Jenne, Craig, Georgia, Sheila, Kara

We had a couple of locklosses this afternoon when the dither/soft loops were engaging.  We have made some changes in the guardian that we think will help prevent this in the future.

• Housekeeping: we moved the turning on of all the dithers into PREP_ASC, we also deleted the gaurdian states that were for the slow ASC engaging, because these haven't been kept up to date and weren't doing things correctly.  We also made sure that the SOFT pitch blends are off. 
• We have split the SOFT and dither loops out of ENGAGE_ASC into a state called ENGAGE _SOFT loops.
• before we were turning on the CHARD loops at high bandwidth before engaging the dither loops.  This means that if the alignment isn't great, the A2L noise from the CHARD loops covers the dither lines in DARM.  We are now leaving CHARD in the lower bandwidth state (~40dB lower gain than used for power up) while the dither and soft loops are engaged, and then increasing their gain once the dither loops have converged.
• We also ran into the soft limiters on SRC1 P when the soft loops were engaging once.  We are now waiting for all the ASC except the soft loops to converge, turning off the soft limiters, and then engaging the soft loops so that the other loops can follow with their full gain.

We see that spot positions on the TR QPDs are rather different now than they were before the dither loops were commissioned (attachment), especially the Y B QPD.  We would like to try moving these spots while in low noise to see if we can reduce the fringe wrapping we are seeing, but in the past we have difficulty moving our spots after powering up because of the dP/d angle instability.  Georgia is trying right now to move the spot while we are at 2W to be more similar to the old spot.