the atf damping ring bpm upgrade nathan eddy, eliana gianfelice-wendt fermilab for the atf damping...
TRANSCRIPT
The ATF Damping RingBPM Upgrade
Nathan Eddy, Eliana Gianfelice-Wendt
Fermilab
for the ATF Damping Ring BPM Team
New Hardware
Improvements on the analog downconverter
CAN-bus controls, IF filter, remote diagnostics, etc.
New RF, DC & CAN-bus distribution. Grounding of tunnel hardware.
Switch to in-house VME digitizer8-ch. ,125 MSPS ADC (serial outputs), Cylcone III FPGA, PLL-locked CLK distribution
Able to measure Injection TBT, Narrowband Orbit, Narrowband Calibration , and Last Turn on every injection
6/30/2010 ATF2 Project Meeting 2
LO & CAN-bus Distribution
New Timing Module& Custom Digitizers
New Downmix &Calibration
Current Status
• 95 of 96 Ring BPMs were switched to the new system during the May shutdown
• Beam commissioning began the last week of May– Initial Timing tests revealed an issue with clock synchronization
– The echoteks used a 69.2MHz clock (32 samples/turn) synthesized from the 714MHz
• The turn by turn data collection was initiated by an external turn marker
• Fine for 1024 turns
– The new system counts turns internally from injection to provide turn by turn data at any turn and last turn data
• The synthesized 69.2MHz clock was found to drift over a full machine cycle
• This caused problems with the turn by turn data at the end of the cycle
– The solution was to bypass the clock synthesizer on the Timing Module
• Simply use the clock divider, 714MHz/10 = 71.4MHz (33 samples/turn)
• Solved the locking issue but required a major system modifications
• Operation of all bpms was demonstrated over the remaining shifts– Orbit data was read into the ATF control system
– First beam studies
• Two Sets of Narrowband Orbits were collected, without and then with calibration
• Several Turn by Turn data files were collected both at injection and by kicking the beam after 500k turns
6/30/2010 ATF2 Project Meeting 3
First Look at Narrowband Orbit
• Narrowband orbit for each injection (shot) is measured – The mean orbit and rms over 160ms is reported for Horz, Vert
• For each data set, 128 shots are collected– Large shot to shot mean orbit variations are observed in both– Horizontal shows larger RMS than vertical -> Beam related
6/30/2010 ATF2 Project Meeting 4
0 10 20 30 40 50 60 70 80 900
10
20
30
40
50
60
Orbit Variations - No Cal
BPM #
RM
S (
um)
Horz - Mean 14.1um
Vert - Mean 9.7um
0 10 20 30 40 50 60 70 80 900
10
20
30
40
50
60
BPM #
RM
S (
um)
Orbit Variations - with Cal
Horz - Mean 16.6um
Vert - Mean 11.6um
Narrowband Orbits
• Single shot RMS is much smaller than shot to shot orbit variations– Plotted as mean shot RMS with variation on RMS as error bars– Observe larger RMS in Horizontal than Vertical -> Beam related– Data with Calibration has larger RMS than without Calibration
6/30/2010 ATF2 Project Meeting 5
0 10 20 30 40 50 60 70 80 900
1
2
3
4
5
6
7
8
9
10
BPM #
RM
S (
um)
Shot RMS Variations - No Cal
Horz - Mean 1.1um
Vert - Mean 0.3um
0 10 20 30 40 50 60 70 80 900
1
2
3
4
5
6
7
8
9
10
BPM #
RM
S (
um)
Shot RMS Variations - with Cal
Horz - Mean 1.4um
Vert - Mean 0.6um
Preliminary SVD Orbit Analysis
• Use Singular Value Decomposition to look for correlated motion and estimate resolutions
– Ignoring beam effects, the horizontal and vertical resolution of the pickup/electronics should be the same…
6/30/2010 ATF2 Project Meeting 6
Preliminary SVD Orbit Analysis
• See much larger residual RMS in Calibrated Data– Still 6um Horizontal and 4um Vertical after removing first 15 modes!
– Something is not working correctly with Calibration…
6/30/2010 ATF2 Project Meeting 7
Analysis with Echotek System
• Perform same analysis on Data taken with Echotek system 12/07– 48 shots were analyzed with 21 BPMs instrumented– See resolution of 540nm after removing 7 modes– Largest mode is all Horizontal
6/30/2010 ATF2 Project Meeting 8
0 10 20 30 40 50 60 70 80 90 1000
2
4
6
8
10
12
14
RMS vs Mode - Echotek 48 Shots 12/07
Horz
RM
S (
um
)
Remove 0 Modes (8852 nm)
Remove 1 Modes (3967 nm)Remove 2 Modes (1883 nm)
Remove 3 Modes (1517 nm)
Remove 4 Modes (1077 nm)
Remove 5 Modes ( 818 nm)
Remove 6 Modes ( 686 nm)Remove 7 Modes ( 540 nm)
Remove 8 Modes ( 468 nm)
0 10 20 30 40 50 60 70 80 90 1000
1
2
3
4
5
6
BPM Number
Vert
RM
S (
um
)
Remove 0 Modes (2465 nm)
Remove 1 Modes (2422 nm)Remove 2 Modes (1269 nm)
Remove 3 Modes ( 979 nm)
Remove 4 Modes ( 879 nm)
Remove 5 Modes ( 780 nm)
Remove 6 Modes ( 609 nm)Remove 7 Modes ( 541 nm)
Remove 8 Modes ( 496 nm)
Re-Analysis Uncalibrated Data
• Perform the same statistical analysis on new data– Use 48 shots and the same 21 bpms as in the echotek system
– Resolution is 390nm after removing 8 modes
– Larger initial variations and all modes are mixed Horizontal/Vertical
6/30/2010 ATF2 Project Meeting 9
Summary
• The Damping Ring BPM upgrade installation is now complete and commissioning has begun– Major change to digitizer clock during commissioning
• The Narrowband Orbit data and Turn by Turn data are available
• Preliminary Analysis of the Narrowband data– Horizontal RMS > Vertical RMS -> measuring beam not noise!– Clear problem with the Calibrated orbit data
• Should be resolved before operation resumes
– Large shot to shot orbit variations…• Variations are larger than observed with Echotek system in 12/07• Simple estimate of system resolution from shot to shot data is slightly
better with new system
• Results from Turn by Turn Analysis are presented next
6/30/2010 ATF2 Project Meeting 10
The Future
• Continue Commissioning the system – Optics and Machine studies – more analysis needed– Determine resolutions for all measurements
• Update the position scaling for the injection bpms– Use correct bpm geometry for bpms 20 & 21
• Investigate odd behavior of bpm 53– Flagged by SVD and TBT analysis
• Implement a history for the calibration data– Allow tracking of channel performance– Show large drifts which may indicate impending failure
• User requests?
6/30/2010 ATF2 Project Meeting 11
Backup
6/30/2010 ATF2 Project Meeting 12
Overview of the BPM Upgrade
• New read-out hard-, firm- and software, BPM pickups (button-style) stay unchanged.
• R&D activities over the last couple of years on 20 BPMs in the arcs, utilizing mixed analog/digital signal processing– Test of different analog downconverters (w/o CAL)
– Digital signal processing based on spare Echotek digital receivers.
• Final upgrade scenario (96 BPMs, plus spares)– 714-to-15.1 MHz analog downconverter with CAN-bus controlled
calibration tone, located in the tunnel.
– VME hard- & software, in 4 rack locations• 8-ch. 125 MSPS digitizer with an Altera Cyclone III FPGA• 12 ch. VME timing generator (Fermilab).• Motorola 5500 VME controller, with CAN-bus interface,
running VxWorks & EPICS software
– Auxiliary hardware, e.g. power supplies and distribution, LO-signal distribution, CAN-bus distribution, etc.
1/14/2010 LER 2010 - ATF Damping Ring BPMs 13