RHIC BPM System Modifications and Performance
ABSTRACT The RHIC beam position monitor (BPM) system provides independent average orbit and turn-by-turn (TBT) position measurements. In each ring, there are 162 measurement locations per plane (horizontal and vertical) for a total of 648 BPM planes in the RHIC machine. During 2003 and 2004 shutdowns, BPM processing electronics were moved from the RHIC tunnel to controls alcoves to reduce radiation impact, and the analog signal paths of several dozen modules were modified to eliminate gainswitching relays and improve signal stability. This paper presents results of improved system performance, including stability for interaction region beam-based alignment efforts. We also summarize performance of recently-added DSP profile scan capability, and improved million-turn TBT acquisition channels for 10 Hz triplet vibration, nonlinear dynamics, and echo studies.
- , - - I
, , , - - - v
R H K BPM System Modifications and Performance
T. Satogata, R. Calaga, P. Cameron, P. Cerniglia, J. Cupolo, A.
Curcio, W.C. Dawson, C. Degen, J. Gullotta, J. Mead, R.
Michnoff, T. Russo, R Sikora
Presented at the Particle Accelerator Conference (PAC’OS)
Brookhaven National Laboratory
P.O. Box 5000
Upton, NY 11973-5000
www. bnl .g ov
Brookhaven Science Associates, LLC
for the United States Department of Energy under
Contract No. DE-AC02-98CHlO886
This is a preprint of a paper intended for publication in a journal or proceedings. Since changes may be
made before publication, this preprint is made available with the understanding that it will not be cited
or reproduced without the permission of the author.
This report was prepared as an account of work sponsored by an agency of
the United States Government. Neither the United States Government nor
any agency thereof, nor any of their employees, nor any of their contractors,
subcontractors, or their employees, makes any warranty, express or implied,
or assumes any legal liability or responsibility for the accuracy,
completeness, or any third party’s use or the results of such use of any
information, apparatus, product, or process disclosed, or represents that its
use would not infringe privately owned rights. Reference herein to any
specific commercial product, process, or service by trade name, trademark,
manufacturer, or otherwise, does not necessarily constitute or imply its
endorsement, recommendation, or favoring by the United States Government
or any agency thereof or its contractors or subcontractors. The views and
opinions of authors expressed herein do not necessarily state or reflect those
of the United States Government or any agency thereof.
FOR UNCLASSIFIED, UNLIMITED ST1 PRODUCTS
Available electronically at:
Available for a processing fee to U.S. Department of Energy and its
contractors, in paper from:
U.S. Department of Energy
Office of Scientific and Technical Information
P.O. Box 62
Oak Ridge, TN 37831
Phone: (865) 576-8401
Facsimile: (865) 576-5728
National Technical Information Service @TIS):
Available for sale to the public from:
U.S. Department of Commerce
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22131
Phone: (800) 553-6847
Facsimile: (703) 605-6900
Online ordering: http://www.ntis.gov/ordering.htm
@ Friritnd an rgc*p~nspapnt
RHIC BPM System Modifications and Performance
T. Satogata *, R. Calaga, P. Cameron, P. Cerniglia, J. Cupolo, A. Curcio,
W.C. Dawson, C. Degen, J. Gullotta, J, Mead, R. Michnoff, T. Russo, and R. Sikora
Brookhaven National Laboratory, Upton, NY 1 1973-5000, USA
The RHIC beam position monitor (BPM) system pro-
vides independent average orbit and turn-by-turn (TBT)
position measurements. In each ring, there are 162 mea-
surement locations per plane (horizontal and vertical) for
a total of 648 BPM planes in the RHIC machine. Dur-
ing 2003 and 2004 shutdowns, BPM processing electron-
ics were moved from the RHIC tunnel to controls alcoves
to reduce radiation impact, and the analog signal paths of
several dozen modules were modified to eliminate gain-
switching relays and improve signal stability. This pa-
per presents results of improved system performance, in-
cluding stability for interaction region beam-based align-
ment efforts. We also summarize performance of recently-
added DSP profile scan capability, and improved million-
turn TBT acquisition channels for 10 Hz triplet vibration,
nonlinear dynamics, and echo studies.
The RHIC BPM system [l, 2, 31 consists of 160 23-
cm cryogenic striplines per plane per ring. 72 dual-plane
BPMs are distributed through the interaction regions (IRs),
and 176 single-plane BPMs are located at each arc PmU.
Signals are cabled through 6 dB reflection attenuators and
20 MHZ lowpass filters to analog/digital integrated front
ends (IFEs). Each IFE contains independent electronics
boards for two measurement planes, including active 20
and 40 dB gain stages, 16-bit digitizers for 1 pm resolution
over a f32 mm measurement range, and Motorola 56301
fixed-point digital signal processors (DSPs) for data reduc-
tion and acquisition control. Arc IFEs are in tunnel alcoves
for radiation protection, while I R IFEs are located in equip-
Each IFE calculates digitizer status, digitized raw sig-
nals, and beam position once per turn for the first bunch
after the abort gap. Upon receipt of a beam-synchronous
trigger, up to 1024 turns are streamed to a local DSP buffer,
then passed along IEEE1394 (firewire) to VME memory
and the RHIC control system. Upon receipt of a separate
trigger, TBT positions are averaged over 10,000 turns to
provide an average orbit measurement. Each IFE can ac-
quire simultaneous 1024-turn TBT and average orbit infor-
mation. Details of RHIC BPM electrodes and acquisition
electronics are available elsewhere [I, 21.
There are 20 dedicated BPM VME crates and controls
front-end computers (FECs) distributed around the RHIC
*Work performed under Contract Number DE-AC02-98CH10886
with the auspicies of the US Deparhnent of Energy; author email
rings. Several console-level servers collect and correlate
TBT and average orbit data from all BPMs, and handle both
orbit logging and orbit delivery to application programs.
RADIATION AND RELIABILITY
During previous RHIC runs, unexpectedly high numbers
of BPM IFE digital boards failed. These were initially re-
motely resettable, but progressively deteriorated to nonre-
settable failures over 1- 2 week timescales. The dominant
radiation-driven failure mode for BPM IFEs in the RHIC
tunnel was radiation damage of the Motorola 56301 DSP,
particularly failure of the core memory. The process of
spare DSP procurement and surface-mount replacement is
expensive and both manpower- and spare-intensive.
During an earlier shutdown, 11 8 BPM IFEs (three al-
coves) were moved from their locations above the cryostats
to equipment alcoves. These showed a 95% reduction in
DSP hang rate, and had no radiation-induced DSP failures
through the following runs . During the FY04 sum-
mer shutdown, 284 BPM IFEs (with 568 cables in 9 al-
coves) were also moved from locations above the RHIC arc
cryostats to equipment alcoves in the RHIC tunnel. To date
there have been no further DSP radiation-induced failures
during the RHIC FY05 run.
To track detailed BPM reliability and injection perfor-
mance, 1024 turns of injection data at all BPMs are saved
for the first bunch of every physics fill. Fig. 1 shows his-
tograms of peak to peak values for several hundred RHIC
Peak-peak signal value [mm]
Figure 1 : Histogram of RHIC BPM injection peak-to-peak
signal values during the FY05 polarized proton run. Aver-
age injection quality has improved by a factor of two over
previous runs .
fills. Good injection consistently gives 0.6 mm peak-to-
peak values, consistent with AGS extraction jitter; these
oscillations are removed with injection dampers. About 2-
3% of BPM injection readings are faulty, indicated by very
large or near-zero peak-peak signal values. This failure rate
is reduced by a factor of five over previous runs .
GAIN RELAY MODIFICATIONS
The IFE electronics have a total of 15 gold-plated high-
frequency DPDT analog relays . Seven are in the sig-
nal path: a calibration signal relay, and three per signal
path for switchable 0 dB, 20 dB, and 40 dB active gain
stages. Six additional relays are in the calibration sig-
nal path for switchable attenuations. While attempting to
diagnose radiation-induced hardware failures and system
performance problems, diagnosis reproducibility problems
were traced to contact instabilities in these relays due to
organic membrane buildup and lack of routine contact cy-
cling during RHIC operations. These relays contain gold-
plated contacts but are not hermitically sealed.
For the FY05 run, 56 planes of BPM electronics in two
service buildings were modified to remove these relays, and
matched 50 ohm coaxial jumpers were installed in their
place. A 3 dB fixed gain stage was included to maintain
reasonable dynamic range for Cu-Cu and polarized pro-
ton operations. These modifications were made for inter-
action region BPMs in the low-beta areas around the large
RHIC experiments STAR and PHENIX to assess modifi-
cation stability. Initial beam-based alignment studies on
both modified and unmodified BPMs show that offsets of
BPMs without relays are stable to 100-200 pm, the accu-
racy of the measurement, while similar BPMs with relays
produced unreproducible errors of over 1 mm . Plans
are underway to modify additional modules during the up-
coming RHIC shutdown.
DSP TIMING SCANS
The RHIC BPM system contains over 1200 signal paths
(two per BPM plane) to verify, and over 600 BPM chan-
nels must also be timed properly. This task is particularly
onerous when moving from self-trigger to dead-reckoned
timing, as each of these timings must be set within 1 ns of
the beam peak to produce a meaningful signal. A LabView
VI was previously used to for timing scans, but one scan
for 12 BPMs would take up to five minutes, and tests of
one RHIC ring would take over one shift.
In April 2005, the DSP code resident in BPM IFEs was
modified to permit acquisition of longitudinal beam pro-
files by locally scanning timing registers. Two parameters
specify start delay and scan width inns, and on-board pro-
grammable delay generators are used to generate timings.
Fig. 2 shows a scan profile for two BPMs for the new
code, showing profiles of the first two bunches in RHIC
separated by a nominal bunch spacing of 105 ns. Each
point represents an average digitizer value at that timing
Profile timing delay [ns]
Figure 2: Profile timing scans for BPMs bo6-bh18 and bo6-
bh20, with 105 ns bunch spacing. The vertical axis is digi-
tizer voltage from each BPM stripline.
over 10,000 consecutive turns. These profiles are acquired
simultaneously for all RHIC BPMs in the course of one
minute. Analysis code is being developed to automatically
set all BPM timings from this data in both dead-reckoned
and self-triggered acquisition modes.
MILLION TURN PERFORMANCE
Ten BPM modules at RHIC have had 512 Mb SDRAM
PCI mezzanine carrier (PMC) memory cards installed, in-
terfacing Motorola 56301 DSP built-in PCI interface. Four
of these in the RHIC blue ring were modified in late April
2003 , while the remaining six (four in yellow, two in 2
o'clock blue DX horizontal BPMs) were modified through
2005. Million-tum acquisition has the same TBT noise lev-
els (about 50-75 pm) as normal 1024-turnTBT acquisition.
The million-turn BPMs have been used to observe and
study well-established beam oscillations in RHIC at fre-
quencies of about 10-14 Hz. The power spectrum for
a million-turn acquisition at store energies is shown in
Fig. 3, showing clear peaks at 9.8 Hz and 10.3&, and a
substantial 60 Hz component that appears in other beam
measurements. 256 ktum measurements have be acquired
about every 30 seconds to correlate to possible cryogenic-
driven triplet vibration sources [ 8 ] . Million-turn BPMs
have also beenused during electron cooling solenoid beam-
based alignment experiments to attempt to improve signal
to noise; here a 1 Hz modulation is applied to an alignment
quadrupole, and the beam response at 1 Hz is measured .
The RHIC BPM system has demonstrated substantially
improved performance and stability in 2005 compared to
previous RHIC runs. All BPM electronics have been
moved from the RHIC tunnel to limit radiation upsets and
. . . . ..............
Figure 3: Million-turn acquisition and spectrum for the RHIC blue ring horizontal BPM bo2-bhl0 at storage energy,
showing 100 pm IR triplet vibration components and 60 Hz oscillations .
__ _ -
7 - -
.. . -
Figure 4: Million-turn acquisition and spectrum during blue ring electron cooling studies at injection where a nearby
quadrupole is being driven at 1 Hz . 22 Hz power in the spectrum comes from coherent synchrotron motion near
radiation-induced DSP failures. A test sample of BPM
electronics had their gain relays removed; beam-based
alignment and operational experience demonstrate an order
of magnitude improvement in readout stability. Fast DSP
code for signal path validation and global timing has been
implemented and tested, allowing full routine timing and
signal tests for al1,BPMs. Ten BPM modules have million-
t u r n capability, and are being routinely used for 1 1 Hz mod-
ulation feedback calibration, electron cooling beam-based
alignment, and transition instability studies.
[l] T.J. Shea and R.L. Witkover, “RHIC Instrumentation”, Beam
Instrumentation Workshop 1998 (Stanford).
 T. Satogata et a Z . , “RHIC Instrumentation”, Nucl Instr Meth
in Phys Res A 499 (2003) pp. 372-387.
 T. Satogata et aL, “RHIC BPM System Performance, Up-
grades, and Tools”, EPAC2002 (Paris, France).
 R. Calaga and R. TomLs, “RHIC BPM Performance”,
EPAC2004 (Laceme, Switzerland), p. 2867.
 Aromat Corporation internal
high frequency relay
 J. Niedziela et al., “Quadrupole Beam-Based Alignment at
RHIC”, these proceedings.
 T. Satogata et aZ., “Multi-Million-Turn Beam Position Moni-
tors for RHIC”, PAC2003 (Vancouver), p. 2697.
 C. Montag et al., “Measurements of Mechanical Triplet Vi-
brations in RHIC”, EPAC2002 (Park, France).
 P. Cameron et al., “Beam-Based Alignment in the RHIC
eCooling Solenoids”, these proceedings.
technical specifi cations,