Feb C.Smith UVA EC energy calibration – g13 pass0 For pass0 data were cooked with CALDB calibration constants reset to nominal 10 channels / MeV –HV have not been changed since 2004 –Good opportunity to check absolute calibration of PMT without software corrections. In the early years absolute calibration based on cosmic muon runs, while minimum ionizing pions from data used to cross-check and make run dependent adjustments. More recently pion data used for absolute energy calibration Need to check consistency between pion and muon calibrations.
Feb C.Smith UVA Wrong pedestals for ADC in slot18 ADC swapped out after Run New pedestal run performed not until Run two days later ! This means between ADC in slot 18 was loaded with incorrect sparcification tables EC energy calibration – g13 pass0 On-call detector experts should insure that new hardware is updated promptly before data-taking resumes.
Feb C.Smith UVA Gain matching is reasonable but absolute calibration is low. Part of this is continued PMT gain drift since HV last adjusted in Another component is due to outdated attenuation length corrections (actual attentuation lengths are smaller than assumed). Currently Avto recooking several runs at my request to iterate software gain adjustments. EC energy calibration – g13 pass0
Feb C.Smith UVA Flash ADC SIS Channel 200 MHz 12-bit ADC / Wave Form Digitizer Sector 5 inner V13 - V20 3:1 UVA splitter For details see Serguei Pozdniakov, CLAS NOTE /3 1/3 EC Cosmics Test Setup
Feb C.Smith UVA Flash ADC Studies – Digitized PMT Pulses 200 time samples per event per channel Signal summed over time samples Remaining ~ 182 samples used to calculate baseline (pedestal) and subtracted event-by-event. Direct hit on single pixel Sector 5 Inner V13-16
Feb C.Smith UVA Two day cosmic run in January Cosmic run Saved FADC raw banks 2 GB file ~ 10 minutes 272 files…almost 2 days Missed events in FADC ? 1881 sparcification threshold
Feb C.Smith UVA Possible non-linearity Comparison of LC1881 and SIS3320 – charge
Feb C.Smith UVA 1881 ADC Channel SIS 3320 SUM Comparison of LC1881 and SIS noise 1881 noise (ac coupled) SIS 3320 noise Main issue for calibration purposes is noise relative to mip peak. Use pixel cut to isolate mip peak in cosmic runs
Feb C.Smith UVA Pixel cut for minimum ionizing muon peak u v w inner outer Acceptance fraction ~0.0009
Feb C.Smith UVA Pulse height versus distance to PMT - LC1881 and SIS3320 sigma / centroid = 0.30 sigma / centroid = 0.27
Feb C.Smith UVA Pulse height versus distance to PMT - LC1881 and SIS3320 sigma / centroid = 0.32 sigma / centroid = 0.33 Shapes of mip peaks similar – somewhat noisier FADC makes negligible contribution to sigma of mip distribution.
Feb C.Smith UVA Absolute light yield EC laser pulses attenuated to match ADC range of cosmic muons Measured sigma of pulse height distributions Two components of sigma: –Laser fluctuations –Poisson statistics Poisson component (and number of photoelectrons) could be separated using linear fit to sigma/centroid data Yield of 3-4 p.e. / MeV obtained