Drew Alton University of Michigan

1. Status of the Central Fiber Tracker(CFT)Central Preshower(CPS)and Forward Preshower(FPS)or CFT/PS for short Drew Alton University of Michigan

2. Outline • Introduction • History • Status • CFT Efficiency • CPS results • FPS progress • Issues • Plans • Conclusions

3. Introduction • ~100,000 Channels • CFT (76,800 channels) • 8 axial layers • 8 stereo layers • 4 u and 4 v • CPS (7680 channels) • 1 axial layer • 2 stereo layer • 1 u and 1 v • FPS (14,884 channels) • 2 MIP layers • 1 u and 1 v • 2 shower layers • 1 u and 1 v

4. Data to Level 3 Scintillating or wavelength shifting fiber G-Link Cable PS VRB VTM CFT Sequencer Waveguide VLPC cassette AFE boards Electronics LVDS cable to L1 Cryostat AFE backplane PS and CFT readout Led Illuminated

5. Recent History • July-Oct. 7, 2001 2 prototype AFE’s on Platform • Jan 21, 2002CFT Axial complete (again) • April 17, 2002CFT and CPS complete • June 29, 2002FPS complete

6. VLPC Cryostat • Cold block stable to 100mK • Cassettes stable to 50mK • Impossible to state how important this is • It’s a statement on how well behaved things are that no one (except GG and RR) thinks about this anymore

7. Issues overcome this year Everyone • AFE Power supply ‘flickers’ and trips • Developed and improved calibration techniques • Latching boards • Readout turning off • Split pedestals • Non-uniform pedestals • Pedestal moving when applying zero-suppression threshold • Move readout by 264 ns

8. Pedestal Stability 68 • Pedestals for an SVX as a function of time. • Mean of the 64 channels • RMS of the 64 means (uniformity). • Mean stable to <1 ADC for months • RMS stable to < 0.5 ADC for months 58 9 3 March October J. Warchol, Y. Jiang Seq Firmware to remove ped shift after threshold Swapped board… cryo modifications Clockgen for Uniform ped First 264 Read attempt 264 Read all ticks

9. Make a ‘hit map’ (fired fiber distribution) Fill a histogram with # of fires for each strip in a layer. Find mean and RMS Define ‘hot’ as >mean+5s Using s from standard run ‘hot’ is usually only warm Definition of hot still under refinement Define ‘dead’ as< mean-5s Using s from standard run Data Quality Efforts D. Lam, M. Hildreth

10. Hot and Dead Channels

11. Layer A AU B BV C CU D DV E EU F FV G GU Efficiency 98.880.02% 97.700.02% 99.130.02% 97.160.02% 99.140.02% 96.690.02% 99.170.02% 98.180.02% 97.060.02% 97.910.02% 99.620.02% 97.690.02% 99.220.05% 95.970.02% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Efficiency B. Wang, S. Kulik, A. Kharchilava, M. Drazil • Layer efficiencies • Build 15 out of 15 layer tracks, ask if 16th layer is there. • Z->ee tracking efficiency (in ||<0.8) • e =46.4%-65.2% (depending on alg.) from TARC • e = (64.0 ± 4.7 ± 7.0)% from EMID yesterday • Z->ee efficiency, from L3 tracking (in ||<1.2) • e = 65% • If Run>150,000 e = 80% • If 7 or more hits e = 94% R. Zitoun D. Whiteson

12. Min-Bias Occupancy A. Bellavance • Occupancy/layer • Zero bias occupancy • Min-bias events • Physics trigger occupancy • 1-8 Axial, 9-16 stereo (two conventions) Min-bias events Zero Bias running Physics running

13. CPS MIP studies A. Askew • MIP peak in ADC counts for two gains. • As we currently run (in MIP mode) • A small step (x2) towards shower mode.

14. EM Identification • pt>15, |ID|< 20, iso<0.15, Emfr >0.9, HMx <100 • Rejection • Run>151,831 and 1 of two EM in CPS fiducial. • One of two EM has cps_cal_match_prob>1e-4 • Each EM in CPS fiducial has match. • Efficiency • Track match • Track and CPS match • Rejection  0.8(0.6), Efficiency 0.9

15. FPS A. Patwa, M. Hildreth, Y. Mutaf, A. Turcot • Fully instrumented for 3 months • P13 now has an FPS cluster chunk, so one can match the FPS to tracks or calorimeter objects in the thumbnail. • Still issues to resolve

16. Efforts in progress • Trying to monitor tick and turn to insure that there is no event mixing • VRB software (G. Brooijmans, etc) • Sequencer software (F. Borcherding, M. Utes, L. Xuan) • Developing Software for Monitoring the data quality. (D. Lam, M. Hildreth, DA) • Improve Operations • Minimize Dead time • Eliminate FEB’s • Eliminate Missing (incomplete) events • Respond to unusual events • Optimize AFE performance (F. Borcherding, J. Warchol) • Improve performance (Discriminators/Gain/Bias studies) • Hot Regions/ Dead Regions

17. Hot/Dead Regions • Regions of the detector (multiples of 1024 channels) are on ‘all’ the time, or are off ‘all’ the time. • These problems appear without warning • Easy to ‘fix’ • Used to appear around once a day, they now occur a few times a shift. • This appears to be related to the Sequencers. • This is our highest Priority issue.

18. Plans • Resolve Hot/Dead regions • Reduce discriminator threshold • Efficiency: Attempt to identify problems clustering • Automated run quality measurement • Improved control room tools and documentation • Areas which need attention • Use DB for CFT/PS reconstruction • Offline analysis aimed at understanding and improving detector performance • Tuning CFT/PS Significant Event Server alarms. • Preparing the CFT/PS for Comics downloads

19. Conclusions • CFT • Over 99% channels operational • Layer efficiencies ~99% for axial, ~97% for stereo • CPS • Three D cluster to track matching efficiency >90% • Beginning implementation for EM identification • FPS • Offline reconstruction • True studies can now begin • Optimization continues