The distributions of directly measured preamp gains and preamp/VPTT gain products still do not match the indirectly measured ones (from the units already glues) - mean and width are different
Is there a problem with the light pulser reference? Tobias was asked to remeasure some of the oldest glued units again to see if the same normalised response results
Is the statistics of only two units the calculation of the gain of the already glued units is based on too weak?
Is the light pulser measurement suitable at all? We inject form the read out side of the crystal; the light has to pass the crystal 'back and forth' and the influence of the reflective foil alignment is still questionable
We currently estimate the effect by measuring the distribution of light entering the two sensors of an APD read out crystal (earlier estimations of that kind were dominated by the "1:2 intensity effect" of failed APD couplings)
Comment form Pawel: Is it insured that the preamps do not change gain during measurement because of heating up components? Werner in the meantime measured the gain of the preamps as a function of ambient temperature: 1% change between -25 and 70 degrees C