I did some studies last night on the sideband PRMI which was then interrupted by an earthquake at around the midnight.

(Yaw motion)

Last night, the yaw motion in the Michelson looked larger than usual. The fringe pattern in the dark port camera fluctuated horizontally at a low frequency roughly 100 mHz -ish. According to the oplev signals, the BS was the loudest below 1 H and louder than ITMs. Though both ITMs and BS showed a high coherence below 1 Hz with the dark port DC signal, i.e. ASAIR_A_LF. I did not take a close look, but it seems the BS glitch issue is now much better.

(REFL_RF45)

After locking the PRMI on the sidebands, I checked the signals on REFL_A_RF45_Q. The motivation is that we would like to switch the sensor of MICH to REFL_RF45. As Rana told me, the signal fluctuated by a several 100 counts around zero, even though the PRMI had been locked with REFL_RF9 and ASAIR_RF45 for PRCL and MICH respectively. I noticed that the REFL_RF45_Q seemed coherent to POP18 and I got an impression that the signal depended on the PRC alignment fluctuation. I then moved the REFL beam to a point where both WFSs were centered by engaging the DC servos, but this did not change the situation or the amount of the DC light on REFL_A.

I was not able to adjust the demodulation phase because the Q signal had been sensitive to a PRCL excitation at 41.3 Hz regardless of how big the demod phase is. Something is not right. According to free swinging PRMI fringes, the demod phase should be close to 0 deg or so. It maybe interesting to look at REFLAIR_RF45 as well to see if this nonsense is common in both in-vac and in-air PDs.

Also, I noticed that as I steered RM1 and RM2, the amount of DC light on the WFS_A decreased by roughtly 30% while the WFS_B did not change. This smells like a clipping in the WFS sled and therefore does not explain the large fluctuation in the REFL_A_RF45.