Not Just a Coincidence

Analyzing a 91 KB data file from overnight run, started 9/6 (2011) at 5:04:03 pm, snapshotted for data analysis on 9/7 at at 3:18:13 pm; that's about a 22-hour run (22 hours, 14 minutes, a.k.a. 1,334 minutes).

SciLab is slow as a dead dog; I am looking forward to the day when we will be processing the data in real time in our Python server.  Maybe the students can work on that.

According to Scilab analysis script (anal-pulses.sce), over this period, there were 337 coincidences, for an average rate of one every 4 minutes.

We might conceivably see a still higher rate if we turned up the sensitivity, and/or figured out why the one paddle is responding so much more weakly than the other, and fixed that.  Perhaps its scintillator is less polished, or its optical glue is more bubbly.

Let's take a look at some of the pulses, to make sure everything is copacetic:

Coincidence #1:

Coincidence #1 with delta = 0 ns:

    3.679D+10    351.    2.    1.    0.    5.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0. 
    3.679D+10    2730.    1.    1.    0.    8.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0. 

Raw data:

Tue Sep 06 17:07:07 2011 + 237 ms: < PULSE,2,351,36791996524,1,(0,5)
Tue Sep 06 17:07:07 2011 + 237 ms: < PULSE,1,2730,36791996524,1,(0,8)

Both pulses crossed just 1 threshold, but the PMT1 (window) pulse was ~15 ns longer.  The time difference was zero, so the pulse came from close to the east-west-zenith plane.

Coincidence #2:



Coincidence #2 with delta = 10 ns:

    1.398D+11    1410.    2.    2.    0.    2.    1.    4.    0.    0.    0.    0.    0.    0.    0.    0.    0. 
    1.398D+11    10397.    1.    1.    0.    6.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0. 

Raw data:

Tue Sep 06 17:15:42 2011 + 163 ms: < PULSE,2,1410,139780145512,2,(0,(2,1),4)
Tue Sep 06 17:15:42 2011 + 162 ms: < PULSE,1,10397,139780145514,1,(0,6)

This shower hit the North (PMT2) detector ~10 ns before the South one.  Currently, distance between them is about 22', so this shower came from about acos(10/22) = 63 deg away from the N-S axis.

The pulse from the North detector crosses 2 thresholds (-0.5V) and is total width 7 time steps (~35 ns):
   *******
     *

And the one from the South is width 6 time steps (~30 ns), but only crosses 1 threshold (-0.35V).

Let's skip ahead to one of the larger pulses:

Coincidence #24:

Coincidence #24 with delta = 5 ns:

    8.166D+11    59690.    1.    4.    0.    1.    1.    1.    6.    3.    2.    8.    0.    0.    0.    0.    0. 
    8.166D+11    8182.    2.    2.    0.    1.    9.    3.    0.    0.    0.    0.    0.    0.    0.    0.    0. 

Tue Sep 06 18:12:05 2011 + 940 ms: < PULSE,1,59690,816571367255,4,(0,(1,(1,(1,6),3),2),8)
Tue Sep 06 18:12:05 2011 + 941 ms: < PULSE,2,8182,816571367256,2,(0,(1,9),3)

This one hit the South paddle first (slightly).  The pulse profile of the larger (PMT1) pulse was:

**********************
 *************
  **********
   ******

With the present DAC settings, that one is over 800 mV amplitude!  The PMT2 pulse was smaller:

*************
 *********

This is consistent with our observation that the pulse rate on PMT2 is lower.  In general the gain on that paddle is lower, as we determined yesterday.

Ray says the rate of showers we are seeing now (1 every 4 minutes) is reasonable.  We're going to keep the current run going one more night (to look for diurnal variation).  However, in the data so far, there doesn't seem to be any diurnal variation per se, but just a slow settling down of the PMT1 pulse rate near the start of the run:

Pulse rates for paddle #1 (middle), #2 (bottom), and their sum (top).  Horizontal axis is minutes in run, vertical axis is pulses in each minute-wide bin.  Note there is a slow settling down of the pulse rate for PMT#2, which was first put into service shortly before the start of the run.  The bias voltage for both PMTs is 1.2 kV, and the detection threshold for these (negative) pulses was -350 mV.

That's a good stopping point for today.