Here is a plot showing the PRCL gain thermalization (same as from alog 70211), but with information from Camilla's measurement today added in the small grey marker at about 155 mins.  It needs to be high relative to what we'd been using (pink line) since June 6th.  Recall that these PRCL locklosses we've been having are likely because our PRCL gain is too low.  So, I'll see if I can make another thermalization curve that increases faster.

I've got a new candidate thermalization, the yellow curve in this (becoming much too busy) plot.  This won't quite get the gain as high as it wants at the 2hr35min point, but it also won't overshoot the 4 hour target from 6 June 2023 by too much.  The thermalization guardian just holds the gain at the value at the end of this plot (it doesn't keep following the curve forever), so the final gain won't be very different with this new candidate (yellow) versus what we've been running with after 6 hours of lock in pink.

I'm going to implement this quickly before we get relocked.  This means using a GAIN_SCALE of 5.25 (has been 5.55 for pink trace) and a TIME_CONSTANT of 5571 (has been 7800 for blue, orange, pink, green traces). 

I modified and loaded the THERMALIZATION guardian, and checked it in to the svn, while we were relocking.  Hopefully this will reduce the number of PRCL locklosses we're seeing.