1. HF PMT Issues HF PMT Choice: Hamamatsu R7525 Main Properties Operating Conditions Gain Choice Criteria – lifetime – saturation – timing – SPE monitoring – sourcing – Raddam HV Operating Range Short-term strategy INFN Aldo Penzo HCAL Topical Meeting, 05 March 2009 1

2. 8 stages Rise time: 1.3 ns Trans time : 14 ns 2

3. 3

4. Almost all parameters checked at Iowa 4

5. Distribution of PMT vs  PMT boards (powered up independetly) Mapping to detector These towers are in the shadow of HE Radiation doses varies by 4 orders of magnitude as a function of  1 HV supply power the PMT boards of 9 wedges: 36 EM towers (and corresponding HAD) will have the same HV. 5

6. Physics Priorities Most important towers for physics are EM towers ( HAD towers improve 1 TeV shower resolution by 14% at most) and low  towers (Pt, physics, backgrounds). In June 2003 CMS week it was decided to assign low gain tubes to HAD and within these samples lowest gains to high  QIE non inverting input saturation at 26772 fC From 2003 testbeam assume 30 pe/100 GeV electron or 20 pe/100 GeV pion. 6

7. HF Gain Distribution CMS IN 2005/046 7

8. Lifetime limit The average energy deposited in a em tower at the highest  is 34 GeV/beam crossing (F.i. S.Abdulline, CMS IN-2003/027 ): hence 1.4 10 9 Gev/s, assuming 8 10 6 s of running per year (160 days) at high lumi the most exposed tower (area ~13 cm 2, amounts to ~12 10 15 GeV/year (≈ 4 10 15 pe/year) The PMT is quoted to loose 20% of its gain after 3.6 KC. If we set that as a reasonable limit we must stay below 0.36 kC/year. 0.36 10 3 /4 10 15 = 90 fC/pe (this is not an issue if we run below 1600 volts). 8

9. HV setting proposal HV set to have saturation above 3.5 TeV in EM towers (low and intermediate  ) Fine adjust HV for HF quarters so to have highest possible gain compatible with these constrants (PMT have been assigned to various quarters in increasing relative gain groups in the following order: HF+Q1, HF-Q1, HF+Q2, HF-Q2 and so on) Q1Q2Q3Q4 High  1300 Med  13001250 Low  1250 1200 Very good situation: max operating voltage of PMT is 1700 so we have ample margins for compensating for gain loss. Moreover calibration with sources will be eased by having relatively high gains. Tiziano’s CMS Week June 2003 9

10. Saturation Levels 2744011.767.8439.22450001400 184807.925.2826.41650001300 144486.1924.12820.641290001250 112004.83.2161000001200 87363.7442.49612.48780001150 72803.122.0810.4650001100 Had 3.5 TeV in fCEM GeV in fCHad Gev in fCp.e. in fCGainHV 10

11. Summary Table HVGainCalibrationSaturation -11006.5x10 4 23 fC/pe4.7 TeV -11507.8x10 4 29 fC/pe3.7 TeV -12501.2x10 5 43 fC/pe (10.7 fC/GeV) 2.7 TeV -13001.5x10 5 52 fC/pe2.1 TeV -14002.3x10 5 79 fC/pe1.4 TeV -15003.5x10 5 118 fC/pe0.91 TeV Nural’s comment 11

12. Things to remember… Lifetime (HV ↓ ) Timing (HV ↑ ) SPE (HV ↑ ) Source (HV ↑ ) Equalization of gains? –Trigger ?, LUMI ? 12

13. HF HV Short Term Strategy Calibrations at Iowa 1100V In Bldg 186: 1350 V and 1250 V (sourcing) In 2007 UX5 we used systematically 1350V Since July 2008 1250V Suggestion: –Global Run: 1250V flat –Take Local Runs 1150 V 13