Yokozawa, Michimura, Ushiba, Tanaka

This morning, Yokozawa-san found PRMI kicked when its 3f lock acquisition. We found that the phasing of RF51 seemed to be bad when the demod. phase value was 60 deg. So we rotated the phase to maximize the Q output when we excited MICH. The value after the rotation was 145 deg.

Next, HANDOVER from ALS to IR got hard because the glitch in MICH error signal when IR beam flashed in the arm cavity became large. This issue is the same as the one in klog30912. So we need to make the demod. phase worse in order to reduce the glitch

On the other hand, Michimura-san found that the RF51 and RF135 signals had dark offsets (~ several cnts). We subtracted them (fig.1) and then, the PRMI could locked smoothly even though the demod. phase of RF51 was 60 deg. HANDOVER also succeeded with this demod. phase. 

## NOTE

I also subtracted the dark offset of AS_PDA1_RF17_{I,Q}, REFL_PDA1_RF45_{I,Q}, POP_PDA1_RF{17,45}_{I,Q}, POP_PDA2_RF90_{I,Q} manually.