Daniel, Evan
The tidal offloading from IMC-F to the ETM UIMs contained suspension plant inversion all the way up to ~10 Hz, which seemed unnecessary since this should be a slow offloading. We replaced the inversions in FM2 of LSC-{X,Y}_ARM_DRIVE ("control") with a less aggressive inversion in FM1 ("ctrl2"). These filters now have 1 s ramps.
Tagging CAL: while even doesn't explicitly say this, this is COMMON arm tidal feedback to the UIM (as implied by the reference to IMC-F). See the attached figure that described the current LHO tidal topology, from T1400733. Even if this was DIFFERENTIAL arm tidal feedback (which [a] takes a low-frequency pickoff of the DARM control signal sent to the UIM from DARM_CTRL through the ETMX actuation distribution system {represented only by DARM |> ISC L |> ESD |> the L1 path in the below diagram} and [b] feeds it to HEPI), the calibration group doesn't model that pickoff either, and its cross-over frequency is "low" (where folks shrug their shoulders and say "probably around the microseism or lower;" this has been a traditionally poorly documented cross-over, since rarely do the folks who commission the offload filters have the patience to characterize the systems before or after their change). Tagging SUS: This should mean that there's much less longitudinal drive going to the UIM above 5 Hz. Presuming that there's large length-to-angle coupling on the UIM, I wonder if that means there's less angular motion about 5 Hz now? There should also be much less of the actuation range being used as well. Does that mean we'll have a different DAC noise level? Lots of fall-out questions to be answered from this change... Tagging Detchar: This should mean that there's substantially less low-frequency motion driving the UIM stage, which in turn (because the UIM actuators have coils on the reaction chain and magnets on the main chain) should mean a lot less *differential* longitudinal motion between the ETM's main chain and reaction chain (in the frequency region above 5 Hz where you see the new red filter is much less than the former blue). This may reduce the amount of scattered light left between the test mass and the annular reaction mass due to less velocity in frequency regions that can up-convert. It may also not, my intuition could be wrong. Regardless, this might be an interesting "before" vs. "after" study. Note, while we've recently re-tuned the reaction chain "compensation plate" alignment offsets on the *Input* test masses, there thus far still remains NO offsets on the ETM reaction chains.
Before (blue) versus after (red) in terms of DAC counts to ETM UIMs. The change is only seen on EY, since EX is dominated by the DARM drive.
Not immediately obvious to me why EX UIM should be pushing 10 times more microseism control than EY UIM. Just guessing from IMC-F, we aren't really offloading much of our microseismic CARM fluctuation to the UIMs, while DARM (which goes to EX only) is tightly controlled.
The bandwidth of the IMC-F offloading to the ETMs is less than 0.1 Hz -- the attachment shows a crossover measurement with IMC-F_OUT calibrated back to the excitation point.
Via step response, the bandwidth is of order 0.03 Hz.
