(Sheila Daniel)

Concerning alogs 31573 and 31679, and FRS 6787.

Increasing the thresholds didn't work because the tidal system is engaged, when we are first locked on RF at the AS port which is before the CARM reduction. At this stage the thresholds need to be small, since the arm power is still low. It won't trigger later because we loose the ALS. So, we made a change to guardian to increase the ON/OFF thresholds to 8000/4000 (from 30/10) at the same time that we change the NSUM whitening gains (which is after CARM reduction). They are reset in the down state. We also removed the 3 sec ramp time from the NSUM filter modules. They introduced a false down transient in the arm powers. Without the ramp it will now produce a false up transient.

At the moment these thresholds are optimized for a 2W lock. A further improvement could be to scale them with input power.