On Thursday, Betsy and I moved the OSEMs back in on the OFI, so I took the opportunity to take some quick transfer functions this afternoon. They are rough and will need some tweaking, but look as we expect. The most notable change is in Yaw, where our resonant frequency started at 1.04hz with our old sled, to 1.148hz with our new sled and supporting optics and dumps. LLO found that their peak shifted up as well, but ours is a bit higher than their shift (~1.12 vs our 1.146). I talked with Jeff and he assured me that our numbers look good since the modeled moment of inertia decreased we should see the frequency increase, and since our balance masses are slightly different, we should see a slightly different frequency as well. They are close enough that he isn't worried.
Jeff and I played with the filters briefly to try to clean them up, but ultimately gave way to figuring it out another time. In the, possibly, near future, someone will take these again with some better filters.
"I talked with Jeff and he assured me that our numbers look good since the modeled moment of inertia decreased we should see the frequency increase, and since our balance masses are slightly different, we should see a slightly different frequency as well. They are close enough that he isn't worried." I would say this a bit differently: "- Jeff sees that in LLO:56189 Arnaud states that the SolidWorks assembly D0900623 change from -v12 to -v13 predicts a *decrease* in Yaw moment of inertia once we change the breadboard from O3 to O4 configuration. - Jeff agrees that a *decrease* in moment of inertia will *increase* the resonant frequency, as both L1 and H1 OFIs now see. Cool! - Jeff suggests that the difference between arriving at L1's 1.12 vs H1's 1.146 Hz resonance could *very* easily be a small difference in Yaw moment of inertia of lots of things, especially balance masses, so he's not surprised to see a small, 20 mHz difference between the two real instantiations of balance from the sites. - Jeff is reminds the crowd that the OFI is still using crude, starter damping filters that are just "velocity damping," i.e. a zero at 0 Hz, and a way-too-high a frequency of two poles at ~100 Hz, coupled with a way too aggressive (high Q) elliptic with too high a corner frequency (~50 Hz). In short -- there's nothing (e.g. a gain boost filter) that's tuned to any specific resonance frequency. So a shift of resonance frequency from the O3 1.04 Hz resonance to the higher H1 O4 resonance frequency of 1.146 Hz does not impact the robustness of the damping loops in any significant way whatsoever, because the design is robust against such small changes. (These filters, that we've been copying around as "starter filters" for years were designed to "just work" on any LIGO suspension that has resonances at any frequencies between 0.1 and 10 Hz.)"