As noted in alog 27683, turning on the 10dB gains in the M0 Lock filter banks for increased offloading speed was killing the lock. We saw that again today.  I turned on all the filters by hand, and it seems that it's the ETMY Y that is the problem - all others are fine.  Sheila pointed out that ETMY has filters for Y2Y and P2P in the M0 drivealign banks, but no other quads do.  We turned off the invY2Y filter, and we were able to run the EngageSoftLoops state.  We didn't turn off the invP2P, but we probably should so that all the quads are the same.