Today we tested the effect on DARM of additional resonant gains at 2.8 and 3.4 Hz, as described in 73698.
DARM RMS was reduced as expected from 1.6e-9 to 1.2e-9.
Looking at DARM spectrograms, there is no evident difference in the level of non-stationary noise.
Comapring the DARM PSD during two times, there seems to be a small improvement, but that might be a fluctuation.
In summary: the resonant gains improve the DARM RMS, but we don't see a clear improvement in noise; they don't hurt though
I've created a low frequency boost that generally gives more gain at low freq, rather than just having the resGs. I turned it on starting at 23:11:00 UTC. Until 23:19:38, the squeezers were working. Then around 23:26:00 I changed the CARM gain. So, maybe this isn't such great 'quiet time' to evaluate whether we've improved our nonstationarity, but it certainly does seem fine, and it also does nicely reduce the low frequency DARM_IN1 and its RMS. The filter was back off around 23:31:00 UTC so we could go back to Obsering.
In the first plot, the blue filter is the one that Gabriele had created, and we tried early in the commissioning period today. Red is my new proposed filter.
In the second plot, the blue is DARM_IN1 under nominal circumstances (so, no resG or boost), and red is with my version of the boost filter. As noted in my 'timeline' above, there isn't really a lot of good quiet time to evaluate its effect on the nonstationarity, but the significant reduction in RMS seems like a good thing. If we want to leave this filter on, Louis will need to do some work to prepare the calibration pipeline, and we'll need to restart the GDS pipeline during the same commissioning window that we turn on the new boost.
EDIT: please disregard the y-axis label on the DARM_IN1 specra plot. I don't think that should be there.
