Paddling Upstream

When I left on Thursday I did a Quartus compile to pull in the latest firmware; loading that image.  It sets the DACs at -300 mV.

Hooked up the 2nd paddle (at PMT1); it worked, but I wasn't getting any pulses from the 1st paddle (at PMT2)!  Left it sitting a while, and it eventually started working.  Weird.  Perhaps the FPGA had to cool down first in order for the PMT2 datapath to start working reliably?  That is the one that uses logic registers instead of RAM blocks in its FIFO; perhaps it's a little slower for that reason...

Interestingly, the pulse rate is higher for PMT1 (the paddle near the hallway).  I would have expected a higher pulse rate for from one closer to the window!  The supply voltage is slightly (like 10 mV) higher for the one near the hallway, but that does not seem like enough to account for the difference.

I think I'll try swapping the PMT bases and see if that changes things.  Stopping the current run.  PMT1 (with base #1) had 46392 pulses, PMT2 (with base #2) had 13220, for a ratio of 3.5:1.

OK, swapped the bases; now retrying.  PMT1 input (ICDP channel #0) still has the higher pulse rate - stopped it after running for a while and the ratio is 7,513:1,991 = 3.77:1 (even worse!).

Next, I'll try swapping the inputs, in case there is some difference in the way the two different pulse-capture datapaths are responding.  No, after swapping the inputs, input PMT2 now has the higher pulse rate.  Let's stop the run for a moment to check the ratio - 2,463:556 = 4.43:1 (even worse!).

Let's now try swapping the paddle locations (without changing which cable is going to which side of the room).  That restores input PMT1 to the higher pulse rate; ratio is 1,885:581 = 3.24:1 (back down below what we had originally).  This means it's the scintillator/PMT assembly that's causing (the bulk of) the difference in response that we're seeing.  The paddle assembly that's near the window now (input PMT1) is the one that's responding more strongly.  We just need to remember this, and maybe turn down the base voltage on that one somewhat to compensate (or turn up the base voltage on the other one).

OK, I'm starting a new overnight run to check for coincidences, so we can figure out what the shower pulses will look like with the paddles.  If they are very large compared to the present voltage ladder, then we may want to increase the spacing between the voltage steps somewhat.  It might make sense to also increase the base voltage of the ladder to cut down on the piddly little noise pulses.

Using the ladder:  Base -350, step -150.

Ray suggests tomorrow also trying a test with the paddles on top of each other.  This should give us a much higher rate of coincidences.