Figure 1_pad2: But... we should still be getting 10 counts per second from the T8 pulser (1024 Hz prescaled down to 10Hz) during this beam trip. If I cut on DBB.evtype==8 (should be fine, since it's not very likely that there are T2s or T6s in coincidence), we see the same gap. Where did the T8s go during this time.
Looks like there are periods of time 3-5 seconds (looking at previous
gaps) where the DAQ stops taking data. Maybe this is related to the
network saturation that has been observed (Figure 2).