Andrea Giammanco CMS Tracker Week April DS ROD Prototype: “final” optohybrids “final” CCUM integrated in the rod with new FEC_to_CCUM adapter (Guido and Farooq) Noise (briefly) Noise (briefly) Physics signals Physics signals CCUM 5 XDAQ latest release (to be used in forthcoming beam test) New FEC supervisor Substantially higher speed Commissioned by Paolo & Laurent Problems of PMC very low trigger frequency cured in the XDAQ software (Laurent) for cosmic runs
Andrea Giammanco CMS Tracker Week April The ROD test setup See previous talks at the last tracker and cms weeks ! Alu box, gas tight, with patch panel for pipes ( C 6 F 14 cooling and dry air) and other services (it can house 2 rods) Modules, optohybrids, CCUM, high voltage Daq PC with 1 TSC, 1 FEC and 3 FED cards, DAQ Software
Andrea Giammanco CMS Tracker Week April Basic tuning procedure of DAQ settings well established in XDAQ discussed in previous system test presentations (Paolo) Standard APV settings Time alignment in the FED PLL scan Optimization of laser bias and gain [N.B. I sha =80 V psp =30 inverting mode] (24 and 2, with non individual tuning) Detailed study of noise & correlated noise as for SS rod repeated for all modules Excellent behaviour low common mode flat noise distribution
Andrea Giammanco CMS Tracker Week April Noise studied in all modules… CCUM ped i = ev tot tot : RMS of ADC i -ped i d d : RMS of 0.5(ADC i -ADC i+1 ) nrm nrm : RMS of ADC i -ped i -CMN 0 lin lin : RMS of ADC i -ped i -CMN i CMN 0 = strip CMN i = b+a i a = CM Slope FNAL modules old ceramic hybrid - DCU2
Andrea Giammanco CMS Tracker Week April Study of physics signals 106 Ru 61 (Q=3.5 MeV) Cosmics Analysis: tools developed by Roberto, Paolo, Andrea Trigger rate ~500 Hz Trigger rate ~0.5 Hz TSC gate = 6 ns Low voltage Hamamatsu PM Well assessed delay scan to find signal (Laurent)
Andrea Giammanco CMS Tracker Week April Cluster finding 1.S/N > 2 for every strip in a cluster 2.S/N > 5 for the highest signal strip Signal: sum of the signals of the strips Noise: noise of the seed Very simple cluster definition: 5 2 S/N strips cluster not a cluster (“seed”) No special treatment for bad strips
Andrea Giammanco CMS Tracker Week April Physics signals - source Source on the center of a sensor Deconvolution mode / inverter ON S/N=21.2 Det #1
Andrea Giammanco CMS Tracker Week April Physics signals - source peak mode inverter ON S/N=33.3 Det #1
Andrea Giammanco CMS Tracker Week April Det #7 tot tot : RMS of ADC i -ped i d d : RMS of 0.5(ADC i -ADC i+1 ) lin lin : RMS of ADC i -ped i -CMN i CMN 0 = strip CMN i = b+a i a = CM Slope nrm nrm : RMS of ADC i -ped i -CMN 0 Group of missing and shorted bonds missing bond (PA) missing bond (S1-S2) Partial signal recovery from neighbours:
Andrea Giammanco CMS Tracker Week April Physics signals - Cosmics ~20% higher than rays Deconvolution mode / inverter ON S/N=25.6 Det #1
Andrea Giammanco CMS Tracker Week April Time delay scan N clusters /N triggers vs delay N clusters /N triggers t=0 Not the detector efficiency! upper =N upper&lower /N lower =99.9% lower =N upper&lower /N upper =99.8% noise of scintillator spurious triggers (not ideal collimator) Use only events seen by the other detector: Det #7
Andrea Giammanco CMS Tracker Week April Source on the bonding N clusters /N triggers t=0 subtracting Source on the center Source on the bonding Det #7 Evidence for region with lower charge collection efficiency near the edge normalized to same area
Andrea Giammanco CMS Tracker Week April Source on the overlap CCUM x axis
Andrea Giammanco CMS Tracker Week April Alignment 7-1: strips 7-8: strips CCUM Rod alignment pins were not integrated! x axis 2-8: strips ( m) ( m)( m)
Andrea Giammanco CMS Tracker Week April “Tracking” with 2 hits CCUM Deviations from linearity due to the finite size of the source 3 mm 7 1 m=tg Slope of the line passing through the hits in 7 & 1 vs Position on 7 x position of the source
Andrea Giammanco CMS Tracker Week April “Tracking” with 4 hits CCUM Slope of the best fit line for the hits in 7 & 1 & 2 & 8 vs Position on 7 3 mm 8 mm x axis m=tg x position of the source
Andrea Giammanco CMS Tracker Week April Conclusions Fully equipped DS rod thoroughly tested also with physics signals: no problem found. Analysis of physics signals provides a complementary view of detector defects. Simple method to qualify rod geometry commissioned. Calibration of 106 Ru 61 vs high energy cosmic muons performed 106 Ru 61 : Cosmics: deco: peak: S/N=21 S/N=33 S/N=25 If muons 500 m) = electrons noise deco) = 1600 electrons identical to predictions.
Andrea Giammanco CMS Tracker Week April Electrons vs capacitance
Andrea Giammanco CMS Tracker Week April Noise: module #7 FNAL module - old ceramic hybrid - DCU2 To be compared with performance on physics signals (deconvolution) ped i = ev tot tot : RMS of ADC i -ped i d d : RMS of 0.5(ADC i -ADC i+1 ) nrm nrm : RMS of ADC i -ped i -CMN 0 lin lin : RMS of ADC i -ped i -CMN i CMN 0 = strip CMN i = b+a i a = CM Slope
Andrea Giammanco CMS Tracker Week April Noise: module #1 FNAL module - old ceramic hybrid - DCU2 To be compared with performance on physics signals (deconvolution) ped i = ev tot tot : RMS of ADC i -ped i d d : RMS of 0.5(ADC i -ADC i+1 ) nrm nrm : RMS of ADC i -ped i -CMN 0 lin lin : RMS of ADC i -ped i -CMN i CMN 0 = strip CMN i = b+a i a = CM Slope
Andrea Giammanco CMS Tracker Week April Noise: module #8 FNAL module - old ceramic hybrid - DCU2 To be compared with performance on physics signals (deconvolution) ped i = ev tot tot : RMS of ADC i -ped i d d : RMS of 0.5(ADC i -ADC i+1 ) nrm nrm : RMS of ADC i -ped i -CMN 0 lin lin : RMS of ADC i -ped i -CMN i CMN 0 = strip CMN i = b+a i a = CM Slope
Andrea Giammanco CMS Tracker Week April Noise: module #2 FNAL module - old ceramic hybrid - DCU2 To be compared with performance on physics signals (deconvolution) ped i = ev tot tot : RMS of ADC i -ped i d d : RMS of 0.5(ADC i -ADC i+1 ) nrm nrm : RMS of ADC i -ped i -CMN 0 lin lin : RMS of ADC i -ped i -CMN i CMN 0 = strip CMN i = b+a i a = CM Slope
Andrea Giammanco CMS Tracker Week April Det #7 ADC i -ped i -CMN i CMN 0 = strip CMN i = b+a i a = CM Slope Non gaussian noise! ADC i -ped i -CMN 0 kurtosis