The results of CAL pedestals calibration in LEX8 and HEX8 ranges for the first 44 runs collected on orbit showed a significant bias with respect to the pedestals actually used for the data processing in the pipeline (they were measured during cold thermal balance period of Observatory TVAC in January 2008).
The plot below shows the trending of pedestal difference with respect to official pre-launch pedestal file for all LEX8 channels over 44 runs:
Similar plot for all HEX8 channels:
The reason for this pedestal drift is higher temperature on orbit with respect to cold_TVAC. Pedestals for the first 12 runs demonstrate some evolution, while CAL temperature was increasing and after that became stable. The maximum pedestal drift is 30 adc units, which corresponds to 1 MeV in LEX8 range and 60 MeV in HEX8 range. It is not very big bias for energy measuremet, but it strongly affects the position measurement along the crystal as the 1% bias of the signal at one crystal end leads to the 5 mm bias in position. This effect is the biggest in HEX8 range starting at 1 GeV, so 60 MeV bias means 6% leading to 3 cm position bias.
To avoid these consequences we request to replace the existing pedestal file with the new one:
/nfs/slac/g/svac/chehtman/calibGenCAL_analysis/flight/ped/calPed_28jun08.xml
produced genMuonPed application of calibGenCAL package revision calibGenCAL-05-06-00 from calibOps run 236323982 collected on Jun 28,2008.
Anders Borgland has reprocessed 10K events from the run 236323982. The following two plots show the sum of pedestal histograms for all LEX8 channels and for all HEX8 channels. 
