Bruco tells us that the pitch damping of the IMs has significant coherence with the 45 MHz REFL WFS pitch signals from about 8 Hz to 30 Hz. The first page shows this situation for both O3b and for a more recent lock (June 5). Reference traces are the recent lock and non-reference traces are the O3b lock.
As with the 9 MHz WFS signals, it seems that the appearance of HAM1 motion, mediated by the RMs, is stronger in the rf signals these days than during O3b (see LHO:63547 for that story). I suspect that explains the reduced coherence with IM motion in the more recent lock below 20 Hz. In other words, the IM motion is still there, lying below the RM noise.
The second page shows some extra traces, including a modest coherence between the 9 MHz WFS signals and the IM motion around 25 Hz during O3b. Again I hazard a guess that the IM motion is lurking below the RM motion in these signals.
Bruco of REFL A RF45 I from O3b is here. Gabriele's recent instructions on running bruco are at LLO:60390. I used Elenna's REFL WFS calibration numbers again (LHO:62976).
Roger! Great find! We're still using the former filter HTTS filter design on the RMs. We should be able to improve OSEM sensor noise rejection there in the frequency region you call out by upgrading to the "level 2" design shown in G2200722 (which for now is just a link to LHO:60645, but the DCC entry is easier to find.) Over the coming weeks (once I finally figure out what's causing the systematic error in the DARM calibration LHO:63607) I'll cook up a similarly improved damping loop design for the IMs (which are HAUX).
Sheila and I took a look at the IM1-4 damping loops after Evan pointed this out. We noticed a couple things: the damping filters are exactly the same for each degree of freedom and the gains are wildly different (-20 vs -0.02). I took spectra of the IM1 degrees of freedom (first screenshot) and noticed that the undamped spectra (red) do not show some features that the damped spectra do (blue). We think that might be evidence of gain peaking.
Also, I reran the IM1 L OLG measurement that Jeff took in Dec 2021 and noticed a very different result (second screenshot). In the first plot, the blue trace shows Jeff's measurement from 2021 while the red trace shows the measurement now. I used the exact same excitation, but the coherence was very bad (plot two). Plot 3 shows the unexcited SUS_IM1 suspension in pink and green as compared to the excitation in blue and red. Something is not right. I also checked that the gains have not been changed since these measurements were taken (third screenshot).
