To check that this worked as expected, we did a sweep of the MC IP by 600um (seeimage) and observed a change in FmodErr of 20Hz (with a IB DSP gain of 0.5). This is a little low, so I put 0.65 in the DSP gain for this signal to get the nominal 23um/Hz calibration. The conclusion is that the current signal is not the same as the old signal, differing by at least a gain factor of 1.5, and possibly by an offset. Unfortunately, the images from the original calibration of the FmodErr signal seem to be lost.
In the end, the ITF locks with the new FmodErr set to zero, but for noise
purposes we might need to do better than that. Of course, it is not
certain that the signal was right before, so we might just need to
remeasure the coupling... for now I guess we leave things as they are.
