Terra, Sheila

After recovery tonight we did a few things:

• Reduced the gain of the PRC2 loops, by a factor of 5 each. (PRC2 P gain from -5000 to -1000; PRC2 Y gain from -10000 to -2000) because we saw an instability at around 1.5 Hz that could be reduced with the lower gain.  These loops have had higher gain than normal the last few weeks (the story is here)
• Reverted the TMS matrix to the one that is insensitive to transmon motion.  The thinking was that TMS drift is the reason we are not fining repeatable offsets, and that by reverting at least things would be more repeatable from day to day.  The old offsets worked OK, giving us a recycling gain around 34 at 2 Watts and 24 at 50 Watts, although at 50 Watts we saw the CSOFT instability slowly growing.  We also see that at 40 Watts we can recover a recycling gain of 32 by using our old pop A offsets (0.4 pit, 0.2 yaw)
• We modified the a2l script so that we can run it while the SRM dither loop is on, and ran it.  (for now I saved the modified scripts with _LHO at the end of the name.)
• We have continued to have trouble with the CSOFT instability and tried to improve things by turning up the gain of the ISS 3rd loop by a factor of 10.  However, the third loop seems to have no effect on the instability tonight, although in the past it has clearly damped it. We weren't able to measure the open loop gain because each time we tried to measure we broke the lock, even with a fairly small excitation.
• We powered up with the dither lines on, and saw no evidence for spot motion on the test masses (consistent with what was seen in the past). 
• In our last two locks we powered up with dither lines on the PR optics and beamsplitter, and can see clear changes in the spot positions as we increase the power (power ups starting at 8:56:52 and 9:50:30 UTC), and we can see that that when we adjust offsets to improve the power recycling gain we are partially resotring the spots to their orgiginal positions before the power up.  We will work on calibrating the data, the attached screenshot shows uncalibrated spot positions for PRM in maroon, PR2 in orange, PR3 in red, and BS in taupe.

We have noticed that since we are going to higher powers, things are misalined after a lockloss and become realigned as they cool off.  We might want to account for this someday to save time.