Sheila, Lisa Since we had lost good alignment references, and the sub-optimal IR alignment was what limited the build up in the last full lock attempts , we decided to invest sometime working on the initial alignment procedure. As already known, achieving a good alignment for both IR and green is easy for the X arm, but kind of tricky for the Y arm. So, after Sheila aligned both green and IR to the X arm, we decided to commission the IR lock to the Y arm (feeding back the REFL IR signal to the MC length) with the idea that this step will become part of a future automatic alignment procedure to ensure the IR is well aligned to Y. As observed for the X arm, because of the 27 Hz and 41 Hz MC2 resonances , achieving a stable IR lock feeding back to the MC is indeed painful. We eventually succeeded by re-designing the loop so that we could keep the 41 Hz notch in the MC path always on. Phase and gain margins are not great, but this worked well enough that we could keep the cavity stably locked while optimizing BS, ITMY and ETMY so as to reach maximum build-up for IR (1 in LSC-TRY). At that point we used TMSY and the green QPDs offsets to recover some green build up, and we could relock the cavity on green; with both cavities locked at the same time we could find a way to keep maximum build-up for IR, and achieve 0.7 for green. Not fantastic, but probably not too bad; Keita's green WFSs will hopefully make the green alignment better. Happy with the IR in the arms we went back to PRMI. The PRM alignment was totally off (about 15 urad both PIT and YAW); Sheila patiently brought it back, and we could relock the PRMI stably on 1f sidebands. We leave the IFO with the green arms locked for Keita tomorrow.
Lisa, Sheila
we used Refl 9 to lock the PSL directly to the Yarm this afternoon. We rougly followed the procedure in alog 9644.
We disengaged the notch for the MC2 roll mode (ISCINF FM6), but leave FM7 on to help avoid ringing up the vertical mode at 41 Hz.
We used Refl A 9 I, with the first two whitening filters on and a whilething gain of 27 dB. This gives us a gain of roughly 1 Hz/ count.
We send Refl9 I to the Y arm filter bank. We copied the filters that were used in alog 9644, but added an additional boost with a zero at 3 Hz . The boosts and the feedback are triggered on TRY. (FM3 nd 4 are engaged, FM1 and 2 are triggered.) We are curently using a gain of 0.15.