STEP seems to see a lot of self-triggered events with dt < 20µs
after events deep in the landau tail. One proposal is to raise
min dtime threshold in L2 up from 0. This may starve the daq at very
high trigger rates but gives reliable PHA for those events that pass
the L2 trigger.
The other option is to set the I-DefX delayed startup bit, which delays
the reconnect of the shaper by 32µs after a readout. That cannot
starve the daq, but may result in pileup and or balistic deficit errors
at high rate.
This patch adds infra to set the min_dt cut in step.ppss and
step.base_config and changes the default for all units to 20µs.
55624a99af