I spent some time this morning trying to get the full IFO to match the beam spot from Saturday (alog 47101) that seemed to be mostly away from the big point absorber on ITMY.  This new candidate position is about 9.7 mmm low-of-center on ITMY, whereas the previous nominal position was 9.9 mm high-of-center.

In the attached first figure, we are locked at 2W DC readout starting at about 5000 seconds.  A little after 6000 seconds, I start moving ITMY's spot position by changing its A2L gain.  You can see that the POP18 buildup went from its usual value of 60-ish counts up to 70-ish.  However, the power recycling gain and the buildups in the arm cavities had decreased.  I moved the ETM spots by changing their A2L gains, and I also moved PR3 to try to get the PRC axis nicely aligned with my best arm axes.  This brought the carrier buildups back to about where they had started, and also increased the POP18 a little bit more. 

Around 22,000 seconds I started increasing the laser power, first to 10 W, then in steps of 5 W. The POP18 buildup didn't seem to go down quite as fast as it has in the past, which would be consistent with having avoided the ITMY point absorber.  However, the carrier buildups went down much faster than usual.  Afterward, Dan found that it looks like I accidentally put the ETMX beam spot right on top of its point absorber.  Since I was doing my alignment at 2W, the absorber on ETMX wasn't bothering me, so I didn't realize that I should avoid it.  This is likely the reason that the carrier buildups went down so drastically.  

In the afternoon, Dan and I decided to try the ITMY spot that is even higher of center than the point absorber, since this location had given us better carrier buildups on Saturday.  Note that in the second attached figure, we started moving spots while already locked at 30W, so the POP18 and PR gain look lower than in the first figure which was at 2W.  In this second figure you can see that I started moving ITMY's A2L gains at around -5500 seconds.  Before that time, the PR gain was less than 45, the POP18 buildup was less than 52, and the Yarm's circulating power was around 149 kW.  After the spot moves, I had the power recycling gain above 46, the POP18 buildup about 55.5, and the Yarm circulating power just barely at 154 kW.  So, that's an extra 5 kW circulating in the arms.  We seemed really stable for most of this moving, but at the very end something started ringing up in ASC, mostly Yaw although also Pitch, and I couldn't recover, despite trying to move some things back a few steps (ETMY Y2L gain and PR3 yaw slider).  The oscillation was at the 0.49 Hz frequency that we often associate with dP/dTheta.  I think the oscillation itself was mostly under control, but the noise in DARM got high enough that it was drowning out the ADS signals, so the arms and PRC started moving around a lot.

To relock: We wanted to keep these A2L gains, so we moved PR3 back to it's original position according to sliders (about -119.8 pit and 149.8 yaw), then locked DRMI, and got to ENGAGE_ASC_FOR_FULL_IFO.  During this state, I moved PR3 back to where we had it at high power last lock.  The ADS loops moved the rest of the degrees of freedom to put the spots where we had them.  So far, we were able to power back up to 30W with this situation. POP18 has stayed fairly flat during the power up and after 10 minutes of thermalization (yay!), and the carrier recycling gain and arm buildups are increasing as we thermalize (also yay!).  The POPX PZT is unhappy, but seems to be doing better at 30W as our loops converge.  I changed the checker in ENGAGE_ASC_FOR_FULL_IFO to let you pass with a notification if the centering of the spot isn't too bad.  Formerly, it would just Return False if you were at all mis-centered. 

If we decide we like these positions (I think we will), then we're going to have to move the green initial alignment setpoints, the ALS beatnote paths on ISCT1, and perhaps also relieve the POPX PZT.

Other locking notes:

• Lost lock once at DRMI ASC with the 0.5 Hz ringup in MICH ASC.  2nd time TJ had improved the alignment before allowing ASC to come on.  That, combined with increasing the MICH ASC gain from -0.25 to -0.4 seemed to help.  Once we had switched sensors to AS36 from AS45, the nominal gain of -0.25 was fine, and the oscillation was gone.  So, seems to be a problem only when using AS45.
  • Next lock (several hours later) this didn't seem to be a problem.
  • Perhaps the ring-up is more likely when the alignment is not so great?  We just had another lock with the DRMI ASC successful on its own.