Tue., Nov. 1st

The specs of last night's run are as follows:

• SMA#1 <-- Gun Case #1 @ North end of room (near interior hallway)
• SMA#2 <-- Gun Case #2 @ South end of room (near window)

Both paddles were set to use an HV supply of 1,200 V (last I checked).

The start and end times on the run (current cut) are:

• Start:  Mon Oct 31 18:00:18 2011 + 835 ms: < DAC_LEVELS,-0.200,-2.500,-0.299,-0.447,-0.669,-1.000 (logarithmic ramp)
• End:  Tue Nov 01 14:40:51 2011 + 166 ms

Total length of run:  20 hrs, 40 min., 33 secs. = 1,240.5 mins. (approx.)

Number of pulses:  4,136
Number of pulse pairs:  2,068 (assuming all pulses were in pairs instead of larger clusters)

Rate of showers:  1.67 showers/min.  Tweeted about it.

Histogram of pulse heights:

• 1 threshold (200-300 mV) => 954 pulses
• 2 thresholds (300-450 mV)   => 2,244 pulses
• 3 thresholds (450-670 mV)   => 721 pulses
• 4 thresholds (670-1000 mV)   => 216 pulses
• 5 thresholds (>1000 mV) => 1 pulse - but time values were scrambled; probably have to disregard

Anyway, here's a plot of that data:

Rough histogram of pulse heights in last night's 20-hour run.

And (after running anal-pulses.sce on the data file), here is the histogram of the absolute time differences:

Histogram of time differences between the two detectors on last night's dataset.

For absolute time differences of 40 ns and above, the count of coincidences levels off, to the range 20-22 per 5 ns wide interval.  This is presumably the background rate of coincidences due to random happenstance.  For time differences of 35 ns and below, the rate is substantially higher, indicating that these are "genuine" concidences due to passing showers.  The detectors are only about 22 ft apart, which makes the existence of genuine coincidences with time differences at and above 25 ns somewhat mysterious.  Either the shower fronts are moving significantly slower than light, or the pulse amplitudes or rise times at the two detectors are substantially different, or some other, unidentified source of error is smearing out the measured arrival times of the pulses.  If the shower density is low, the amplitudes could be different due to shot noise in the number of shower particles intersecting the paddle area.  Low particle-number fluctuations could also more directly affect the arrival time of the pulse, since the shower front pancake has nonzero thickness; one detector might see a particle near the leading edge of the pancake, the other one a particle near the trailing edge.

It would be interesting to see if these anomalous time differences persist even after we are doing pulse shape reconstruction with its more precise identification of the pulse peak time.

Tomorrow Ray wants to do another histogram test on the scope with the overlapping orthogonal paddles, but this time with the paddles intersecting closer to the middle of the paddles rather than at their far ends.  This will maybe help us distinguish any effects due to differing light transmission properties of the two scintillator materials.