ITF locked for all the shift; at 14:13 UTC, upon request of Fiodor, I performed another step for the etalon test:
2023-09-10 14h13m51 UTC berni 'NI:NI Set point [NI_RH_SET=20.366,rampTime=21600]' sent to LSC_Etalon_Acl
ITF left locked with autorelock engaged.
Software
At around 14:56 UTC the SL_TempController started to provide fake data; at 17:44 UTC I restarted the process and this fixed the issue.
As the last step with Etalon resulted in a worse sensitivity, I reverted it:
Figure 1 tries to summarize the etalon steps done during the weekend.
The solid lines show the start time of the ramp of the etalon step (which lasts 6 hours). The corresponding dashed line is when the highest slop of the temperature on the ring heater is roughly happening, this is delayed by ~10h for the first step, and ~15h for the second step. The third step occured shortly after (~30min) the maximum of the temperature slope of the second step occured. The Etalon correction is clipped at either zero or maximum (500V) most of the time, which probably explains why the delay is not the same. The Etalon correction has also been clipping at zero for half of the time before the weekend.
For the first two steps the CMRF has gotten worse with BS TY INPUT getting more negative when the maximum of the temperature slop was reached, so the effect is systematic. Changing the direction of the etalon set point, has started to revert this trend.
B4 56MHz mag becoming stable on Sunday morning for a few hours is due to SRCL SET saturating at 50, it has became smaller since then.
The etalon transient from the last step will continue all day today. So this need to be taken into account in any analysis of the data from Monday's shifts.
