Sheila, Thomas, Craig

We wanted to spend some time today looking at the DHARD P loop.  On Friday Jenne reduced the gain of this loop by 10% to make it stable, but as you can see from the measurement attached to her alog (46636), this loop is very marginal.  Hang has done some modeling of how this could be caused by a cross coupling, 46469, This situation is very similar to the phase rotation problem we had in CHARD Y in Septemeber 43946, which turned out to be due to a stray gain in the top mass offloading for one suspension.

There are small differences from 1 in the L2_LOCK_P/Y gains which were set in 2016 27494 and haven't been changed.  Since these were set 2 test masses have been replaced, so we might need to re-tune these.

When I arrived this morning, the interferometer was locked at 30W without the ASC cut offs engaged.  I measured the DHARD P loop from 0.5-2 Hz, and tried turning off the DSOFT P loop to see if that was the source of the cross coupling.  The DHARD P transfer function didn't change at all, so dsfot P is not the problem.

The attached screenshot shows several DHARD P OLG measurements, taken at different powers, after we recovered from the spontaneous reboots.  In the last measurement we turned off the top mass offloading for all four test masses, remembering the story from September, but we still have the same phase situation with the top mass offloading off.

I think the next step that we should take to investigate this is to measure the response of each suspension from ASC input to the optical lever, when the suspensions are set up as they are in lock, as we did here:  43946