After today's commissioning meeting I was asked to look into the FSS oscillation threshold setting and why the FSS guardian was not detecting when the FSS was oscillating (thus necessitating by-hand adjustment of the FSS common gain every time the IMC loses lock and throws the FSS into oscillation).  Looking at the FSS model (1st attachment, middle of the picture), H1:PSL-FSS_OSCILLATION is set when H1:PSL-FSS_MIXER_PP is greater than H1:PSL-FSS_OSCILLATION_THRES.

Knowing this, I trended back a couple hours, looking for times when H1:PSL-FSS_FAST_MON_OUTPUT was oscillating, and compared them with the value in H1:PSL-FSS_MIXER_PP; see 2nd attachment.  This makes it clear why an oscillation threshold of 3 was not working to set H1:PSL-FSS-OSCILLATION; the mixer peak-to-peak voltage never exceeded 2.5V during times of oscillation, therefore the threshold was never crossed.  From talking with Jenne, apparently a value of 0.6 doesn't always catch an oscillation, so for now I recommend a value of 0.5 for H1:PSL-FSS_OSCILLATION_THRES.  This change has been made and accepted in SDF.