Minutes of CALCOM meetings, December 1997. ========================================== During the Commissioning and E1 runs in December, more or less regular CALCOM meetings were held to discuss issues related to the ongoing run, Not all the goals were achieved, however, this effort proved still very useful as it initiated activities in the analysis sector, and focussed the analysis work on tasks that were relevant for the progress of the run. I would like to thank everybody who contributed to this effort and I wish you happy holidays and all of us a very successful CLAS physics year! Hope to see you all in January 1998, Volker Burkert ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Contributions by: S. Stavinsky - Cherenkov counter efficiencies L.C. Smith - Trigger studies and EC calibration R. Minehart - Elastic ep studies and theta/phi anomalities R. Thompson - Trigger studies K. Joo - Trigger efficiency studies S. Taylor - TOF calibration studies L. Elouadrhiri - Status of particle id using tof/momentum E. Smith - TOF summary report W. Brooks - EC timing calibration status J. Manak - Elastic ep scattering rate studies Additional contributions were made by: B. Niczyporuk, but no report was received. +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ S. Stavinsky: Cherenkov Counter efficiency studies. ------------ Cherenkov counters efficiency for 1,4,5,6 sectors have been tested in off-line for raw data accumulated in run 7710 with EC trigger (threshold 250 mV). 10^6 events have been included into time-based analysis. Efficiency was estimated for elastic scattered electrons with test of EC-DC matched point and EC deposited energy. Results for Cherenkov counters efficiency are in a good agreement with simulations. (See /home/stavinsk/efficiency7710tb.ps). L.C. Smith: Trigger studies and Calorimeter calibration ---------- Absolute calibration and energy resolution of the forward angle calorimeters is being checked using data taken during the Sept. commissioning run. PMT HV were adjusted prior to the run using cosmic-ray minimum ionizing events to produce an average ADC calibration of 10 channels/MeV. Individual PMT calibrations were also extracted from cosmic-ray data. Pizeros were identified from events w/ two neutral EC reconstructed hits/sector. Electron ID was based on EC reconstructed hit matched with DC negative track. Run 6343 (4 GeV) was used for pizero events (w/TRK HBT), Run6079 (1.6 GeV) for electrons (w/SDA TBT). SEB was used for track matching. Results: Sector 1 2 3 4 5 6 pi0 cen (x100) 6.8 0.7 3.0 1.4 1.7 3.2 pi0 rms (x100) 13.7 11.4 10.3 12.8 11.7 11.1 e- cen (x100) 7.3 -2.5 3.1 -0.5 6.7 6.4 e- rms (x100) 10.1 10.5 8.5 10.8 9.6 9.5 pi0 cen: centroid of (invariant mass)/mpi0 - 1 e- cen: centroid of (ec energy/dc momentum) - 1 Further improvements can be expected as in situ attenuation lengths from cosmic ray runs are introduced (currently the original bench measurements are used). Also the effects of pedestal noise are under study. More commissioning data will be analyzed as TBT at 2.4 and 4.0 GeV becomes available. cole@apollo.phys.virginia.edu R. Minehart: Dependence of W position of elastic ep on Theta and phi. ------------ Ralph Minehart showed some early results of a study to check the combination of torus field, wire-chamber survey, and whatever else enters into the determination of track momentum. He used a set of 1.6 GeV negative track momenta that had been calculated with Bogdan's SDA program. He calculated the missing mass associated with these tracks and studied the width and location of the elastic peak. He found that the peak position varied from one sector to another over a range of about 35 MeV. He also found a theta dependence within a single sector. He will look at this in more detail to try to find the source of the variation, or a correction factor for it. R. Thompson: ----------- An examination of the trigger electron selectivity at 1.6GeV beam energy was made. Electrons were identified my requiring negative tracks with 4.0 degree hit matching in the Cerenkov and 60cm hit matching in the calorimeter. The tracks were from hit-based reconstruction. A 10cm fiducial cut was applied to the hit matching in the calorimeter. Each event was classified into one category in the order shown below. For example, if an event was found to have at least one electron, that event was categorized as an "electron" event and it was not classified into any of the categories that follow in the list. The following table shows the percentage of events of each category for the runs shown at the top of the table. The run conditions are listed below. 7718 7721 7725 7851 electron 22 26 28 30 n.f. electron 8 9 9 10 positron 2 2 2 2 n.f. positron <1 <1 <1 <1 other negative 11 11 11 13 other positive 9 9 9 11 no FC match 19 16 16 16 no track 29 26 24 18 n.f. == non fiducial other == not an electron or positron no FC match == there was at least one track in the event, but none matched to the forward carriage all runs are with EC*CC trigger 7718: 75mV EC thld., 20mV CC thld 7721: 100 " , 20 " 7725: 100 " , 30 " 7845: 100 " , 20 " It was observed that with 120mV EC threshold, the W spectrum was being cut into at around 1.55 GeV relative to 100mV EC threshold. Runs 7851 and 7721 differ only in time and the difference in electron selectivity between these two runs is not understood. Future studies: Repeat above study for 2.4 GeV. Normalize spectra to Faraday cup. Examine bias of EC*CC trigger by studying EC trigger with same discriminator thresholds. Kyungseon Joo: ------------- I did electron trigger studies using SEB (Simple Event Builder of RECSIS). I analyzed two runs, run 7893 with EC.CC trigger (100 mV for EC and 20 mV for CC) and run 7841 with EC trigger (250 mV). I analyzed sector by sector basis and by dividing one sector into 6 small regions, I studied trigger uniformity over a sector. I also studied electron id using different detectors (EC, CC, SC and DC). From run 7893, without fiduical cut on EC, I got 39.4 % electron events from all trigger events from electron particle id using DC and EC and got 33.7 % electron events using electron id from DC, EC, SC and DC. There are 5-10% variations between sectors. I found pretty good uniformity by comparing small regions over a sector. For run 7841 (EC trigger only), I made a fiducial cut on EC to make EC the same size at CC. I got 11.7% electron events using electron id from EC and DC. I got 10.7 % electron events using electron id from EC, DC and CC. S. Taylor: TOF Calibration/Efficiency studies --------- The status of the gain-matching and energy calibration of the TOF was presented. The pedestal-subtracted geometric means of the ADCs corresponding to normally incident minimum ionizing particles were measured and the energy-calibration values for left tubes and right tubes were calculated from the intercept of plots of ln(ADCL/ADCR) vs. position along the scintillator. The ratios of these NMIP ADC values were approximately one to within +/-30% or so with some exceptions, indicating that the quality of left/right gain-matching needs some improvement before future runs for which the TOF is part of the trigger. The attenuation lengths for each of the counters were also obtained from the ln(ADCL/ADCR) vs. y plots. The values agreed with expectations, since the longer paddles had longer attenuation lengths than the shortest paddles. Finally, the quality of the left-right time-matching was shown to be quite good except for the very forward angles. The second topic presented was a study of the TOF efficiency using cooked data from run 7845. Tracks from hit-based tracking were projected to the scintillator planes and predictions for which scintillators should have fired were made based on the TOF geometry. Two cases were presented: one for which an exact match between predicted id and actual id was required and one for which the prediction could be off by +/-1 paddle. For the former case the efficiency hovered around 80-90% with a drop in the back angles for sectors 5 and 6 and drop in the forward angles in sectors 2 and 3. Expanding the cut (the latter case) marginally improved the efficiency. The "inefficient" events seem to fall into three categories: (1) events with tracks that clearly pointed at scintillators that did not fire; (2) events with particles whose trajectories were bent sufficiently toward the beam pipe so that they missed panel 1; and (3) track reconstruction failures. A set of 50 inefficient events were studied using ced. Case (1) occured 40% of the time; case (2), 20%; case(3), 40%. During the meeting a suggestion was made to use time-based tracking to beat down the percentage of tracking failures. L. Elouadrhiri: Particle Identification and Tof calibration -------------- This requires matching the information for each track between the drift chamber and the TOF system, together with a good electron identification. Using SEB (Simple Event Builder) we selected events for which an electron and at least one other charged particle are present. The tracks reconstructed with DC are required to match a hit in the TOF system. The electron is identified with the calorimeter and Cherenkov. Knowing its path length we determined the flight time, by subtracting this time from the TDC in the TOF we can determine the t0 of the event. From this information and the TOF TDC for the other charged particle, present in the event, one can calculate the mass of the hadron. In the first step of this calibration procedure, we selected elastic reaction by that the the electron is in Panel 1, (sector 1,4,5 and 6) and the hadron in the opposite Panel 2 (opposite sector) and with an additional cut on dE/dx we selected that the hadron is a proton. Assuming that the calibration counter to counter is done within one panel one can use this to calculate the timing calibration constant from panel to panel. The result of this study showed that the counters within one panel were calibrated to a few ns which is not sufficient to extract the panel to panel calibration. We showed also that this method works to calculate the time constant from counter to counter by requiring that the electron and the proton are in specific counters. To determine these calibration constants from counter to counter, we will use pion and proton electroproduction at 2.4 GeV with an automated procedure. E. Smith: TOF Quick Overview 12/22/97 --------- HV for all channels was updated before the beginning of e1 data taking and not changed. 1. Bad Channels Sector 1 2 3 4 5 6 16R - 48AL 1L ADC 48AL 1R 44R 48B OFF 33R 48B OFF 30R ADC 45R 2. Gain Matching (Exceptions are about 2 channels per sector which were fixed and HV not updated before run began): Channel m.i.p. matching corresponds to 600 +/- 200 channels, +/- 100 channels in most cases. Matching Between left and right sides: ratio = 1 +/- 0.5, +/- 0.2 in most cases. Some systematic shifts are observed between the forward and back angles. 3. Time matching between channels is of the order of a few ns and updated in the map. 4. Matching of Tracks to TOF scintillator hits (Approximate averages, all tracks): Sector 1 2 3 4 5 6 Panel 1 >90% 60-90% 60-90% >90% 90% 90% Panel 2,3 >90% 90-70% 90-80% >80% 80-70% >70% Panel 4 >70% >70% >75% >70% 65% 70% From 50 events scanned that had no match: 40% no TOF hit, 40% poor tracking, 20% FA acceptance miss. Notes: a) Find more misses for positive tracks b) misses seem to be correllated with triggering, probably due to enhancement fraction of electrons vs protons, etc. W. Brooks: --------- The CLAS forward angle electromagnetic calorimeter timing reconstruction relative constants have been determined from October data to an accuracy of 1 - 2 nanoseconds for all six sectors, and have been entered into the calibration mapmanager. This will permit the start time for the drift chamber reconstruction to be calculated for tracks which impinge on a TOF bar which does not have two operational tubes, up to 45 degrees. The present method of calibration uses internal calorimeter data only, reconstructing the inner and outer calorimeter mean times independently and using these to fix the tube-to-tube (relative) constants for events with a simple topology. The procedure is implemented using an automated PAW routine for all 6 sectors, deriving 1296 constants in an iterative correction algorithm. An accuracy sufficient for drift chamber reconstruction has been achieved without performing walk corrections. Achieving the ultimate detection accuracy of approximately 200 ps will require further effort. An evaluation of the EC-TOF plane relative times is underway. J. Manak: Elastic ep scattering -------- The elastic rate as a function of beam current was measured by Joe Manak. 4 runs were examined: run 7730 at 1nA(80% live), 7731 at 2nA, 7744 at 3nA and 7745 at 4nA(40% live). The number of elastic events per faraday cup count was found to be constant as a function beam current at about 8 elastic events per 1E-10C of beam as measured by the Faraday Cup. The elastic cross-section was constant accross all runs and was found to be 50-70% of expected(w/o rad. corrections). A Time based analysis of run 7745 was also shown, 50% of events with a single negative track seen in hit based tracking were reconstructed in time based, in the central region of the drift chambers (along the centerline) the hit to time based eff. was 70-80%. The postscript files of this analysis can be found on: http://www.cebaf.gov/~manak/offline_analysis_dec97.html