J. Kissel, S. Aston

I've gathered TFs of the IMs before the HAM2 doors are installed. They're free, see first attachment.

More importantly -- confused by a large discrepancy between model and measurement, I noticed in
/ligo/svncommon/SusSVN/sus/trunk/HAUX/Common/MatlabTools/plotHAUX_dtttfs.m
that the coil driver transconductance had not been updated (reduced to 0.998 [mA/V]) since we reduced the coil driver strength (see LHO aLOG 8758). Correcting this, the calibration *still* was off by a factor of 2ish. Exploring further, I found the OSEMINF filter's "to_um" filter was a gain of 0.0391251 [um/ct], instead of the expected 0.0233333 [um/ct] (see LLO aLOG 4291) that installed in every other suspension. As such, I multiplied the measurement by a correction factor of (0.039 / 0.023) ~= 1.6, and achieved the expected match of model and measurement.

I've now corrected the OSEMINF "to_um" filter, and corresponding "norm_um" filters in the DAMP bank, changed at 10:20:21 PDT (17:20:21 UTC). I've also cleared out all unused extra filters in the DAMP bank to avoid future confusion.

Comparing against previous measurements we see resonances in the same place, but off by the expected calibration factors (see last 4 attachments).

I would have compared against LLO's results, but data from each of their SUS show a wildly different overall scale factor, making comparisons difficult at best. I'll see if Stuart can chase down the difference once he get's back to LLO.