Sheila, Pep, Jenne Georgia Ed

We were able to increase the input power to 37W requested from the PSL, without any big problems.  We have also reverted the change to the AS A 36 phases that caused some extra noise last night.  We have not yet retuned the LSC feedforward for the new input power, or run calibration sweeps. 

Details:

• reverted ASA36 phasing change from yesterday, which is used for BS angular control.  We saw that POP18 was improved, PR gain became more stable (measured by POP/IM4), excess noise below 40Hz gets better.  Phasing settled at 16:26:30 UTC June 5th
• We saw that AS72B responded with a good SNR and went closer to 0 when the BS alignment changed.  We phased individual quadrants of AS72B (not in loop) to zero the signal in the Q phase here, see screenshot attached. We put offsets in the MICH P+Y loops to estimate that if we want to switch to AS72BQ for MICH we would want an input matrix of 5000 for pitch and 800 for yaw.  We saw that the high frequency sensor noise in AS72 is much larger than for 36, so that when we attempted to switch over there was large noise in DARM from AS72.  We tried to engage one more stage of whitening filter, but the noise in 72B isn't ADC noise.  We may want to think about either reducing the bandwidth of the MICH ASC loops, cutting them off more aggresively, or blending to try to use 72 for MICH control. 
• AC coupled the ISS using the guardian, worked well!
• measured PRCL OLG, found ugf of 35Hz (with gain setting of 16), increased gain to 25 to get ugf of 50Hz.
• Powered up in 1W steps to 37.4W from the PSL.  Georgia noticed that the ADS YAW3 loop was running away, so we opened it and adjusted PRM yaw manually waiting for ADS Y4+Y5 to converge.  We checked that the YAW3 demod phase looks good by changing the Y2L gain on ITMY, and checked that the zero crossing of the error signal looked good by moving PRM by hand. 
• We increased both ADS YAW4+YAW5 gains by 30dB to do this, why don't we normally operate with higher gains?  After doing these checks we closed YAW3 again and it seemed to work fine, maybe the increased gains for YAW4+5 have helped avoid some very slow cross coupling problem.
• We tried to power up to 38W requested, which gave us 38.5W from the PSL and 33.8W at IM4, and saw that we had some glitches in the PR gain and things didn't look very stable.  For now we have decided to stay with 37W input power requested, which is giving us 37.5W from the PSL, 33W at IM4, 184kW in the Y arm according to the TMS monitor and 170kW in the x arm according to our monitor. Up to this ooint, our power increase today gave us a 6% increase in circulating power, but kappaC dropped from 0.97 to 0.96 while we increased the power. 
• Jenne is trying to move the spots around to see if we can find better recycling gain for 33W input power (measured at IM4). Lost lock at 18:55 UTC without any sign of trouble: 1243797554
• Relocking:
  • PRMI alignment was way off, Ed moved IM4 to recover the build ups.  We used the DRMI ASC with the 36A phasing that was good in full lock, it seemed OK so we are leaving it that way (121degrees). 
  • One lockloss when I tried to increase the ADS DOF3 gains at engage ASC, they need to be low until the other ASC loops are at full gain.  Also made some changes so that the ADS loops would be managed in a more uniform way (ADS 4+5 have same filter modules in all states now, which increases the YAW4 gain by 30dB in full lock and the yaw5 gain by 20dB in full lock).
  • Due to a bug in the maxiumu power state we requested 40W input power when we were intending to request 37W. We did all of the low noise steps but skipped 9MHz modulation depth reduction.  ADS YAW3 did not hold it's error ponit to 0, and we lost lock with what looked like an ASC instabliity: 1243803997
  • Next lock we re-wrote MAXIMUM_POWER state to make the request what we set it to in lscaparams, this worked fine, we reached nominal low noise with 37W out of PSL, 33W at IM4 around 21:30UTC
• We saw that the ASB72 signal responded to lowering the 9MHz modulation depth, and hypothesised that what is really happening is that the ASA36Q signal used for BS control is responding.  Jenne put an offset of -400 counts into the MICH pitch input to bring the 72 signal to 0, this temporarily improved the POP18 build up but the offset was too large after the thermalization (AS72 overshot) and we removed it.  We can try more exploration of BS ASC offsets at some point.
• Jenne moved spots again, with ADS open, and found a better position for PRM Y (ITMY Y2L gain adjustment from 0 to 0.1). 
• We checked that the phasing of the REFL WFS is OK with a frequency excitation.  We also measured the sensing matrix for INP1, PRC2, and CHARD.
• Lockloss 23:20UTC, winds reaching almost 40mph.  We haven't been able to relock mostly because of the wind since then, although it is starting to calm down now. 

We are going to leave the input power request set to 37W, if there are any problems with this it can be undone by going to userapps/h1/guardian/lscparams.py and changing NLN power to 35 on line 17 and hit load on ISC_LOCK, and on LASER_POWER.