I also took a look at the output widths on all of the Cerenkov related discriminators and set them to 100ns. Most had been set to 40, but there was one that was only 20ns. There is some high frequency ringing that might have been causing some multiple hits with the old shorter pulses.
PMT 12 is now showing the high-frequency (~5ns period) bipolar noise on
the pulse that was seen on PMT 2 yesterday. It just occurred to me that
maybe PMT 12 was confused with PMT 2 last night -- I foolishly didn't
check the analog signal of PMT 2 this afternoon. I did check the TDC
signal for PMT 2 and it wasn't too bad.