Description of mMvdDbase

Here is the calling sequence with links to the source code:
mMvdDbase dMvdGeo dMvbPar dMvcPar dMvdPar dMvdDbasePar dMvbDbase dMvcDbase dMvdDbaseStatus

This module is called only once -- at the beginning of the first event. It reads various "database" files and fills the appropriate structures. It will eventually be modified to read from a database rather than from local ascii files.

The "database" information is read from 6 local ascii files:
MVD_noise_barrel.dat
MVD_cross_talk_barrel.dat
MVD_noise_pad.dat
MVD_cross_talk_pad.dat
MVD_dead_channel.dat
MVD_bcal.dat
If these files are not found, the code will setup some default values, but this is not recommended and you will not get the "standard" results without these files.

This module fills the MvbDbase table. The entries in this table contain

  • flags for dead/alive channels (MvbDbase[index].flag_dead_channel = 0 for a dead channel),
  • the noise level in a channel (MvbDbase[index].noise expressed as sigma in keV, where 119KeV corresponds to a 1 mip signal, so bDbase[index].noise = 11.9 keV is defined as signal/noise = 10/1 = our default).
  • estimated crosstalk with neighboring strips is stored in bDbase[index].cross_talk, expressed as a fraction from 0 to 1. 0.01 is the default value.
  • MvbDbase[index].calib[0] ... MvbDbase[index].calib[4) contain parameters used to convert an ADC value into the estimated number of particles.

    More details on how the noise and crosstalk parameters are used can be found in the documentation for mMvdSetUcal

    There is a similar table for the pads, called MvcDbase, but it does not contain the calib constants. The ADC values from the pads are used directly in the later analysis with a default calibration which converts ADC channels to mips (31.75 ADC channels = 1 mip) is assumed. The number of particles is calculated from this by assuming all particles are mips. This is clearly an area that needs work -- there should be calibration constants.

    We are not happy with the way the calibrations work currently. We are working on an improved calibration procedure. This does not significantly affect the vertex-finding procedures, but it does effect the dN/deta calculations.

    John Sullivan
    comments to: sullivan@lanl.gov
    updated 23-Dec-1999