In the afternoon shift we started by trying to continue the recovery of the lock acquisition; given that at the beginning of the shift the situation was worse than in the morning (with no apparent modifications) we stopped and we focused mostly into understanding what our issues are:
- the biggest one we have is the additional noise we have when we start to reduce the CARM offset before the CARM_MC handoff; these bursts are around 75 Hz in its main component (Figure 1) and the louder ones are also seen by the RFC error signal, which is still in loop at this stage; however, they are not present in the initial stages of the lock acquisition (the other increased noise in LOCKED_ARMS_IR looks different and today was barely present); the INJ team did several checks (to be reported in a separate entry, in Figure 2 the result of actions aimed to reduce the noise in the 6 kHz region) but the situation did not change; in order to mimic (poorly) the situation we locked the PRWI configuration, in order to have PR aligned and some signals to look at (around 19:30 UTC); we could see increased noise on PSL_PMC_TRA_Audio; to be discussed whether checking the Faraday isolator on INJ can help or not;
- when looking at the RFC we noticed that its transmitted power is really low: a quick look at a few figures of merit (Figure 3) shows that we lost some power, with possible causes being a slight misalignment of the PMC and a slightly different IMC power due to finite precision of the input power tuning routine; IMC, PMC and arms are somewhat coherent, but the RFC is clearly lower than it should be; we thought this could be related to some change in temperature, maybe due to a bad position of the PR mirror when (mis)aligned before the recovery of the DRMI (Figure 4); indeed the change in temperature is there, and it is big (although possibly converging now) but the correlation with PR is not strong; a wider picture in Figure 5 shows also the QPDs of the RFC changing a lot their readings and it also shows the clear difference with respect to before the shutdown;
- one other thing that we noticed for the RFC, in addition to the long-term trend, is the presence of quite quick rises and falls of the transmitted power on a shorter time scale; these were linked to the engagement of the dithering control of the arms when locked in standalone; in Figure 6 the comparison with the night of the 1st of November (disclaimer: the values of 1st of November have a fake Y offset for visualization purposes): the quiet periods in IB_BENCH_TY is where there is no dithering engaged, while when the dithering is engaged the effect is visible, with an initial behaviour which looks like a spike; this already happened before, but it may induce to think that the one minute we wait with the dithering on may not be always enough (especially when coming from different alignment strategies); the more important point is though that now this behaviour also induces a change in RFC_TRA_DC, which did not happen before (top plot, purple trace); maybe there is something not well centered for the RFC?
- as shown in Figure 7 (same disclaimer as Figure 6), they also happen on longer time scales, with the arms kept locked; another thing we noticed here is that the natural drift of the MC Z position (locked on the RFC) is more accentuated now than in the past;
- during the evening the noise issue of CARM_OFFSET_MID became less impacting (but the DRMI harder to lock), so we refocused a bit back on the lock acquisition: we understood that the unlocks at DARM_IR (after the handoff, 40 mW of arms power) were caused by acting on the ALS_CEB Flip Mirror in the Atrium; this was suspicious as the unlock happened always after a fixed amount of time after the ramp completion, which is in the end 1 Metatron cycle + 1 second for the cm command (Figure 8); looking more into it (Figure 9), it looks like that such Flip Mirror has an effect on the PSTAB loop, maybe induced by EOM_CORR (which is closer, hardware-wise);
- being able then to move on, we increased MICH_GAIN at STEP1 overcoming a low gain oscillation visible on the sidebands power;
- we reached again CARM_NULL_3F, and the jump (more a drift, actually) was quite slow, and we unlocked due to the DIFFp engagement; a change of pi of the demodulation phases fixed the issue;
- in the following lock we saw an oscillation rising on most signals; we pinpointed that to the engagement of the input beam alignment control, and opening it made it stop; we checked the phase, which was off by only 0.24 rad; it should be checked if the gain is wrong or something else is going on;
- at the end of the shift we left the arms locked on the IR.
Minor/other points:
- we saw another couple of missing data events from Sc_NI;
- the position loop of SNEB opened once, in correlation to a bad lock on the beating signal, but it was recovered by just closing it again;
- we turned off again PSTAB0_line (43 Hz); to be saved in the configuration file if we want it as default;
- the input beam alignment control (ASC_PR_X/Y) has been left disabled in CARM_NULL_3F, to be checked manually;
- we reduced the threshold for the jump from -0.2e-3 to -0.15e-3.
