Sheila, Jenne

We're looking into why we lost lock here after our nice long overnight + most of the day lock. 

Hang's new lockloss tool (result plots in /opt/rtcds/userapps/release/isc/common/scripts/lockloss/python/rec/lockloss_1124145611.531/) indicates that some ITM yaw LOCK drive outputs are some of the first to go out of "norm" after the lockloss, which led us to look more closely into ASC-related channels.

It looks like DHARD is our culprit here (see attached plot), with yaw going out of normal a teeny bit earlier than pitch.  Also in the attached plot are oplevs for all 4 test masses, and we see that none of them gets a kick of any kind by the time that the DHARD loops start to go unstable.

The frequency of the DHARD oscillation is about 18.9 Hz, which is awfully similar to one that Sheila saw earlier this week: aLog 20601.

The UGF of the DHARD loops is a few Hz, and the phase crosses zero degrees below 10 Hz, so it's not a loop oscillation.  Is there something crazy happening, like BS roll mode doing something that lets a signal out to the AS port, which then gets fed back into the DHARD loops?