J. Kissel

I've processed ETMX-BSC-ISI-ST2-coupling-to-DARM data that I've been taking over the past few weeks prior to using the Test Mass Discharge System on ETMX, and now including that taken immediately following Friday's IFO recovery to nominal low noise (see LHO aLOG 38494) after the successful discharge of ETMX last week. 

The results show more excellence: All DOF's of ETMX BSC-ISI ST2 coupling has been reduced by about a factor 100, bringing this noise source from "just below" to "several orders of magnitude below" the current sensitivity.
It's exciting that we can pretty definitively say that we've measured and subsequently mitigated real DARM noise that results from electrostatic charge.

See attached plots:
Figures 1-3: Noise budget style projections of the ETMX ISI's coupling during ambient conditions, comparing before vs. after. The "before" data is the most recent data I collected on 2017-08-28.
Figures 4-6: Screen captures of the DTT templates showing the raw measurements. It shows the three data points for each degree of freedom  I collected before the discharge, and the one after the discharge (The three "before" data points for each DOF are date-separated by about ~1 week).

This reduction in noise coupling corroborates with the reduction in effective bias voltage seen LHO aLOG 38469. However, we know that this effective bias voltage is sensitive to all charge coupling mechanisms, so it's difficult to tell from that data alone which mechanism has been neutralized. With this new data showing drastic reduction of BSC-ISI coupling, it may indicate that some large fraction of this reduced effective bias voltage is from a reduction the voltage difference between the cage and the ESD system, decreasing the "C" and "D" terms in Eq. 1 of T1500467.

Also of note: We hadn't been watching this BSC-ISI / Cage coupling for long enough prior to discharging to notice any statistically significant trend / drift in the coupling like we see with the effective bias voltage measurements and/or the relative longitudinal actuation strength, but there is *some* evidence that it was evolving in this data: the X / L data shows a 10% descrease from 2017-07-25 to 2017-08-22 (but no change from 2017-08-22 to 2017-08-28). Neither the RX / P or RZ / Y data show any change. Tough to conclude any time dependence from this little of data. 

Details:
%%%%%%%%
The data are using the same excitation parameters as described in LHO aLOGs 37752 (for X), 38122 (for RY), and 38132 (for RZ) for all times.

The data was gathered using DTT templates,
/ligo/svncommon/SeiSVN/seismic/BSC-ISI/H1/Common/
    2017-09-02_H1ISIETMX_ST2_BB_X_Injections_ETMX.xml
    2017-09-02_H1ISIETMX_ST2_BB_RY_Injections_ETMX.xml
    2017-09-02_H1ISIETMX_ST2_BB_RZ_Injections_ETMX.xml

The noise budget plots were generated with a special offshoot of Shiela's simple noise budget, that lives in the same location:
/ligo/home/sheila.dwyer/Noise/Noise_projections/simple-noise-budget/
    SimpleNB_ISIETMXST2_Only_TMDS_BeforevsAfter.m           << Model
    noise_projections_ISIETMXST2_Only_TMDS_BeforevsAfter.m  << Noise Projection