may be tightened with an A11en—head wrench

et the opening provided in the rear bearing housing. The

voltrcter may new be boltcd to tnc tank und the roter and

stator ettnch«d fror inside the tank. The voltmeter should

be disasserbled in reverse order.

-g2•

CALIBLKATIOH

Although the generating volhaeter reading is independent

of the ion bean, it ls necessary for ealibratlon purposes to

bombard e target with protons. For this reason a brlcf de-

scription of operation of the eq zipuent ia given.

The first step in _;utting the accelerator into operation

ls to initiate a dlseharge in thc ion source. The lens pro-

duced are cxtracted by application of a 'probe" voltage ln the

dlscharge vessel. The spray voltage is now ra! sed until e pre-

determined voltage between terminal and ground is brought to

equilibrium, with the amount of charge being brought to the

terminal by the charging belt equal to the charge being lost

by the terminal. The ion bc··x is now b·-ing aecelcrnted denn

the tube with the chosen energy. The beam is allowed to fall

upon e quarts disk where it is focused by an eleetrostatlc

lens et the high voltage end of the acceleratlng tube. The

beam mst now be analysen, that 1s, separaten into mass one,

uns two, etc, components. This is done with the magnetic

analyzer•‘8 With propcr adjustment and a reasonable current

thrc gh the magnet eo that the dcsired mass number will be

beat by the magnet, the beam she ld now be at the slit box.

With the bcam implnging on the sllt jaws the coronn feedback

stabilizationl holds the beam on target with the eeslred en-

·33¤-

erg. The spread of energy on the target is nov a function

of the sllt width thro gh which the been passes and the dls-

tance from the magnet to the sllt• The approximate energy

ot the beam is nov known h-cn the knowledge of the nagnetlc

field es determined by the currentJ8 It is now desired tc

determine more exactly the energy correepcmdlng to the gen-

eratlng vcltmeter reading at this vcltagc„

Counting ggaiatul

A

prctons fall upon the target a reactlon may

take place which results in the mission er ga¤ma•reys. It

ls desired to find the "relative yield" of this reactlon,

that is, a number proportional te the ratio of the gonna-

rays euitted from the target te the number ef bombardlng pro-

tons. It is then necessary to measure a quantity proportimal

tc the number of gamma-rays emltted and also a quantity pro-

portional to the number of protons lmpinglng on the targ•t•

A scintillatlen counter was used to deteet the games-

x·ay|•This counter consisted of a so<Eiu¤ iodlde crystal13

end e 6292 Dumont photemultiplier tube. The seintilletiens

caused by the gam:a~rays are arplified by the photomultiplier

tube and the output pulse, after suitable ampliflcation, was

applied to a decade scalcr. Th: dlscrininatcr of the ampll-

fier was set so that the x-ray background ot the accelerater

was not detected. 1'nerefore, the 'background" radlatien wu

-3b-

redueed essentlally to that due to cosnlc radiation.

A qunntity proportional to the n·rher of bombarding

pretons was obtained with the use of a "current integrator'.

The ”current integr tor" wa: of the neon glow t pe. lt een-

sisted of a 150 micrenicrefarad cupaeiter, a 1.2 nagohm re-

slstor and a neon glow lamp. The charge which accnmulated

on the target charget the capacitnnce until the brcakdown

potential of the neon lanp was reachot. The glow tube then

“firod', discharging the capacitor through the reslxtanee,

giving rise to a pulse which was proportional t the charge

ac umulated. The cycle was then repeated. The output pulse

was fad through a preon lifler into a decodc scaler.

Te obtain the rrl tive ‘1eld, it wu: only necessary to

allow the pulses from thc xclntillation counter ana eurr¤nt

lntagratcr to register sinulteneously for n s ecific length

ef tina. By diviping the number ef ¤cale« scintillation

count: by the current integrator counts, the relative yield

is obtained.

Caligratigg Exrcrircnt

Host of the light olments are known to yield gaxua-

rays when they are bemberded with protoas. The intensitr of

gunma•ray radlatlon free next light elencnt: exhibit reso-

nance phenomena; near resonance the amount of garca-rar

radiation is much grratrr than fer neighbering energiee.

For example, in the bombardhent ot fluorinc with protons,

-35-

resonances are known to occur at encrgles of .3*-60 Mev and

.*+86 Hev in addition to other well known rescnnnces.22 To

lllustrate hxrther, examining the yleld from protons on fluo-

rine, it would be expected that the yield would increase with

lncreasing energy until the energy ot .3*+0 Hev was reached.

Increesing the mergy beyond this would result in decreasing

yleld until the next resonance ls approached. This of course

ls the case it the target is a thin target. Let us consider

what would happen lf the target is thicker. If a proton hav-

ing an energy slightly greater than .3*+0 Hcv lmplnges ax such

a target lt will have a lover probabillty of being captured

by the first layer of ateuxxs than one of energy of .3*+0 Hav.

Ext as lt penetrates the target lt will lose energy by calli-

sione with electrons and the probebillty of capture increases.

The net result than is a shitting of the aaxlmxm of the yield

curve to higher energies. If the target ls thick enough to

stop all the protons the yield curve will edxibit no naxlmm

at all but will level off at the highest value of the yleld.

The position of resonance is taken to he half way up the yield

curve. These so called "thick target" yields arc well known

and are suitable for calibraticn purposes.

Thick target yiclds of tlxxorine were obtained and the

calibration ot the generating voltneter made from them. A

typical yield run was made in the following manner. hu gen-

-36-

erating vcltmetcr eelcctor switch was placed in the "Ca1"

positim. Alter a. fefinito recycling pattern ef the magnet

to minimiu hysteresis effects the ren net current was brought

to a dcsired current to bend the mass one c ¤„omnt of the

beam with approxinntely e predetermined energy. The vcltage

ol the accelerntcr was rai:ed until the bcen was seen to be

lrzplnging on the target. Slxmltaneous ecuating ef game-ray

end lntegrator ccunts were made tor two minutes and the rele-

tive yleld oalculated. The gcncratirrg voltneter was read

during the time period. The magnet current was read by

reading the potential ..rop acrosu a standard .01 ohm resis-

tance in series with the nagnct with a Leeds and Zlorthrup type

K-1 potentiometer. hc magnet currcrzt was then ralsed .05 ul-

peres und simultaneous reading taken again. This process vu

repcated until the :.eai;~c<; yield spectr n uns cbtained.

Buulu

T’he .3*•0 Hev and .*+86 Mev resonances ol fluorine were

obtained. The yield curve nay be seen in Figure 11. The

.3*•0 Hev und .*+86 Mev 1·e.;onancos corras ond to generatlng

voltmeter reading: of 36.0 und 56.3 percent cf scale range

on the one million vclts range. Since the voltnetcr should

be linear it was felt that it was only necessary to obtalu

two p.·1¤ts tc establish calibration. The calibraticn curve

ot the generetlng voltneter may bé seen in Figure 12. M

-37.

use of this curve a reiterstlon process was made by |witch•

ing from the "Cal" position to the "Pull" position of the

scale selcctor and the meter was sdjusted umtll full scale

deflection in the "Full" position sho· ld correspond to one

million volts. This curve may also be seen in Figure 12.

The dotted line from the origin to ovne million volts is the

idealized calibrstion curve. The distance between the cell.

brated "full scale" curve and this idealised curve represente

the actual error reading interpretlng the voltage nom the

meter directly. This error should amount to within five

pcrcent ln half scale deflection with decreasing error fo!

higher reading on the one million volts range. On the tvo

nilllon vclts range the error should not exoeed plus or minus

three perocnt being a minimum near half soale• If it ls

necessary to recallbrate the meter, good results should be

obtained ln the following manner. With the been on target

and the current in the magnet 5.0+ amperes, the generating

voltmeter ehe ld read 38 and 32 percent scale reading om the

'Gal" and "Fnll" positions ot the seele selecter respeotively

on the one million volts range•

@

@

#+0-

Hüllen!

A generating vclueter capable ot measuring one, two,

or tour mllllm volts has ben designed and oonetrueted tor

une with the Virginia Polytochuie Institute elaetroetatie

eceelerator. The voltmeter ie e grouuded ahutter typo, the

reetitied output ot which is measured by e vecmm tube v¤lt·

meter. ‘l'he voltmeter wu calibrated by kamm uucleur reso-

mmcoe of fluorine. '1'ho callbretiou showed the meter to he

eocurate to within tive peroent at hal! ecale deflectim ¤¤

the cue mulln volts range md lese than plus or minus three

poromt cu the two million volts range,

.1+1..

aßllultäßünuütl

The author wishes to thank Dr. T. N. Hahn, Jr. tor his

inspiration, guidanee and gansrous advice during thispro•

jeet. He also wishes to express his appreoiation to ,- 1

-tor his many untirlng oonsideraticms. He especially

wishea to thank-tor his advice sad hal) in

the aaohining ot the voltneter. Since any work done on the

aooslerator is always the result of group effort, it is in

this oapaoity that the author wishes to thank all the staff

and graduate school uho have helped in this initial effort•

.b2.

BIm.10GRA1HY

1. Ball G., Unpublished H.S. Thesis, Virginia Iolytechnic

metlzuze, 1 57.

2. Bonner, T. W. and Evans, J. E., Phys. Rev., 73, 666

(19k8).

3. Ghao C. Y. Tellcstrup, A. V. Fovlcr W. A. und

Lmalzsea, Ö. c., (bye. Rev., 79, 108 (1950).

M. Elnore W. C. und Jands, H., Eloctgonics, p. 52, McGraw

H111 (19M9).

5. Gernt E. J. Herb R. G. and Parkingson D. B. Phys.

¤.v.,'sM, 6M1 (193f).’ ’

6. Gunn, Ä., Ihys. Rev., H0, 307 (1932).

7. Harnuell G. I. and Van Voorhis, S. N., Rev. Sci. Inst.,M, 6Mo (1933).

8. Haxby, R. 0., Shoupp, U. Y., Stcphcns, H. E. und Wells,

I. H., Ihys. Rev., 56, 1035 (19h0).

9. Henerson, J. E., Goss, V. H. and Rose, J. B., Rev. Sci.

mst., 6, 63 (1935).

10. Herb R. G. Kerst, D. W. and Hcxibben, J. L., Ehys. Rev.

:1, 691 (1937).

11. Herb, R. G. Iarkingson, D. B. und Kerst, D. U., Phys.

Rev., 51, 75 (1937).

12. Herb R. G. Jnowdon, J. C. and Gala, 0., Ihys. Rev.,

vs,§¤613.

n¤rsta6ex·, R., Ihys. Rev., 75, 796 (19*4).

1h. Korn, B. D., Private Corvunicatien.

15. Kirkpatrick, I., Rev. Jci. Inst., 3, M30 (1932).

16. Kirkpatrick, 1. ana Hiyako, I., Äev. dei. Inst., 3, 1

(1932).

17. Lauritscn, T., Luuritsen, C. C. and Fovlcr, H. A., Phys.

Rev., 59, 2h1 (19%1).

J;}-

18. ¤11v•r, ¤. H. ¤¤,)uh11shs«1 1:.8. Thss1s, v1x·g1n1a Po1y•

t•chx11c Inst15ute, 1956.

19. P *.1.11 ·>11 D. B. Ilcrb :1. G. G ::11; E. J. and

raäiibbä, 5. L., Ähys. Äsv., 55, ghz (1938).°

20. Wan, J. L., Uznpublishcd M.S. Thesis, Univusity of Kan-

tucky, 1952.

21. Thomas, H. A., Rev. Sci. Inst., 8, @+8(1937).

22. Tramp J. G., Sattord F. J'. and Vand•

Gruft, H. J.,

nev. 5¤1. mst., 11, 5*+ (who).

23. rz-up J. 0. ma vu.a•

orssrr, 11. J., Phys. nw., 55,

1160 (1939).

zh. 1-WVG M. A. Hsrszsd, L. R. and Dahl, 0., Phys. mv.,*+8. 515 (1955).

25. Van Atta, L. C. Horthrup, D. L., Van Ath C. H. lhd

vu. as 0;·ss::, Ä. J., Phys. mv., h9, 761 (1936).

26. Hella, W. H., Hexhy R. 0., Stephans, I. E. md Shoupp,

w. 1:., Phys. mv., 58, 162 (19*+0).

27. vI1.111sms J. J., Brmbaugh, L. H. amd Tsze, J. F., 13,

202 (19*+2).

The vita has been removed from

the scanned document

,-1'I0-24NC f'oEEP- 4 HOL[S

1t19 DRILL - 4 HOLES

CLkM, :;c c,iNG r I H.L RE(iD . "

I .. 4-

11 I

@'; ~ CC .. L:.R

STUL •'. REQ'D

- !J 16

.. .1. ... •

- .J...ORILL-2 HriLES 16

-l~ j

• Q CHAMFER f2. .. 4~•

SHAFT S. COLLAR ASSEMBLY

NOTE· DRILL a TAP COLLAR AND • SHAFT AT SAME TIME

II 32

/ ~33 DRILL- 2 HOLES

ROTOR ALUMINUM I REQ'O

t 22f a,

-. 15• -'-, ~ 16 I

•33 DRILL-C'S'K 225 DIA -2HOLES

;( STATOR i ALUMINUM 8 REQ'O

VIRGINIA POLYTECHNIC INSTITUTE

SCALE:

DEPARTMENT OF PHYSICS

f ULL

~::-:=;:-;;------i GENERATING VOLTMETER

DRAWING 2 OF & JANUARY' 19&1

~

t-16 ~C- 4 HOLES EO SPO

•

~ORILL-TAP,iPIPE TH0-4 HOLES.,.

2

'* 19 DRILL- c·eoRE !/(i DEEP .., TAP.; PIPE THO -4 HOLEf

STEEL I REC'D

c·s·K e2",-l. DIA -4 HOLES 16

-~l

-IN <O

-IN

FILLETS 8. ROUNDS __LR 8

-IN ,n

-IN .

CYLINDER STEEL 1 REQ'O

0

1·, \.

~ '

i'oRILL- 8 110LES EQ SPO ~ I ~

.. ~ !....

- /

32

f--. t .. -

I I. t. ...

,n N

<D

0 0

"'

I

-IN ,n

" j I ' p 1

G,- 'tl T

REAR PLATE STEEL I REQ'O

•

J-----.---,--~-----~1 VIRGINIA POLYTECHNIC INSTITUTE DEPARTMENT OF PHYSICS

- - - - -GE"JERATING VOLTMETER

__ ~CALE- ~-;~_1. __ D_R-,;l~G 3 ~0~5 ]_ JA;~~Y, _1958 - ;;-

foRlLL-BHOLES EQ SPO

- I '"

"~) (,) (:)

.. \ ... 2

(J () .l !

PRESSURE HOUSING ASSEMBLY 0 8

TOP PLATE STEEL I REQ'D

I VIRGINIA POLYTECHNIC INSTITUTE l DEPARTMENT OF PHYSICS

-------------~---- -· --

~

ART f N~'J ART

'-,PH.) 2 56~" 8 F._A· HO ~E.i, 2 NO 4 40• "4 F HO SCREW

3 N 1.1 4 40• /2 F H.) SCREW

4 N..1 8 32• 518 RD HD .SCREN

5 ) NJ 8 J... 11& RO HD S..:RE N

6 NO I.., it4 • 518 FIL HO SCREW

7 NO O l <t -. 3 RO HO SCREW

+

' REO

r 8 314 16 ~ J/4 H[.( H.:) AP SCRE"" t,-

9 3/4•16 ~ 3 HEX HO BULT

0 314-16 HEX HD NJT

114 PIP[ PLUO

12 NO 2 ~6 HEX HO NJT 32

+-1

PAR•L NAME '.)F FAR r ,, ~ COLLAR

16 t- "'ISULATING RING

1 FRON"' BfARl,l,,IG HOUS,NG

18 + REAR BEARING HOUSING + 19 CiLINOER

2 0 BALL BEARING

2 I SHAFT

2 2 t- COUPLl-.iG

23 t- CLA'-tPING RING l41 SYNCHRONO..,S MOTOR 2e PRESSURE HOUSING

26 TOP PLATE

27 0 R1NG GASKET

2 8 BJSHING

1120 HP

II

16

,, 25

~) 26

r j

>

i r

J J'

t J 1 r

11 27

J 1 20

17 , 11 rl t ,,, . ""

0

VIRGINIA POLYTECHNIC INSTITUTE DEPARTMENT OF PHY' :s GENERATING VOLTMETER

SCALE 1/2 DRAWING e Qt= 5 1 J~"4JARY, 1958

~

FRONT STEEL

f ,., OJ ,,..

"' .. ,...,

~

I

j __

.1

BEARING HOUSING I REQ'O

r ,----'

•

CYLINDER BRASS I REQ'D

.. 19 DRILL 4HOLES EQ SPO \

rCHAMFER ~2·•s ·

3 3 8

"' "'

--=--:i

I ••O 24NC~2 DEEP 4~0LES EQ SPO

I· I

2' .

! DRILL

"' .,., "'

REAR BEARING HOUSING STEEL I REC'D

1, 1. ,-\t'-3~ DRh .. i.-C'80REi10!:!P- 4 HOLES

.-r44 ORll: · 6 HOLE.S '7 2 ' 7 2 30 ~ t'-h·t- 30 .. ~ 1

' " I I

15' ~; I . 15' r · 1

• I . ' '\ f

I 2' I ,. 1.-4--~ • 30' 30·

k~ I . ' <, 5 ,

"' a ·<,: \ ··IN "' I 'ct'\ 15' I ,I, !).:." I 15'

r ~ 1 L1~ \ .l'....; /

\ I . ; l I

.lO' :-. .,~ (· 30' ~ _1

~ ~1 \ ' i ' -,,.

/'"' L. _j I .... ~ DRILL

,o• 30 INSULATING RING '5~ _..

B~~fil:.1JE _I_ R~(fO

r I VIRGINIA POLYTECHNIC INSTITUTE l DEt'ARTMEN T OF PHYSICS

i GENERATING VOLTMETER

.1' ~CALE: FU LL l DRAWING I OF 5 [ JANUARY, 19~8