[Jenne, Hang, Craig, Gabriele]

Summary

We can improve the recycling gain and simultaneously improve the beam centering on ITMs by moving PR3 in yaw.

We didn't get all the way to centered mirror, but we gained 3% more power.

Details

Motivated by the observation that the beam is mis-center by 2cm in yaw in the ITMs, we tried again to move it, continuing yesterday's attempt.

The strategy was different this time: we locked up to PREP_DC_READOUT_TRANSITION, with all ASC loop engaged and converged. We then moved PR3 in yaw by small steps. The idea here is that the PRM angular control will keep the beam spot in the same position on PR2, the soft loops will maintain the beams fixed on the ETMs, and the REFL WFS loops will ensure that the beam axis is well aligned into the IFO. The only free degree of freedom is then the beam spot position on the ITMs.

We dithered ITMX and ITMY in yaw at 21.3 and 21.9 Hz, and we were using H1:ASC-ADS_YAW2_DEMOD_I_OUT and H1:ASC-ADS_YAW3_DEMOD_I_OUT to monitor the beam centering.

By moving PR3 in yaw we could increase the recycling gain, increase the power in the arms and reduce the mis-centering. We lost lock a couple of times during the move, likely due to loop instabilities triggered by the larger circulating power.
We had to change the MICH gain from 2.8 to 2.0, SRCL from -25 to -15, and change the PRCL to SRCL sensing matrix from -0.045 to -0.02 to get rid of a weird shape in the SRLC open loop transfer function.

At this point we started losing sideband recycling gain (mostly RF18), so we tried to move the PRM set point in yaw (we could not improve RF18, but we lost most of the carrier power gain). We also tried to move MICH, but lost lock.

Unfortunately the new PR3 position is not compatible with the initial green lock, so we have to revert to the old positions every time we lose lock.