This afternoon we were sitting in NLN_CAL_MEAS for a couple of hours doing commisioning for the RF mdoulation depth and test mass coil drivers. While we were sitting there we noticed DARM getting noisier in the 10-30Hz band. When we went back to NOMINAL_LOW_NOISE this noise slowly went down again. In NLN_CAL_MEAS we have the ADS off, so the alignment slowly drifts (notably in pitch), making our A2L coupling worse. In the attached plots the top 2 plots are the DARM BLRMS in the 10-20 and 20-34 Hz bands. The second row is the PIT ADS inmons and outputs (from t-2 to t+2 these are off since the ADS ligns are off). Third row is ADS yaw, and fianl row is the guardian state.

These loops are incredibly slow. We used to have an additional 20dB of gain in the PIT loops which we commented out because we were having some locklosses. I've just now modified LOWNOISE_ASC to set the gains to 4 instead of the usual 1. We might want to push this up further.