LHC- Longitudinal Profile Monitor. Non intercepting diagnostics based on synchrotron light from a bending magnet (started as “piggy back” on transverse profile system) • • • Longitudinal Beam profile (3564 bunches) Abort Gap Population Bunched/Debunched beam at injection Empty RF Buckets (aka “ghost bunches”) Longitudinal wings (high resolution) Core Measurements: length, distribution, oscil. . Functional Specification: CERN/EDMS Doc. 328145 SCH: LEADE LPM+AG 15/12/03 1
Point 4: Echenevex RF accelerating cavities SCH: LEADE LPM+AG 15/12/03 2
Specification Requirements Ultra-high sensitivity 60 High sensitivity Standard sensitivity 104 106 3 x 10 -7 5 x 10 -5 5 x 10 -3 60 to 6 x 104 to 3 x 108 5 x 106 to 5 x 108 100 ns 50 ps 100 ms 10 s 1 ms 30 p/ps 4 x 103 p/ps 1% <1 s 1 min 100 ms Abort gap monitor X Tails X Ghost bunches X De-bunched beam X Core parameters X MODE Sensitivity (p/ps) Sensitivity/Ult. Peak Density Dynamic range (p/ps) Sampling period Integration time Accuracy Transmission rate APPLICATIONS SCH: LEADE LPM+AG 15/12/03 3
Longitudinal Profile Monitor: situation Sept. 2003 LPM is a LARP contribution/“collaboration”, not in-house project LPM is studied by Berkeley lab, (need similar instrument in ALS) (n. b. same group responsible for LHC Luminosity measurement) Initial plan: 2002 R&D (APD/ Laser mixing) 2003 Choice of technology + Prototype design 2004 Production electronics and instruments 2005 Install in LHC (2006 reserved for transverse instruments) SCH: LEADE LPM+AG 15/12/03 4
Light production at injection energy too low: SC undulator added Undulator and D 3 magnet at LHC Point 4 SCH: LEADE LPM+AG 15/12/03 5
Abort gap, 3 s (Protons) For Ions: spacing 100 ns, total 890 bunches, 1: 40 RF cycles SCH: LEADE LPM+AG 15/12/03 6
LHC Beam profile 3564 bunches+ 32, 076 empty buckets, RMS Bunch length 0. 28 ns: sample time 50 ps 89 s/50 ps = 1, 780, 000 data bins…. Integrate over 40 ms… … 50, 000 data points/second SCH: LEADE LPM+AG 15/12/03 7
Fast Photon Detectors: Commercial Avalanche Photodiode modules PC card for Time-correlated Single Photon Counting: Time. Harp 200 3 MHz count rate, <40 ps resolution …but only 4096 time-bins… SCH: LEADE LPM+AG 15/12/03 8
Fast Photon Detectors: Avalanche Photodiodes C-SPAD: Cooled Single-Photon Avalanche Diode With active quenching circuit: laser range-finding of satellite in flight, T. Otsubo, CRL, Tokyo 100 ps, 16 bins SCH: LEADE LPM+AG 15/12/03 9
Photon Counting: MCP PMT Hammamatsu R 3809 U Photon counting has its problems too, For high dynamic range, there must be no systematic false counts SCH: LEADE LPM+AG 15/12/03 10
MCP-PMT: Dynamic range 102 Self-generated false counts after event: need to gate detector off for 1 us after each event: count rate drops to 10’s of k. Hz. 900 ns SCH: LEADE LPM+AG 15/12/03 11
Proposed Laser Mixing System: filters Synch light, 633 nm 40 MHz Laser Mixing crystal 8 bit Detector 350 nm Ti: Sapphire 800 nm Very narrow light spectrum is used, ~3 nm (1% of available) Laser pulse timing phase-locked to Machine RF, with offset Laser used to sample with 10 ps pulses at 40 MHz Max. Data rate at laser frequency, could be 80 MHz? SCH: LEADE LPM+AG 15/12/03 12
Laser Mixing: LPM High Sensitivity Mode 25 ns laser sample interval, 500 sample points/scan 400 MHz Bucket 2. 5 ns Increment delay by 50 ps / machine rev If PMT is 1% accurate, still need to integrate over 1000 samples to get spec accuracy: 55 s + settling time ( Spec. 10 -4 in 10 s ) SCH: LEADE LPM+AG 15/12/03 13
SET 3 Compress Scale. . . SCH: LEADE LPM+AG 15/12/03 14
LM Concerns: • Low wavelength conversion efficiency - single photon counting - needs longer integration time • Requires exclusive operating modes -std/high resolution modes • High res. integration time too long: - 10 sec: increase to 1 min SCH: LEADE LPM+AG 15/12/03 15
• LM system is unlikely to meet specs for integration time, and is not suitable for abort gap protection (too complex) • If LM system is used, 3 separate instruments will be needed. • Avalanche Photon counting has its problems too: as the count rate reduces, the number of detectors becomes large. • No work is being done on the APD method. SCH: LEADE LPM+AG 15/12/03 16
Photon Production and the Abort Gap Monitor Calculated photon production (450 -900 nm) : 0. 0014 photons/proton at the extraction mirror (…& ions? ) At injection energy, abort level is now x 700: 60 p/ps x 700 = 5. 88^6 photons/100 ns/turn = 6 10^9 photons / 100 ms integration The 7 Te. V abort level remains at 60 p/ps: 840 000 photons/integrated over 100 ms =0. 3 n. J signal (n. b. this is corrected from the value given in talk, the D 3 bending magnet is more efficient at 7 Te. V) From this must be taken transmission losses, detector efficiency(10%), bandwidth(1%), background noise… SCH: LEADE LPM+AG 15/12/03 17
Abort Gap monitor is important machine operating instrument; needs simple robust solution. A separate instrument is considered. 2003: LARP priority for LPM reduced, no funding given (no work done) LBNL team concentrates on Luminosity. 2004: LDM priority raised: ¾ FTE available for LPM+AG (no material) Priorities now set: 1/ Abort Gap monitor 2/ Luminosity 3/ LPM (R&D tool? ) To increase the reliability/availability and performance of the AG and LPM, separate, warm undulators are considered, cost and initial design for March ‘ 04 SCH: LEADE LPM+AG 15/12/03 18
AG Tests at the ALS 328 RF buckets 276+1 filled (LHC parameters) Bunch width ~50 ps (280 -620 ps) (2808/35640) Bunch spacing 2 ns (2. 5 ns) ~120 ns gap (3. 3 µs) “Camshaft” pulse SCH: LEADE LPM+AG 15/12/03 19
Hamamatsu R 5916 U-50 Photomultiplier Tube Gate min. raise time: 1 ns <2. 5 ns RF bucket spacing Gate voltage: 10 V Low voltage switching required Gain at – 3. 4 k. V: 106 High gain < 10 dark counts/sec Low noise Max duty cycle: 1% 100 ns -> 100 k. Hz max sampling rate -> 3 ms to measure entire abort gap (w/o integration) SCH: LEADE LPM+AG 15/12/03 20
MCP-PMT experimental setup (present) SROC Hamamatsu Streak Trigger Unit 1. 5 MHz Stanford DG 535 Delay ~100 k. Hz HP 8114 A Pulser 10 V Gate Tektronix TDS 754 D MCP PMT Visible Light -3 k. V Hamamatsu C 3360 HV (Pockels cell) SCH: LEADE LPM+AG 15/12/03 21
Empty buckets (gap) Parasitic bunch Regular bunches Camshaft SCH: LEADE LPM+AG 15/12/03 22
Parasitic bunch Gate signal on Gate signal delayed 28 ns Gate signal on Parasitic bunches SCH: LEADE LPM+AG 15/12/03 23
The Situation 12/03: Tests to establish Sensitivity and Dynamic range of MCP-PMT …answers for Chamonix? Studies to establish reliable AG design: accessibility, few interventions AG data needed for warm undulator design. More effort should be available in 2005/6 for LPM system: technology choice still open but time very short. SCH: LEADE LPM+AG 15/12/03 24
Скачать презентацию LHC- Longitudinal Profile Monitor Non intercepting diagnostics based
2c1d30614e11a9c2e1553642b1257e0b.ppt