Results of the Calibration

The quantity of interest in the phototube gain calibration is the SPE peak position. The goal is to have the SPE peak position be in the same place for all 216 channels. This is primarily so that the trigger threshold has a consistent meaning across the entire detector, but is also used in the reconstruction of the hit position in the Cerenkov, and is therefore of some interest in that area as well.

The latest run for which this calibration has been performed is runs 8916-8922, taken on Sunday, 22 February, while the beam was off due to a trip in the CHL. The plot below shows the fitted positions of the SPE peaks.

The points to note from this plot are as follows:

1. The SPE peak positions are coming much closer to channel 150; this is a sign that the "calibration" phase is nearly finished.
2. Sectors 2 and 5 still have some problems in the calibration. Sector 2 is probably due to the fact that it's one of the last two sectors to be installed, and it simply takes more than one iteration to get the gains perfectly matched (although Sector 3 looks pretty good in that respect). The problems in Sector 5 are due to misidentification of the SPE peak, as discussed below.
3. The last four PMTS in Sector 3 had an electronics problem that prevented them from triggering. The points shown for those tubes are estimates based on previous calibration runs.
4. Some of the SPE peaks were hard to identify; when looking through the various plots, I got some of them wrong, and changed the voltage on those tubes based on that misidentification. This was the case for sector 4, channel 29, and sector 5, channels 14, 16, and 22. In the future, the voltages will be corrected.
5. Some of the SPE peaks are extremely hard to fit due to the presence of an exceptionally high background. This is the case for sector 1, channel 19; sector 5, channel 9; and sector 6, channel 32.

The previous calibration run was runs 8630-8635, taken shortly before the February run started. The PMT voltages were not changed between the two runs, so they may be used together to determine whether (and by how much) the PMT gains have drifted. This shift is shown in the plot below.

Positive values on this plot mean that the gain of the corresponding PMT has risen in the (roughly) two weeks between these runs. The dashed lines correspond to shifts of ±15 channels, or 10%. It is curious that most of the large shifts are in sector 2; I do not yet have a good explanation for this, and will be monitoring that sector closely.

The next opportunity to collect SPE calibration data will be this Monday, 2 March. I will do so then, and will update this page with the results.