[Keiko, Masayuki]

We found that the IMC error signal is weak and the S/N ratio was roughly 2. We changed the IMC_REFL optics to increase the laser power on the RFPD and now, the PDH signal seems healthy.
Due to the optical gain change, the servo gain was changed from 4 dB to -28dB. The OLG is attached. (Red: current, Blue: before, Green: with bad cross-oever described as below)

Also I find that the MCL loop was too fast and the cross-over frequency was around 30 Hz. The dip occured around the cross-over frequency. I changed the gain of MCL loop and now the cross over frequency is around 4 Hz.

However, still we have excess intensity noise at IMC.