Corey, Ibrahim, Georgia, Craig, Matt, Caroline, Julia

Corey locked the interferometer after the earthquake.
We were pretty much able to go directly to CHECK_VIOLINS_BEFORE_POWERUP without touching anything.  
Corey did tidy up the 2f signals by adjusting PRM and SRM alignment after DRMI locked, but the other guardian states "just worked" after our changes from last week.

We went up to POWER_10W, no problems there.
Then we started stepping through POWER_25W.
We ended up losing lock after upping the ADS gain from 1 to 2.

It was a very slow lockloss, which seems consistent with this story about it being ADS-driven.
We also had just turned on the SOFT Y loops to high gain (20 for CSOFT Y, 30 for DSOFT Y).
Georgia looked at the ADS control signals and it appears to be the PIT loops which killed the lock.

The above lock was with the ITMX SPOT GAIN at it's new "CENTER" value of -1.25, rather than 0.85.
We figured that since we MOVE_SPOTS soon after anyways, we can focus our efforts on trying to get up to high power with ASC relatively stable.

EDIT: We lost lock after going up to POWER_25W from what was probably that same very slow 0.02 Hz oscillation.  See picture 2.
Leaving the ADS pitch loops on at only a gain of 1 was not good enough.

Georgia added back in +10 dB to both PIT3 and YAW3 for the ADS which was commented out a few days ago while we were sitting at 2W CHECK_VIOLINS_BEFORE_POWERUP.  
This caused a lockloss when going up to POWER_10W.
This indicates it's just ADS PIT3 which is the major problem here.

EDIT2: We are sitting here at 20 W of input power, measuring CHARD P for an hour.  
We turned off ADS PIT3 while going up in power, with the plan of trying to ease it on at lower gain. 
While we were sitting here measuring, the 0.02 Hz oscillation came back.
Georgia stopped it by lowering the PIT4 and PIT5 gains from 1 to 0.1.

EDIT3: We tried to MOVE_SPOTS, but lost lock after about 8 minutes due to a violin mode damping typo.  
Georgia lowered the ADS gains to 0.1 for all, including PIT3 starting at the weird spot position on IX.
The buildups were getting very choppy, but we were surviving.  Unclear how far we could have made it.
Julia was damping violin modes and accidentally typed in -20000 into a filter module gain, which seems to have immediately saturated the IX PUM stage and killed the lock.
This is Julia's first lockloss so we will be nice to her this time.