Sadly, the run that was going on Tuesday through Wednesday ended Wednesday night at about 9:00 pm because the server PC rebooted itself for some reason. I don't yet know whether it was a power outage, or Windows automatically rebooting itself like an idiot after installing automatic updates (although I thought I turned that "feature" off), or something else (maybe a cosmic ray!). I should probably check the system logs for clues. Although, unless it happens again frequently, I probably won't bother with that for now.
I started a new run with the two scintillator paddles stacked on top of each other, to see how this affects the rate of coincidences. We expect that this should increase the rate of coincidences substantially, because now, even isolated particles that pass vertically through the stack will often intersect both paddles and trigger both PMTs. Indeed, I can visually perceive a high number of coincidences happening on the scope - a significant fraction of over-threshold pulses on either PMT are accompanied by a pulse (although not always an over-threshold one) on the other PMT.
Meanwhile, while that is running, I'm starting a re-analysis of the run that ended Wednesday night, just to see if there were any interesting variations in the pulse rate between 3:00 pm (the end of the part I looked at the other day) and 9:00 pm.
Run start time: Tue Sep 06 17:04:03 2011
Run stop time: Wed Sep 07 21:04:34 2011
DAC settings / comparator threshold levels (mV): -350,-500,-650,-800,-950,-1100.
Length of run: 28 hours, 0 minutes, 31 seconds = 1,680.5 minutes.
- PMT1 (near window, in "Case #1"):
- Total # of (over-threshold) pulses = 1,369,385;
- avg. pulse rate = 815 pulses/min. = 13.6 pps.
- PMT2 (near hallway, in "Case #2"):
- Total # of (over-threshold) pulses = 190,183;
- avg. pulse rate = 113 pulses/min. = 1.89 pps.
- We concluded the other day that this 2nd gun case just has a "bad paddle" in it - i.e., there is something awry with the scintillator itself, or the PMT tube, or the interface between them that causes the paddle's sensitivity to be reduced. The base itself is OK.
Also: I added some new code to anal-pulses.sce on Wednesday to check for "false" coincidences, that is ones with time differences that are just outside the range of time differences that could possibly be due to real EAS's passing through. Time differences from 0 to +/- ~25 ns we generally assume to be potentially "real" coincidences (due to actual showers). By counting the number of coincidences in the range from +/- ~30 ns through ~50 ns (which cannot actually be due to shower fronts propagating at c, due to the maximum distance between detectors in our lab of ~25 ft.), we can get a rough picture of how many of the assumed-real coincidences might actually be false positives, caused by simple chance or by other mechanisms not involving lightspeed particle fronts (e.g., power surges?). Due to my earlier analysis of the coincidence rate based on the Poisson distribution, I have concluded that generally the number of false positives will be very small, but it is good to get independent confirmation of this from the data.
This seems to have worked as expected; the number of false-coincidences seen in this run was only 2, compared to ___ (assumed) "real" coincidences. Let's now take a look at the time deltas of the false coincidences - if they were right on the edge (30 ns, say), then they could potentially be real coincidences whose time deltas simply got mischaracterized due to imprecision in our time measurements.
I have to run over to ECE now but I'll be back for the group meeting after 4 pm.
No comments:
Post a Comment