Sheila, Matt, Mayank

Because of yesterday's PR2 move, the beam was very off center on the ASC POP diodes.  I beleive that this was the cause of last night's locking difficulties 82678, rather than the arm alignment references, which should not have been affected (and weren't very different after being reset).  This QPD is used in DRMI ASC, and the yaw loop was pushing in the wrong direction so that the beam was falling off the QPD.  Turning off this loop and not PR2 can lead to some unusual alignment that might cause difficulty locking, which could be avoided by moving CHARD, even though the real problem was in the PRC.

We were dropped out of observing because of the extreme PI damping (more on that later), so we took the chance to pico on these QPDs. We used the poorly named ndscope template /sheila.dwyer/ndscope/ASC/Pico_pop_wfs.yaml (similar to 81849) 

We strategy that worked for us was to use motor 5 (ASC POP steering 1) to center QPD B and motor 6 (ASC POP steering 2) to center QPD A, but we had to walk very far to get both beams on the QPDs at the same time, we ended up moving by about 17000 counts on both picos in yaw (in the attached screenshot you can see wee picked the wrong strategy first).

We ended up with both QPDs centered around 0, and the sums are about 3% higher after our pico'ing than before.  We reset the offsets to 0, and accepted this in both safe.snap and observe.snap.  We have set PR1 +PR2 ASC to off in the DRMI guardian, because we need to set these offsets at 2W with all the ASC on. 

Request for next time we lock: DRMI may not be well aligned because the PRC1+2 loops are off, but if initial alignment gets run that should fix it.  It could probably also be fixed by running PRMI ASC, then locking DRMI.  Next time we lock, we can trend what the POP_A QPD pitch and yaw are in the guardian state PREP_DC_READOUT and set the offsets to -1 * this number. Then we should edit lines 1030 and 1031 in ISC_DRMI to turn back on PRC1 and PRC2 ASC loops.

Note for Detchar: Keita had pointed out that fixing this clipping on the QPDs could possibly fix our large glitches that looked like a scattered light problem, which was apparent in range fluctuations in this morning's lock.  We don't have much data yet since going back to observing, spectragram.