We had at least one lockloss tonight that seemed to be caused by SR2 glitches. Screenshot is attached.
It seems like SR2 was noisier after maintence day today.
But our main problem tonight was that the refl WFS (which we have taken in and out several times over the last week) don't work when the microseism is around the 90th percentile or more.
For such high microseism situations, it may be worth switching back to the 45 mHz blends for all the test-mass chambers, with or without sensor correction.
We are leaving the IFO in observe, but there is scattering from ISCT1 because we can't move to REFL wfs with the higher ground motion. This is something to look at in the morning, but it caused us many locklosses tonight.
The IMC WFS offsets that were working early last week don't seem repeatable tonight.
Nutsinee switched the test mass ISIs to BLEND_45_SC_useism, that went OK except for some large glitches when switch ETMX, but when I tried switching to REFL WFS again we dropped the lock. We probably need to look at finding a better matrix for refl WFS and getting rid of the oscialltion in the centering and IMC WFS.
Speculation about REFL WFS. Yesterday I was concerned that PR2 had moved what seemed like a lot (8-10urad) when I did the Inital Alignment. I would expect something smaller like 2urad. I couldn't find a cause for that, but I did notice IM4 also changed by 20urad in yaw. A change of 20urad in yaw of IM4, assuming PRM moved to match the change of the PRC input beam, would propigate to the REFL WFS, and my estimate for the change in beam position at REFL WFS to be 168um. I don't know the beam size on REFL WFS, but the IMC WFS is about 250um, so if REFL WFS is about the same, the 168um is a beam position change of ~70%.