Since LLO was briefly out of lock, and the DetChar safety injections are way overdue, and those don't need to be done in coincidence with LLO, I ran them.

I used the instructions in 69723, and ran hwinj detchar safety --run.  I specifically did not set a gps time, so that it would just go at 10 sec after I hit Enter, and it automatically avoided beginning on an integer second.  Sidd notes that there is a Google doc with instructions, although the script should take care of most of it, including the output switch.  These filter modules and the gain that are noted in the first section of the document are the same that are needed for the stochastic injections, so should already be SDFed on at both sites after last Friday.

The first time I ran this, initializing took a few seconds longer than I think the developers anticipated, so I got an error about 'Block time is already past'.  I just re-ran it, and it initialized faster the second time, and worked as expected.  Git issue started here.

Once the injections were complete, RyanC noticed that PCALx (used for these injections) had its OFS servo saturating.  RyanC toggled it on and off, and it is happy again.  I wonder if the amplitudes were tuned at a time when the now-always-on PCALx lines were not yet on.  It sounds like the amplitude tuning happened at LLO and at that time they were well away from the saturation limits, so it could also be an issue of we've got a sligthly different set of things going on with our PCALx. Git issue started here.