1?
'9s$
SLAC-PUB—5556 *' t
DE91 015482
Wire Scanners for Beam Size and Emittance Measurements at the SLC*
M.C.
Ross, J.T. Seeman, E. Bong, L. Hendrickson, D. McComiick, L. Sanchez-Chopitea
Stanford Linear Accelerator Center,
Stanford University, Stanford, Ca. 94309
ABSTRACT
The SLC wire scanner beam profile monitors provide
accurate beam size
and
emittance measurements
for each
bunch
in the three bunch SLC beam. The beam size measurement
error for typical 50GcV SLC linac beams (lOOum
o-(x.y))
is
better than 5p.m. Beam profile measurements can
be
performed
throughout much of the SLC with no interruption to normal
machine operation
and
no adverse impact on interaction region
detector backgrounds. The linac input and output omittance is
determined using sets of four scanners spaced by -45' betatron
phase advance. Each scanner contains three wires, x, y and u
(45*),
from which an estimate of the x - y coupling can be
obtained. Advanced high level control software allows the use
of wire scanner data in feedback and beam optimization
procedures. Non-invasive scans are performed almost
continually and the results arc logged so that long term trends
in emiliancc can be examined. In this paper we describe the
design, construction, performance and uses of SLC wire
scanners.
INTRODUCTION
Measurements of the beam size and associated optical
parameters are key to SLC. In the past progress has been
slowed by our inability to measure beam size accurately in a
rapid, non-invasive manner.
Tungsten
Wire
Figure
1:
SLC wire scanner wire support
card
viewed
in
the
beam direction. The translation stage and vacuum chamber are
not shown.
work supported by Department of Energy contract
DE-AC03-76SF00515
In this paper we describe the design, construction,
commissioning and ultimate uses of wire scanners in the SLC
focusing on the linac and upstream systems scanners. Of
particular interest is the interaction between the wire and the
scattered radiation from the wire with the extreme electric Field
of the beam. As this field reaches the level of several
volts/angstrom, as it does easily at the SLC interaction point
(and may in upstream parts of SLC), field emission from the
wire may occur.
A key feature of
SLC
operation is the degree of high level
active control required to keep it optimized. The high level of
demand takes the instrument out of the category of a device
primarily used for machine development or failure diagnosis
purposes
and
elevates it to an online device.
Feedback requires a fast, non-invasive (or minimal impact)
device which in turn means that the wire, not the beam, must
be moved during the scan. The speed, range, vibration and
other mechanical specifications can be generated from this
requirement
and
from the expected beam sizes and rates. Tabic
1 shows
the
expected performance of
the
scanners.
MECHANICAL
The mechanical design effort addressed the following
problems: 1) wire and wire retention, 2) vibration over the
large speed range and 3) positioning errors and position
transducers.
A
particular concern
was radiation
damage.
A schematic diagram of the scanner is shown in figure 1.
Several labs hnve built scanners of a similar design
1
>
2
>'.
The
wire is strung around 1.5mm stainless steel studs set in a
3/16in thick alumina fork in such a way so that it can carry
wires of three different orientations across the beam and
provide x, y and u (45') scans. The carriage motion is actuated
by a stepping motor through a 2mm pitch ball screw, chosen
because of the expected large number of cycles. Some
difficulty was experienced obtaining the small pitch, high
quality ball screw with no plastic parts. A 125um thick
stainless steel vacuum window opposite the wire allows low
energy wide angle scattered radiation to emerge from the
vacuum chamber.
Both the cantilever nature of the wire support and the
stepping motor contribute to wire vibration. We have used a
piezoelectric accelcrometcr to quantify the motor related
system vibration.
The wire chosen for the scanner was gold plated tungsten
with a diameter of 0.3 Obcam- ^
nc w
'
rc nas an
effective
'a'
of
radius/2 which, when added in quadrature to the beam size,
causes a
3%
apparent increase when 0"bcam =
w
irc diameter.
Under normal conditions the wire size can be subtracted in
quadrature from the measured size.
Presented at the IEEE Particle Accelerator Conference, San Francisco
_ CiC '.'i"i,0'-.7IGjN O"
Tj|
«. 'VlPi "s«cv
;.JASP
#>
At Ml SLC currents and rates, the beam can heat the wire
substantially reaching a steady state temperature of lOOO'C. In
the fast scan mode, the wire is subjected to continuous beam
for no more than a few seconds. However, the wire may be
parked in the beam in error or
Tor
diagnostic purposes and musl
be able to withstand continuous beam. By measuring the
increase in resistance of the wire assembly an estimate of die
wire temperature rise can be made. Resistance tests show good
agreement with
a
calculated rise of
about
4' C/pulse.
Table 1: Wire scanner performance specifications.
Beam size resolution
<3%
a (= 3 -
10
urn)
Systematic error
<3%
a
(=
3
- 10
um)
Emitlancc
(E)
error
10% (for
IE=3
x 10"
5
m-rad)
Dynamic range
10
9
- 10
n
particles/pulse
Vibration Peak amplitude < 0.2a
Relative positioning
20um
Speed
and
acceleration
lem/s max
0.3mm/s min; 0.2m/s^ acccl.
Multibunch operation
<5%
signal contamination from
nearby bunch (60ns)
Radiation resistant
lOMrad/ycar
Lifetime 100,000 cycles/year
As the wire is scanned through the beam its position is not
encoded on each successive beam pulse, rather the current
position of me wire is inferred from a check of me remaining
step count. A position measurement using a radiation hard
LVDT (linear differential transformer) is done at die limits of
ihc scan to check mat the expected position was reached.
CONTROLS
A scan consists of
three
steps: 1) move from PARK (near,
but not in, any of the beams) to the start of scan at maximum
speed, 2) scan at the speed corresponding to the desired inter-
point spacing and beam rate and 3) return to PARK at full
speed. Only brief pauses, to allow the device to come to a
complete stop, occur during the scan sequence. During
muliiwire emillancc or skew scans, the wire moves to die next
appropriate PARK. An important feature of this scheme is die
use of machine wide data acquisition codes which coordinate
die readback of
the
scanner step count and the signals from the
wire scanner detector and therefore allow a great deal of
flexibility in the choice of detector including,
Tor
example, the
use of detectors several miles away.
Because die scanners are to be used for feedback, the
application software that controls them must have
sophisticated exception handling, error logging and status
reporting. The control system software built around these
devices allows use of the wire scanner at several levels. The
lowest is the single scan and associated single pulse detector
signal readout. This information allows checking the Tit
quality, scan ranges and other details. At the next level higher,
the fit results can be used in the SLC control system
correlation plot ulilily.'' This extremely powerful tool.
allowing the acquisition of scan data with other beam
diagnostic data and machine parameters, has been invaluable
for commissioning the scanners. All aspects of
the
gaussian fit
to the scan data
arc
available and arc automatically acquired as
an upstream device setting is controlled in
a
programmed way.
Most automated optimization procedures arc built around this
facility. The next level of software docs multiple scans and
accumulates these results in the correlation plot utility. This
includes four wire emittancc scan results and skew scan results.
Finally, feedback can perform the scans as a background task
and implement the desired corrections.
SIGNAL DETECTOR
The purpose of the wire signal detector is to indicate the
amount of charge striking the wire. Secondary emission and
forward scattering, used in many wire scanner systems, are
often not practical at the SLC, die first because of problems
discussed below and the second because or the very different
beam line areas in which these devices have
ID
operate. The
most difficult region is just upstream of a high power
collimator system. Radiation scattered by the wire in the
forward direction is completely overwhelmed by the scattering
from the collimator jaws thus making the use of small angle
scattered radiation impossible.
When me SLC interaction point (IP) wire scanners
5
were
first tested it was found that the secondary emission signal
would increase dramatically when either the beam intensity
exceeded about 5 x 10' or the Cbeam
<
10p.m. The onset of
ihis dramatic increase is indicated by a very unstable signal. In
die SLC linac, where the beam sizes are about 100pm, a beam
intensity of about 2.5 x lO
1
^ is required. This effect appears to
be
field
emission induced by die field of
the
beam which peaks
at about 20V/A.
Because of
these
problems, tests were made to determine if
a significant signal was present at 90' to the beam direction,
directly opposite the wire support card. A strong, very low
energy election signal was seen in a bare phoiomuliiplicr
(PMT) placed about 30cm from the wire. A ihin window is
required so mat this scattered radiation is not absorbed by the
vacuum chamber wall. Substantial shielding (±50 radiation
lengths) is required in some locations to protect the PMTs
from background generated by upstream beam losses. In
regions where no collimators follow the scanners, small angle
scattering monitors have been placed about 10m downstream
to use for comparison with the PMT.
The detector linearity musl be better than a few percent. A
moderate gain, excellent linearity lube was chosen. In order to
cross check PMT performance small fast ion chambers were
built
and
installed downstream of some scanners
6
The linearity
of die ADC is also very important.
PERFORMANCE AND CONCLUSIONS
Performance tcsls have focused
on
understanding systematic
errors. These tests Tall into two broad categories: 1) tests made
by varying beam size and/or intensity and 2)
ICSL>
made with
different detectors sensing scattered radiation from a single
-1000 0
Wire position (|im)
Figure
2:
A typical single wire scan. The data arc fit with a
gaussian with an offset.
wire.
Figure 3 shows the beam emittance measured using
quadrupole scans
7
on three different scanners and multiwirc
scans at
1
GeV and 4,5 x 10
1
" e/'puise, near the SLC nominal
operating intensity. This is a good test of saturation effects
since the beam sizes and signal strengths vary considerably
over the scan range and from scanner to scanner. These tests
were done using the downstream fast ion chamber.
The SLC wire scanners provide beam cmittance dr.ta that
is
reliable
enough to have already yielded new insights into the
performance of the SLC. By late 1991, 32 scanners will be in
use (figure 4). Future linear colliders will have a tighter
emittance budget and will require improved resolution
a b c
Emittancc measurement method
Figure 3: Comparison of emiltance measurements made
with: multi wire (a) and quadmpolc scans
(b-d)
using different
quadrupole magnet / wire scanner combinations. The results
are in good agreement.
scanners. Piezo-elcctric motors, with their very small step size
and ultra-high vacuum compatibility, may prove to be an
appropriate technology for future wire scanners. Two such
scanners are installed for use at die SLC IP.
ACKNOWLEDGEMENTS
We would like to acknowledge the efforts of the linac
group for help in commissioning the scanners, the mechanical
design and engineering group for mechanical design, A.
Tilghman for controls, and C. Field, K. Bouldin and C. Young
for PMT testing and data analysis.
Special
Purpose
Figure
4:
SLC schematic showing the locations of wire scanners. Wire scanners are used for cmittance (e), energy spread (AE/E)
and special purpose functions in the final focus.
1
R. Jung and RJ. Colchester, 'Development of Beam Profile
and Fast Position Monitors for the LEP Injector Linacs', IEEE
Trans.
Nucl. Sci. NS32-5:1917,1985.
2
R.I. Cutler et. al., "Performance of Wire Scanner Beam
Profile Monitors to Determine the Emittance and Position of
High Power CW Electron Beams of the NBS-Los Alamos
Racetrack Microtron', Proceedings of the 1987 IEEE Particle
Accelerator Conference,
p.
625,1987.
3
K.D. Jacobs el.al., 'The Beam Profile Measurement System
at the Bates Linac', Proceedings of the 1989 IEEE Particle
Accelerator Conference,
p.
1523,1989.
4
L. Sanchez-Chopitea et. al., 'Correlation Plot Facility in the
SLC Control System', proceedings of this conference.
5
R. Fulton et. al., ' A High Resolution Wire Scanner for
Micron Size Profile Measurements at the SIX', Nucl. Instr.
Meth. A274:37,1989.
6
D.
McCormick, "Fast Ion Chambers for SLC, proceedings of
this conference.
7
M.C.
Ross ct. al.,'Automated Emiltance Measurements in
the SLC, Proceedings of the 1987 IEEE Particle Accelerator
Conference, p. 725,1987.