daniel.sigg@LIGO.ORG - posted 11:43, Thursday 09 May 2019 - last comment - 15:47, Monday 03 June 2019(49150)
Updated calibration for IMC-I and LSC-REFL_SERVO_CTRL
The IMC-I and the LSC-REFL_SERVO_CTRL readback channels have calibration factors in Hz, which need to account for the optical gain in the IMC REFL path. They were outdated even before the recent increase in optical gain. The new value is 4.4 times smaller than the previous calibration factor.
The REFL_SERVO_CTRL filter banks also contains a filter labeled aogain. This needs to be the exact same value as the IMC-REFL_SERVO_IN2GAIN. It has been changed to -22dB from -24dB.
Here is an updated plot of frequency noise spectra related to the IMC. The horizontal magenta line corresponds to the shot noise in the IMC REFL PD. The IMC servo suppresses the frequency noise from the laser below sensing noise at frequencies below 4kHz. There are coherent peaks in REFL_SERVO_CTRL and IMC_F which are not laser frequency noise but acoustic jitter peaks. Not sure the frequency noise spectrum deduced from PRCL makes sense.
Compare this with alog 31554. Back in October 2016, we had 4.1mW on the IMC REFL PD in full lock. One thing to notice is that the laser frequency noise above 3kHz into the IMC was a tad bit smaller in the past.
Here is a plot showing the noise at frequencies above 7kHz. Above 3kHz IMC-F actually got worse since O2 (the two ~15kHz poles that were recently added to the readback are not compensated). Unclear where this noise comes from.