We had several locklosses early in the CARM offset reduction sequence tonight. In the CARM_150_PM state we servo the ASC_AS_A_RF45_Q_SUM_NORM to zero using the DIFF_PLL_OFFSET. In the past Sheila and I have lowered the gain on that servo (set in guardian) as it seemed to be injecting noise and causing locklosses.

I had a look at these channels and it seems like turning the servo on is not helping to stabile AS_A_RF45. Something in this process needs retuning for these windy conditions (see attached screenshot).

The gain is nominally 0.004 (ISC_LOCK guardian line 1587), I tried decreasing it, and while we did survive a little longer we still lost lock shortly afterwards. I tried increasing it, and we did make it through the state, so I am leaving it at 0.01.

Do we actually need this servo? If this state continues to cause locking issues perhaps future operators could just comment out line 1598 of ISC_LOCK and see if just not turning it on helps.