Detchar
Conclusion: Now with three data points, when PRM M2 coils get a mean value close to 2^16 (it has some RMS too, not shown) we see low frequency glitches in DARM. Second conclusion, today's lockloss at 4:45 UTC was likely caused by PRM M2 hitting the DAC limit of 2^17. So the tidal offset that was tried before should fix that at least.
Dan Hoak on the call pointed us to that there were two more times with PRCL/SRCL glitches seen in hveto. I used these three times to make the attached plot as an update to previous studies. Back then we were not able to make a smoking gun correlation between 2^16 crossings in PRM and these glitches, whcih is unsettling. We will look at these new times to see if it's more convincing.