I've been making a few changes to the ALS_COMM node over the past couple of days to attempt to fix some of the ir finding issues we've been seeing. After pulling myself out of an awful conditional and lack of sleep mess I made, we're now in a good spot. Summary of changes:

1. When finding IR, the transmitted power would be just barely below threshold to recognize it and fine tune the signal. The signal without IR, dark, would often be off by about +/-0.2, making a found IR seem as though it wasn't there. I added a PREP_SEARCH state to move the VCO frequency away, take some dark time, and then reset the H1:LSC-TR_X_QPD_B_SUM_OFFSET. This seems to have helped with the consistency.
2. Finding IR in the fast search consists of moving the VCO to two known good frequencies. Sometimes it would go to a frequency and while there was clearly something visible on the output, it would be just below threshold. With the previous PREP_SEARCH step added, I was able to lower the thresholds a fair amount (0.22 to 0.8) and so far this has been good. This threshold might need to be adjusted, we'll see as we get more relocks.

ALS_DIFF's IR finding works a bit different, but I think some of this can be carried over. Working on that next.