Camilla, Sheila, help from CDS team

We've been having some locklosses during the ESD transitions, some of them seem to be related to a large glitch that happens at the end of the 10 second output ramp time when DARM1 FM1 (a boost) is turned off in preparation to switch back to ETMX control where the boosts are in the pum rather than the DARM bank. 

Erik and Jonathan told us that since the filter ramp is linear, it's not suprising that there is a discontinuity when the ramp time ends.  They have opened a ticket after a discussion about ways to avoid this: 667

We could make this better by increasing the ramp time, but we want to keep this state short because of a slow instability we can get if we wait here: 81430. 

For now we've removed the boost which was being turned on in the state DARM offset before the ASC gets engaged, during the power up. The attachments compare the ESD drives with and without this filter being ramped off, the glitch is gone.