This morning we continued the recovery of the lock acquisition after reaching CARM_NULL_3F yesterday night.
As a first step, we rechecked the calibration between B1p_PD1 and PD2, and we put the correct value (+0.071). Moreover, we also updated the jump trigger to a lower value (-0.00038).
After that, we could reach CARM_NULL_1F. At this point, Michal locked the OMC and started to work on its alignment. We reached again LN2 at 10.25 UTC.
During the first ~2h of this lock, Michal continued the work on the OMC alignment. After this work, at 12.15UTC we started to work on the CMRF optimisation by moving the setpoint of BS_TY. We reached the minimum of the CMRF with a quite large offset on the BS_TY (-0.0013) with ~25Mpc. In these working conditions, we manually tuned the gain of SRCL in order to put its UGF @5Hz and we switched off all the DRMI lines (see trends in fig. 1).
We then injected noise on DARM in order to have an estimation of the DCP with the minimum of CMRF:
- 13.14.10UTC for 5'; "DARM_noise_broad" AMPL: 20 (to be noted: during the injection, we disabled SRCL_SET servo)
First estimation of the DCP gives ~395Hz.
Then, we went back with the BS_TY offset to have a comparison of the DCP without BS offset, but unfortunately, we unlocked at 13.26.UTC. Looking at the data (see fig. 2) we notice that there is a strange sudden saturation of LSC_SRCL_CORR exactly at the second 36. To be checked if it could be a problem related to data losses or similar.
In the following lock acquisition tries, we reached CARM null 3f where we unlocked twice for DARM low gain, and once at CARM null 1f for SDB1 drifting away. The last unlock could be related to a problem on the action of reload config in LSC_ACL (see comment below).
To be noted: since we are now saturating the correction of the MICH_SET servo we increased the clipping thresholds in ACL from +/-10 to +/-20.
