Secondary Emission Monitor for very high radiation areas of LHC Daniel Kramer for the BLM team LHC Beam Loss Monitoring system ~ 3700 BLMI chambers installed along LHC ~ 280 SEM chambers installed in high radiation areas: – Collimation Injection points – IPs – Beam Dumps – Aperture limits Main SEM requirements – 20 years lifetime (up to 70MGray/year) – Sensitivity ~7E4 lower than BLMI 10.6.2008 D.Kramer BLM Audit 2 Secondary Emission Monitor working principle Secondary electrons Secondary Electron Emission is a surface phenomenon Bias E field Energy of SE (below ~ 50 eV, dominant for signal) is independent on primary energy Ti Signal electrode Ti HV electrodes SE are pulled away by HV bias field (1.5kV) Signal created by e- drifting between the electrodes < 10-4 mbar Steel vessel (mass) Delta electrons do not contribute to signal due to symmetry* VHV necessary to keep ionization inside the detector negligible Very careful insulation and shielding of signal path to eliminate ionization in air (otherwise nonlinear response) No direct contact between Signal and Bias (guard ring) 10.6.2008 D.Kramer BLM Audit 3 Secondary Emission Monitor working principle Secondary electrons Secondary Electron Emission is a surface phenomenon Bias E field Energy of SE (below ~ 50 eV, dominant for signal) is independent on primary energy Ti Signal electrode Ti HV electrodes SE are pulled away by HV bias field (1.5kV) Signal created by e- drifting between the electrodes < 10-4 mbar Steel vessel (mass) Delta electrons do not contribute to signal due to symmetry* VHV necessary to keep ionization inside the detector negligible Very careful insulation and shielding of signal path to eliminate ionization in air (otherwise nonlinear response) No direct contact between Signal and Bias (guard ring) 10.6.2008 D.Kramer BLM Audit 4 Secondary Emission Monitor working principle Secondary electrons Secondary Electron Emission is a surface phenomenon Bias E field Energy of SE (below ~ 50 eV, dominant for signal) is independent on primary energy Ti Signal electrode Ti HV electrodes SE are pulled away by HV bias field (1.5kV) Signal created by e- drifting between the electrodes < 10-4 mbar Steel vessel (mass) Delta electrons do not contribute to signal due to symmetry* VHV necessary to keep ionization inside the detector negligible Very careful insulation and shielding of signal path to eliminate ionization in air (otherwise nonlinear response) No direct contact between Signal and Bias (guard ring) 10.6.2008 D.Kramer BLM Audit 5 Secondary Emission Monitor working principle Secondary electrons Secondary Electron Emission is a surface phenomenon Bias E field Energy of SE (below ~ 50 eV, dominant for signal) is independent on primary energy Ti Signal electrode Ti HV electrodes SE are pulled away by HV bias field (1.5kV) Signal created by e- drifting between the electrodes < 10-4 mbar Steel vessel (mass) Delta electrons do not contribute to signal due to symmetry* VHV necessary to keep ionization inside the detector negligible Very careful insulation and shielding of signal path to eliminate ionization in air (otherwise nonlinear response) No direct contact between Signal and Bias (guard ring) 10.6.2008 D.Kramer BLM Audit 6 Secondary Emission Monitor working principle Secondary electrons Secondary Electron Emission is a surface phenomenon Bias E field Energy of SE (below ~ 50 eV, dominant for signal) is independent on primary energy Ti Signal electrode Ti HV electrodes SE are pulled away by HV bias field (1.5kV) Transit time 500ps Signal created by e- drifting between the electrodes < 10-4 mbar Steel vessel (mass) Delta electrons do not contribute to signal due to symmetry* VHV necessary to keep ionization inside the detector negligible and avoid capture of electrons Very careful insulation and shielding of signal path to eliminate ionization in air (otherwise nonlinear response) No direct contact between Signal and Bias (guard ring) 10.6.2008 D.Kramer BLM Audit 7 SEM production assembly All components chosen according to UHV standards Steel/Ti parts vacuum fired Detector contains 170 cm2 of NEG St707 to keep the vacuum < 10-4 mbar during 20 years Pinch off after vacuum bakeout and NEG activation (p<10-10mbar) Ti electrodes partially activated (slow pumping observed during outgassing tests) NEG St707 composed of Zr, Vn, Fe Zr flamable -> insertion after the bottom is welded 10.6.2008 D.Kramer Very high adsorbtion capacity of H2, CO, N2, O2 Not pumping CH4, Ar, He BLM Audit 8 Vacuum bakeout and activation cycle for SEM and BLMI NEG inside the SEM needs additional activation at 350°C Activation means releasing adsorbed gases on the surface which have to be pumped Pinchoff done during the cool down of the chamber Resulting pressure below measurement threshold (<10-10mbar) 400 1.E+04 Bakeout cycle for IHEP IC 350 Temperature [°C] 36xIC 300 Pressure He leak tests 250 200 1.E+02 350 1.E+00 300 1.E-02 1.E-04 150 1.E-06 Vacuum bakeout 100 50 1.E-08 0 1.E-10 0 5 10.6.2008 10 15 20 25 30 35 40 45 Time [h] D.Kramer Manifold stays colder to limit the load to the pumping system Activation temperature limited by the feedthroughs Bakeout cycle for IHEP SEM 400 Temperature [°C] 18xSEM NEG activation Vacuum bakeout 250 Manifold pinchoff 200 150 100 Ion pump started He leak tests 50 0 0 BLM Audit 5 10 15 20 25 30 35 40 45 Time [h] 9 Geant4 simulations of the SEM Secondary Emission Yield is proportional to electronic dE/dx in the surface layer – LS = (0.23 Ng)-1 g = 1.6 Z1/310-16cm2 “TrueSEY” of each particle crossing the surface boundary calculated and SE recorded with this probability Correction for impact angle included in simulation QGSP_BERT_HP as main physics model Model calibration factor Electronic Penetration energy loss distance of SE 0° impact angle Geant4 SEM Response function Comparison to literature values => CF = 0.8 10.6.2008 D.Kramer BLM Audit 10 SEM Calibration experiment in a mixed radiation field (CERF++ test) Response of the SEM measured with 300GeV/c beam hitting 20cm copper target Setup simulated in Geant4 Response of SEM filled by AIR measured and simulated as well SEM Response expressed in absolute comparison to Air filled SEM – Response = Dose in AIR SEM / output charge of SEM 0.259 +/- 0.016 Gy/count 10.6.2008 D.Kramer H4 Calibration setup with Cu target and a box with 16 SEMs on a movable table BLM Audit 11 Calibration results Only 2 chambers out of 250 had higher offset current Not corrected for systematic position errors SEM calibration in H4, expect from Geant4 3.95+/-0.17 - ratio QSEM/Qbeam [e /pot] 8 6 4 2 0 0 2 4 6 8 10 monitor position 12 14 16 300 350 12 SEM calibration in H4 week 45 1bar(0.6 sigSEM / sigSEM AIR) = 0.26 mbar Pressure inside SEMs smaller than this 10.6.2008 D.Kramer offset current [pA] Upper Limit on the SEM pressure: (equivalent to 3 of the histogram) 2 Offset current without beam 1.5 1 0.5 0 -0.5 0 50 BLM Audit 100 150 200 250 SEM ID number 400 Table of SEM measurements and corresponding simulations Test beam Measured [e-/prim] Geant4 Rel. Dif. [%] PSI 63MeV 0.27 ± 0.014 0.2665 ± 0.0043 1.1 PSB 1.4GeV 0.0495 ± 0.0006 0.0416 ± 0.0046 19 TT20 400GeV 0.476 e-cm 0.608 e-cm 22 H4 target 3.40 ± 0.92 3.95 ± 0.19 14 LHC collimator in LSS5 of SPS 4.03 ± 0.25Gy In progress muons 160GeV 0.059 ± 0.016 0.08 ± 0.008 TIDV dump Long term test - 10.6.2008 D.Kramer BLM Audit 26 13 Thanks 10.6.2008 D.Kramer BLM Audit 14 Backup slides Vacuum stand in IHEP for IC production 36 ICs in parallel baked out and filled by N2 For SEMs only 18 chambers in parallel No N2 injection :o) He leak detection done before and after bakeout (and after NEG activation for SEMs) 10.6.2008 D.Kramer BLM Audit 15 Beam dumped on a Closed Jaw of LHC collimator in LSS5. SEM to BLMI comparison 1.3 1013p+ BLMI A SEM Black line – signal not clipped 5*τ_filter = 350ms 10.6.2008 D.Kramer BLM Audit 16 Cable crosstalks study – important crosstalks caused by long cables in the LSS Xtalks CD3 - B 15 CH6 CH7 CH8 counts 10 5 0 0 0.1 0.2 0.3 0.4 time [ms] 0.5 0.6 0.7 Differential signals of CD3 - B differences [counts] 60 CH1 CH2 CH3 40 Ch 6..8 unconnected Xtalk clearly depends on the derivation Signal peak ratio 5e-2 (26dB) (worst case) Integral ratio (47dB) 4.4e-3 Similar behavior for system A 20 0 -20 0 10.6.2008 0.1 0.2 0.3 0.4 time [ms] 0.5 0.6 D.Kramer 0.7 BLM Audit X-talks limited to 1 CFC card only! 17 Standard BLMI ARC installation HV Power Supply HV ground cut here Small low pass filter in the CFC input stage BLMI CFC BLMI 1.5kV CBH50 280pF BJBHT 1M CBH50 integrator 10M COAX < 30m 470 2k2 4n7 0.47u 8X Up to 8 BLMs connected in parallel 10.6.2008 D.Kramer CFC is always close to the quadrupole BLM Audit 18 BLMI / SEM installation for collimation areas 6 HV capacitors in parallel 8 chambers in 1 NG18 cable (up to 700m) HV capacitor removed CFC BLMI BJBHT 1.5kV CBH50 1M 280pF BJBHT 2u8 CBH50 COAX < 30m 150k B J B A P integrator NG18 < 700m 470 2k2 4n7 8X 150k for current limitation 10.6.2008 D.Kramer 280pF = chamber’s capacity BLM Audit ~25pF = SEM’s capacity SEM has not 150k protection! 19 150kOhm Rp resistor for BLMI i/o current limitation between HV capacitor & IC) Limits the peak current on the chamber input to 1500 / 150k = 10mA Fast loss has only the Chamber charge available 280pF * 1500V = 0.4 uC – Corresponds to ~ 7 mGy total loss – Corresponds to ~ 180 Gy/s (PM limit = 22 Gy/s) Slows down the signal collection DC current limited to 1500V / 1Mohm = 1.5 mA – Corresponds to ~ 26 Gy/s (total in max 8 chambers) 10.6.2008 D.Kramer BLM Audit 20 BLMI and SEM in the dump line IR6 on the MKB 10.6.2008 D.Kramer BLM Audit 21 400 GeV Beam scan in TT20 SPS line Longitudinal impact of proton beam r = 2mm Chamber tilted by ~1° Simulation sensitive to beam angle and divergence Negative signal due to low energy efrom secondary shower in the wall Integral of Simulation = 0.608 e-mm Integral of Scan2 = 0.476 e-mm 10.6.2008 D.Kramer BLM Audit Relative difference 22% 22 Prototype tests with 63MeV cyclotron beam in Paul Scherer Institute 0.4 Prototype F -> close to production version C type F type Geant4 Current measured with electrometer Keithley 6517A 0.35 + Prototype C -> more ceramics PSI proton beam 62.9MeV BLMS prototypes F & C Type HV dependence of SEY inside (no guard ring) HV power supply FUG HLC14 Pattern not yet fully understood – Not reproduced by simulation High SE response if U_bias > 2V Detector response [charges/p ] 0.25 0.2 Geant4.9.0 simulated SEY = 25.50.8% 10.6.2008 0.3 10 D.Kramer -3 BLM Audit 10 -2 10 HV [kV] -1 10 0 23 Measurements in PS Booster Dump line with 1.4 GeV proton bunches Older prototype measured Type C {Type F simulated} Normalized response 0.07 Profiles integrated with digital oscilloscope Data Geant4.9 fit Data 0.06 Detector response [charges/p+] – 1.5kV bias voltage – 80m cable length – 50 termination – Single bunch passage SEY measurement – 4.9 0.2% 0.05 0.04 0.03 0.02 0.01 Geant4.9.0 simulation – 0 0 4.2 0.5% 10.6.2008 D.Kramer BLM Audit 0.5 1 1.5 2 Beam intensity [p+/bunch] 2.5 12 x 10 24 Example loss induced by the fast moving SPS scraper. Measured in the collimation area by the LHC BLM system 4 different monitors (2006-old electronics) 10.6.2008 D.Kramer BLM Audit 25 Example of beam losses in the SPS collimation area during a collimator movement of 10um (2006) Coasting beam FFT spectrum 2006 data CWG 19/3/07 10.6.2008 D.Kramer BLM Audit 26 SPS Coasting beam 270GeV 200um Left jaw move and FFT spectra 10.6.2008 D.Kramer BLM Audit 27 Complete FFT from the previous plot Horizontal Tune calculation from the BLM measurement -> oscillations in the beam not in the BLM system 10.6.2008 D.Kramer BLM Audit 28