Jeff K., Edgard, Lee (by providing IIRrational for fitting), Hang

During today's locking we got a brief moment to test the ISIFF at the GRD state RESONANCE (CARM on resonance, DARM on AS45, DHARD ctrl w/ ~ 3 Hz).

In the attached plot, the top-left panel shows the DHARD ctrl output for two consecutive measurements. The red/blue were curves for the ISI to TOP L2P FF on, and pink/cyan ones were without the FF. Also shown in the bottom-left panel was the coherence between DHARD ctrl and the ISI SUSPOINT L. For clarlity we only showed the ones for the DHARD coherence with the ETMs, but the reduction in coherence with the ITMs were similar. In the bottom-right we showed the similar on/off comparison for the local ISI to OPLEV reduction for the ITMs.

It seemed that we could get ~ x2 to x3 reduction of the DHARD ctrl in 0.1 - 0.6 Hz band, and ~ 20% in the total rms (note that our current microseismic was still much lower than average). This should open up the possibilitiy of reducing DHARD ctrl BW in the future when we start to tune the ASC. The price we had to pay was the contamination in the <0.1 Hz band as we were injecting tilt motion to the ASC loops. However, we should be able to do turn on a second boost stage at ~ 0.1 Hz without eating too much phase at UGF of ~ 1 Hz. Thus this should contamination should not be too much an issue.

We edited the guardian to turn on the ISIFF in the RESONANCE stage. Will try to mitigate the contamination to the earthquake band in the future.