Final calibration of Towers
for 2005 run p+p data
using MIPS with 00xx00 signature in SMD


    Goal: determine absolute tower gains from fit of Gauss+Landau to 'd'-spectra and subtraction of pedestal residuum ('a'-spectra)

    Formula: gain(ch/GeV) = gain=2.89*TanH(eta)*(MIP-Ped)
    err^2= errMip^2 + errPed^2 + 0.3^2, last 0.3 accounts for peds systematic error

  1. INPUT: ~700K minB triggered events from runs: 6171037, 6172091,6173068,6173977.
    Those runs belong to 3 fills with very low beam background.
  2. Table with 720 tower spectra , raw Landau fit results

  3. Summary of new failed channels
    1. masked(broken) towers 02TC06, 10TB06, 08TA07, 06TD11,
    2. towers with non-fatal pathologies:
      08TD01-03,09 - bimodal pedestal , ~5 ADC ch separation
      08TE05 - spikes in raw ADC
      04TA04 - pedestal offset >1 ADC
      08TC05, 05TE07 - only pedestal seen
    3. towers with broken post shower due to (temporary) copyCat in crate 75 ch56-58: 02TA11, 02TB03, 02TB01
    4. there are few more smaller problems. I did not tried to fixed them, since gains for a subset of 20 towers with changed HV are needed.

  4. Comparison of CuCu & pp gains for 20 towers with changed HV.

    tower P-name MIP gain (ch/GeV) ratio: XXgains/ideal
    Cu_MIP pp_MIP Cu_slope/.. Cu_MIP/.. pp_MIP/..
    01TA04 P01TA04 14.14 +/- 0.88 17.24 +/- 1.08 0.76 0.57 0.70 , HV changed
    01TA05 P01TA05 26.82 +/- 0.90 24.82 +/- 1.26 xxx 1.00 0.93 , HV not change
    02TE02 P02TE02 13.75 +/- 0.94 21.09 +/- 1.40 0.81 0.66 1.01
    05TD10 P05TD03 54.84 +/- 0.97 38.09 +/- 1.31 1.37 1.50 1.04
    06TD11 P06TD11 78 +/- 3 broken 1.72 2.03 broken
    06TE04 P06TE04 19.91 +/- 0.96 24.99 +/- 1.50 0.85 0.81 1.01
    07TA01 P07TA01 16.07 +/- 0.95 19.64 +/- 0.86 0.84 1.03
    07TB09 P07TB09 71.64 +/- 1.55 42.19 +/- 1.95 1.61 2.07 1.22
    07TC11 P07TC11 34.19 +/- 0.81 41.27 +/- 1.44 0.81 0.88 1.07
    07TE02 P07TE02 15.77 +/- 0.91 20.23 +/- 1.15 0.86 0.76 0.97
    07TE06 P07TE06 19.33 +/- 0.98 26.37 +/- 1.20 0.79 0.67 0.92
    08TA11 P08TA11 36.46 +/- 0.91 43.05 +/- 2.87 0.85 0.94 1.11
    08TC03 P08TC03 17.03 +/- 0.97 19.07 +/- 1.30 0.85 0.75 0.84
    08TC07 P08TC07 39.87 +/- 1.02 31.44 +/- 1.40 1.20 1.30 1.02
    08TD08 P08TD08 36.88 +/- 1.15 43.07 +/- 2.24 1.19 1.13 1.32 , BAD*)
    08TD09 P08TD09 28.982 +/- 1.000 23.22 +/- 1.0 0.84 0.84 0.67 ,BAD, SL*)
    09TE02 P09TE02 16.75 +/- 1.500 21.38 +/- 1.25 0.85 0.80 1.03
    09TE08 P09TE08 27.55 +/- 0.87 31.82 +/- 1.17 0.85 0.84 0.97
    10TA09 P10TA09 44.16 +/- 1.25 36.3 +/- 1.47 1.33 1.28 1.05
    12TD01 P12TD01 25.92 +/- 1.10 28.62 +/- 2.27 1.23 1.36 1.50 ,BAD*)
    12TD09 P12TD09 26.56 +/- 0.88 34.53 +/- 1.99 0.86 0.77 1.00

    'BAD' - if CuMip gain is closer to ideal gain than ppMip gain: 08TD08 , 12TD01
    'SL' - post shower broke, relative slopes used instead of MIP peak 08TD09 ,

  5. Similar table for another 20 towers for which HV was not changed may serve as reference .

    Script used to merge CuCu & pp gains: fixGain4pp.tcl , data : hvChange.dat

    set my(iGain) {19.035 20.846 22.702 24.701 26.724 28.751 30.759 32.725 34.629 36.603 38.643 40.415};# ideal gains vs. eta bin, in ch/GeV