R.J. Feuerbach: ND calibration (Oct 4)

I've taken a more active role in the ND timing calibration, helping Jon to complete it. What has been found:

Cosmics

  1. Comparing "lined-up" timing calibrations, for bar vs. bar+1 compared to bar+2 cv1_xloc.ps. The stars compare the times of neighboring bars, while triangles are comparing to bars 2 away. There is a difference at the 300-500ps level. This is due to WHERE in the upper bar the cosmic ray entered the active region of the detector (and thus the effective pathlength).
  2. Energy deposition pattern for multiple hits adcs_v_pos.ps
    3. : of course the longer path through the active leaves more energy. This was used to determine where in the bar the cosmic "started".
  3. Timewalk correction: found that the form TW  c; (ADC-ped)^EXP with EXP=-1 instead of the prior value of -0.5 works better. This is the dependence to be expected if the "shape" of the analog signal is nearly-linear at the level of the threshold. The values of TW are consistent with a rise time for the 2'' PMT's of 3 to 5ns (checking if this is reasonable).
  4. Final cosmic calibration for run 4402: Mean Time differences, Y-position differences. This run was supposed to be AFTER all HV adjustments to the ND were made. However, comparing to run 4490 shows that something did change again. For cosmics, sigma of the difference of meantimes, eg: 0.5*(lt1+rt1)-0.5*(lt2+rt2), is 0.4 or less for CMU.

Beam data

Used production run 4490 (kin 4). The (not-final) time distributions look like:
  1. Plane-to-plane Mean-time differences prod4490_p2pMnDiffs.ps
  2. Intra-plane Mean-time differences (neighboring bars fired for horizontal tracks -- low energy deposition) prod4490_MnDiffs.ps