Abstract:
REFL 51Q had a large offset when PRMI was resonating, so I changed the mixture phases of I and Q signals of REFL 51.
After changing the phase, I locked PRMI with 3F signals but still PRMI is not so stable.

Detail:
I aligned the IMMT2, PRM, and BS by using the same ADS in klog23076.
The POP 90 buildup was reproduced, so thealignment procedure seems fine.

After the alignment, I checked MICH and PRCL error signals and found that there is a large offset on REFL 51Q (fig1).
Maybe, this is the same situation as klog9406.
At that time, offset was just subtracted but we don't have MICH offset at least during O3GK, so there should be another solution (but I cannot find it on the klog..).

Since there was a log of PRMI unstability due to the phasing of PRCL and MICH like klog15181, I tweeked mixture phase of I and Q signals so that the offset becomes small when resonating.
As a result, phase was changed from 60 to 103, and offset becomes almost zero (fig2).

After that, I tried to lock PRMI with 3F signals but still PRMI is not so stable.
Attached movie is a POP and AS camera image during the PRMI 3F lock.
Sometimes, POP camera becomes bright while AS becomes dark, which is the condition we want to achieve.
Figure 3 shows the signals on various channels when PRMI 3F was locked.
The signals on POP PD sometimes increased while AS PD signals decreased, which is consistent with what we can observed on POP and AS camera.

I'm not so sure but is it a mode hopping?

Note:
I haven't checked FPMI can be locked after changing the phase of REFL 51Q signals.
So, FPMI might not be able to lock in the current setup.