Tonight I spent some time on ASC during full lock.

In addition to the new and improved dETM pitch loop (LHO#17006), and the existing dETM yaw loop, I was able to close high bandwidth loops for ASB36Q → BS using FM2, FM3, and FM10 in MICH_P and MICH_Y. The gain is 0.1 for MICH_P and 0.3 for MICH_Y. [FM10 is a BS plant inversion which seems to still work pretty well. FM2+FM3 make the loop 1/f everywhere.] The BS yaw loop UGF was >1 Hz (when the gain was turned up to 0.7), and I suspect the pitch loop was similar.

With these high-bandwidth BS loops closed, I then requested the BS SEI guardian to take the ISI to FULLY_ISOLATED. We saw yesterday (LHO#17007) that turning on stage 2 of the BS ISI causes a transient in the BS yaw, thereby breaking DRMI lock. But with a high-bandwidth BS yaw loop closed tonight, the interferometer was able ride out this transient, with no excursion visibile in the BS oplev signals.

Then I turned off the BS oplevs (around 2015-03-02 8:45:30 UTC). This improved the DARM spectrum slightly from 8 Hz to 30 Hz or so.

After that, I moved ETMX in pitch and yaw until the cETM error signals were near their zero crossings. Then I closed the cETM loops as described in LHO#16931, but with gains of -1 for pitch and -0.2 for yaw. The lock broke before I got a chance to estimate the bandwidths. Adjusting cETM (and then closing the loops) brought the interferometer visibility up to 95%. However, the recycling gain remained slightly low, at 26 W/W. So there must be yet more loops to close in order to increase the recycling gain above 30 W/W.