Patrick, Jenne, Sheila, Evan

When the wind died down and we were able to relock using 45mHz blends, we decided to have a quick look to see if we can fix the problems we've been having with soft loops.  In the end we just made the low gain that we have been using with the soft loops when we are using 45mHz blends permanent.

Background:  some gain peaking in the soft loops unfortunately seems to be at the same frequency as the gain peaking in the 45mHZ blends, making it dangerous to turn the soft loops up to their full gain when we are on the 45mHz blends because of the microseism.  The operators have been manualy turning down the gain in these loops in the last week when they are using 45mHz blends.  

We wanted to try to fix the instability in these loops but didn't suceed.  Along the way we noticed that we have 20dB more gain in the offloading path for the ITMs than the ETMs.  In the words of Daniel sigg, this is the underbrush of historical growth.  

We made a few attempts to make the offloading paths have the same gain: first we turned the gain down in the ITM top masses, this worked fine until we increased the power when the ITMs saturated the pums.  Then we turned up the offloading gain in the etms to match the ITMS, this caused a problem with the SRC asc loops and also broke the lock.  In the end, we have just gotten rid of the increase in gain for the soft loops all together.  This means that the ugf of the soft loops is now probably 3 mHz or so.  

What this means for operators:

You don't have to worry about the soft loops anymore when you change blends.  

There is no longer a long wait for the soft loops to converge in ENGAGE_ASC_PART3, but there is a check in DC readout to make sure the soft loops converge before we increase power.