TJ, Ryan S.

We've been occasionally seeing some SDF differences come up after relocking related to the ALS Y REFL servo (most recently in alog77478), where the input gain and compensation boosts are being turned off during SHUTTER_ALS. After not finding any Guardian that does this, we decided it must be Beckhoff-controlled, but still wanted to understand why this only happens during some lock acquisitions and not others. When doing some trending, we saw that the gain and compensation boosts are turned on when the arm cavity is locked and turned off again when unlocked, but not turned off when shuttered (everything gets cleaned up after a lockloss). In some cases, after switching actuation from ALS WFS to DHARD, the green arm alignment drifts to the point where an arm (usually Y) unlocks and the input gain and compensation boosts are turned off before ALS is shuttered. As seen in the attachment, there's a ~5 second gap between when the arm nodes move to their SHUTTERED states and the light is actually shuttered, which we tracked down to a 6 second wait time in ISC_LOCK and believe is a remnant of old code. This wait time has been lowered from 6 seconds down to 1 second as there's really no reason to wait here anymore.

SDFs are correctly set and ISC_LOCK.py changes have been committed to svn, but the node will need to be loaded when next out of observing.