V. Lhuillier, J. Kissel

    First, I had nothing to do with the performance of this spectra, but Vincent is far too humble. A factor of 20-50 below requirements between 0.5 and 1 [Hz]? A factor of 10 at 10 [Hz]? Totally spectacular. And this is *without* anything but position sensor control of HEPI, without GND to ST1 sensor correction, and by paying very little attention to degrees of freedom other that Y and Z. Spectacular work, monsieur.

    Second, this 6.18 [Hz] feature has been seen sporadically else where, as reported by the DetChar group at both sites (see, e.g. LHO aLOG 7159 and LLO aLOG 7976). Given the feature's characteristically, un-mechanical, sharp, Q, we are reasonably convinced the feature's source is either digital or analog electrical. The "best" guess -- and it's just one guess -- was that this is the beat note between two chamber's capacitive position sensor modulation frequencies (which are at a few [kHz]), where say, ITMY's ST1 H1 "coarse" sensor's master frequency (to which the remaining eleven are slaved) beats against BS's ST1 H1 sensor's master frequency, resulting in this sharp ~6-7 [Hz] feature. In order to test whether this is the case, we turned off*** all CPS in the entire corner station *except* for ITMY, and measured a quick spectra. The feature remained. So this is pretty solid evidence that it *is not* beating CPS modulation frequencies. The feature is *not* seen in ST1 sensors, and unfortunately the amplitude of the peak is just small enough to be hidden by residual seismic motion when the ISI is only under ST1 control, so checking if the feature is still present with ST2 OFF is tough. More ideas are welcome!! 
    (Putting on my "does it really matter?" hat: unclear. Detchar as shown that the QUAD does a great job of filtering out this motion by the optic, as any good quadruple suspension would. Further, ~6 [Hz] is outside of the gravitational wave band. Given that it's a sharp feature, it really may not affect the performance of the IFO at all. It's just odd, and a glaring feature in the ISI performance plots that begs questions.)

    Finally, in the interest of improving the usability of the chamber seismic isolation (and for fun), I timed the world's expert in aLIGO BSC seismic isolation from a fully-cascaded chamber watchdog trip, to get back up to this level of isolation. I challenged him to it in under a half-hour ;-). In all seriousness though, I timed him. 5 minutes 31 seconds. And note, the clicking of buttons stopped around 4:30; the remainder of the time was spent waiting for things to settle down after the ramp up of the loops. This being said -- there were about 50 clicks to get there, and he was *flying* through them. So, assuming that we turn the 50 clicks into one click, and take the overhead associated with the clicking as the overhead that the one-button-click script would need to complete, I this ~5 minute recovery time is roughly what we should consider to be the irreducible.



*** We turned off the CPS at the rack, using the separate ON/OFF rocker switch for the CPS in the back of the ISI Interface chassis -- which means the GS-13s and the chassis as a whole stayed powered *on* during this test. Note that "every ISI but the ITMY" includes HAM2, HAM3 (HAM4 and 5 are not yet running), HAM6, BS (ITMX is not yet running), and the ISI Test Stand.

Attached is a link the DetChar summary page with more info about these lines
https://wiki.ligo.org/viewauth/DetChar/SUS7HzLineStudyPlots