1947

2 downloads 0 Views 319KB Size Report
High precision of about. 0. 1 -nm is required in the beam posit ion measurement of TRISTAN MR(main ring) which is now under cons true t ion. In this respect,.
IEEE Transaction!

© 1985 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE. on Nuclear Science, Vol. NS-32, No. 5, October IYXS REFIGEMENT

M.

I947

PROCEDURES OF BEAM POSITION MEASUREMENT IN THE TRIST4N ACCUMULATION RING

Tejima. H. A. Ogata.

lshii, T. T. Ieiri

National Laboratory for Dho-machi, Tsukuba-gun,

Shintake. J. Kishiro, and Y. Mizumachi

High Energy Ibaraki-ken,

Physics (KEK1 305 JAPAN

Summary High precision of about 0. 1 -nm is required in the beam posit ion measurement of TRISTAN MR(main ring) cons true t ion. which is now under In this respect, position measurement in the TRISTAN AR(accumulation ring) is reviewed. We found that the statistical study of the accumulated measurement data gives vary useful information on the performance of the monitor system. repeated COD measurements during a beam For example, storage showed not only reproducibility of the position measurement (0. 05 mm) in the beam intensity range of factor 50 but also pointed out troubles caused by poor Moreover. it contact in coaxial switches. suggested the criterion to abandon erroneous data. Programs for such check has been prepared and has improved the efficiency of COD correction. Improvements have also been made in the calibration and in the mechanical setting of the monitor chambers.

tn t roduc The started position performance. accumulated and the

t ion

operation of TRISTAN accumulation ring (AR) at KEK from November 1983 (1; and the beam monitor system has worked with good During these two years. we have experiences which are useful for the design construction of the TRISTAN ,na;n ring (MR). Position

monitors

in

AR

(2,3)

We have 83 monitor pickups for orbit measurement at the side of a Q-magnet. A in AR, each one sitting position electronics system is set in each of four local control buildings on the AR ring. Each one covers one fourth of monitors (20-22). A monitor chamber is just a part of an alumtnum bean duct where four pickup electrodes are welded on to the wall as shown in Fig I. (,I> We use supor -heterodyne de tee t ion scheme and pick. up a harmoni*,s of beam revolution.

Fig

2

Setup

tor

represent

it

by

X= 2 Px(i,

j)*NxL*Nyf

the

mapping

calibration

following ,

third Y= Z Py (i,

order

polynominals,

j 1 *Nx*xNyt

with 0 I i,j < 3 , 0 I i+j < 3. In these expressions, Px(O,O, and Py(tl,Ol gives the deviation of the signal center from the geometrical one. Sumetimes, Px(l,O) referred to as monitor k-values. and Py(0, Ii are Doininant ter:ns in Eq. (li are those of Px(i=odd. j=even) j=odd), and other terms are small due to and Py(i=even, of the pickup configuration. Fig. 2 shows the symmetry 01 the calibration with a coaxial antenna the setup the field in the monitor chamber. exe ting Owing to antenna positioning and data the computer-controlled of one monitor takes only about tak ng. the calibration Fig 3 shows the contour mapping of the ten minutes. man tar coordinates. li”“l

“l..)

Mapping

calibration source at tX,Y) in a monitor u(e set the sign.1 chalnber and measure the output voltages h,B,C,D, and then get the normal i;a t ions, Nx= (AtC-B-D) /(A+B+C+U) and Ny=(AtB-C-D,,‘(AtHtCtl)). In this way, we get the relation between sign.al coordinates (Nx,Ny) and space coordinates (X,Y) for each monitor. We can practically

(a)

Fig

1

Position

monitor

chamber

in

the

bending

(1)

section

DD18-9499/85/1GDC-1947$01.00~~

race track ch&nber in bending set t ion Fig

3

Contour

1985 IEEE

mapping

(b)

of

circular chamber in collision section position

monitors

Closed

orbit measurement The measurement of the cloce~l orbit in the four c.pntrol proceeds in paralle! and takes about 12 lOCd! display of the orbit on the Fig S is the seconds! We also graphic screen at the control console. prepared the display of the voltage output of each monitor electrode as in Fig 6. We consult this display whenever beam orbit looks abnormal because it is very in the system. useful to find bad elements

L”

,.I,, : : ,, : +$j .* : I is ;. : j L L..-$.++& i’ ‘&.. ;r .-+--6,. : ..i...i.. .c7-~--;~.‘-~.‘-tl’~+-++-~~~ 1 : : I ; : ,. h 0 , ,b ,I ,b 21 I” 3; / Id----: I-. j >_

1

j

BCRrl 5. LllR : I I I :

35-03-15 *5:12:*6 : !“.I. L amy j I ( :

; 5 ;’ ; : is

LL , ,* ,x -n u -~U:‘9I!:,t9~:2il..“~““,‘.1::‘“s:.PI:”:YL~~”-

: ;:

:

:

7: 71 su :: :: -w ; .‘_ ;4 .,, I*; *-e 1.:-~ :._~“,t’i-‘,-;-,,‘..~~~.~ v ....I. L--f*......-,* ,b : ( ..T:.y --+-l----4-. t-i-. : : ~ .-.&&--.+.’ ‘0 3; 1*I 0.8 * V2 and V2 .’ 0.8 * Vmax (2) where Vmin i V2 I V3 5 Vmax and these are outputs of four electrodes. Actually, Eq (2) is satisfied by the beam within the area of 20 mm x 12 mm in the central region of the chamber. In the later experiment, we met the cases where Eq (21 was not effective because we had plural number of abnormally small output. Hence, we adopted one more criterion; Vnax and Vnin should appear in a diagonal couple of electrodes. This check works well if used together with Eq (2). Table

Fig

of

j

X rms/Y rms number of PM

I

Distribution in the

of standard deviation repeated measurements

(a) No data check < 0.05 < 0. I >o. 1 67/63 7/l 1 5/5

(b) Check by Eq (21 < 0. 05 < 0. 1 >o. 1 70/67 8111 l/l

I949

Pil

:jr:_:

6

i

(b)

(a) Fig

7

Position

nonitor

(a)

data

during

beam

position (0. 5 mm range) and beam current (O-40 electrode outputs (O-10

(b)

storage

mA) VI

improvements

for

monitor

The coelficients Px(l,O) and Py(0.l) which were obtained in the calibration measurement did not fully agree with the analytical calculation in the case of a In the measurement, the stripped circular beam duct. end of the coaxial antenna was set at the position of The field pickup electrode (position 1 in Fig 8Aj. distribution in this configuration was obtained by “TWA based on boundary eiemen t method (BEM). program” [6,7! It showed that the position of the antenna was not adequate because the field is disturbed at the antenna the pickups did not see such therefore, end and, (TEM) field as the beam with transverse electromagnetic light velocity produces. Instead, TEM pattern appears at some distance away from the antenna end (position 2). The measurement agrees with BEM calculation when the antenna posit ion is set so that the pickups see the TEM Sensitivity calibration of MR monitors is now field. proceeding with the improved antenna configuration, but the revised measurement is no longer possible for AR However, orbit correction of AR is possible monitors. even with small error in the sensitivity data because the coordinate origin was correctly obtained in the a practical solution Therefore, previous calibration. the coefficients Px(l,O) and will be to replace Py(0, 1) in Eq (11 with the calculated ones. 81

in tens i ty dependence Since the beam current decreased from 40 mA to 0. 1 nA in the above measurement. Fig 7a shows the at various beam current, performance of electronics Fig 7b show:. the output voltages of individual too. electrodes in the same measurements Step change in the output mean\ the switching of the programmable The super-heterodync circuit has linearity attenuator. range of about ~10 dR and the overall linearity range can be further extended by an external programmable the full However, attenuator by 40 dH with IOdH step. linearity range oi the detector ctrcuit is not used beca1us.e we must adapt the output love1 to the full get enough precision in the range of II bit ADC to When the programmable attenuator is set norm21 i7a t ion. of position data is recognized at at 0 dB, offset WC suspect that this is due to the several monitors. mismatching at the circuit input and at the pickup The effect of mismatching will be largest electrodes when four electrodes are unbalanced and there is no it seems to be adequate to keep attenuation. Therefore, the attenuation to some finite level.

HI; position

, SENSOR E~rcrnoqr

2

Beam

Esttmation of the setting of unneasurcd monitors Geometrical setting error has not heen measured for 29 [monitor chambers out of 83 because the measuring robot was not applicable due either to irregular chamber shape or to restriction of space in the tunnel. reliable position data is not available from Hence, these monitors as was demonstrated by the following experiment, orbit correction at the 63 measured first by using the data moni tars was made 111 tlyil IuIy:, of measured monitors only and next by using both The first case gave the neasorad and unmeasured ones. We proposed a better result as is shown tn Table 2. beam experiment In place of the geometrical mea