This morning Tony and TJ had a hard time locking the OMC.
We've found that OMC DCPD A and B output are very assymmetric only when there was a fast transient (1st attachment) but not when the OMC length was slowly brought close to the resonance (2nd attachment), which suggested whitening problem.
The transfer function from OMCA to B suggested that the switchable hardware whitening was ON for DCPD_A and OFF for B when it was supposed to be OFF for both. 3rd attachment shows the transfer function from DCPD_A to B, and 4th attachment shows the anti-whitening filter shape.
Switching ON the anti-whitening only for DCPD_A made the frequency response flat. Trying to switch the analog whitening ON and OFF by toggling H1:OMC-DCPD_A_GAINTOGGLE didn't change the hardware whitening status, it's totally stuck.
We tried to lock the IFO by only using DCPD_B, but IFO unlocked for some reason.
After IFO lost lock, people found on the floor that it's the problem of the whitening chassis, not the BIO. It's not clear if we can fix the board in the chassis (which is preferrable) or have to swap the whitening chassis (less preferable as calibration group needs to measure the analog TF and generate a compensation filter).
We'll update as we make progress.
Fernando, Fil, Daniel
DCPD whitening chassis fixed.
We diagnosed a broken photocoupler in the DCPD whitening chassis. Since the photocoupler is located on the front interface board, we selected to swap this board with the one from the spare. This means the whitening transfer function should not have changed. Since we switched the front interface board together with the front-panel, the serial number of the chassis has (temporarily) changed to the one of the spare.
The in-vacuum DCPD amplifiers were powered off for 30-60 minutes while the repair took place. So, they need some time to thermalize.
Unit installed is S2300003. The front panel and front interface board was removed/borrowed from S2300004.
N.B. S2300004 and S2300002 have been characterized and fit already. See LHO:71763 and LHO:78072 for the S2300004 and S2300002 zpk fits, respectively. Should the OMC DCPD Whitening chassis need to be fully swapped, we already have the information we need to install the corresponding compensation filters in the front end and in the pyDARM model to accommodate that change. This, of course, rides on the expectation that the electronics have not materially changed in their response in the interim.