[Sheila, Jenne]

This morning DRMI had been locking reliably although was very sensitive to alignment.  Then it started to struggle to lock.  It looked like when our SRCL input offset (that tries to keep us from being locked on a bad SRCL mode) comes on, we actually transitioned *to* a bad mode.  At least once we have locked easily with the offset commented out of guardian.

Since this seemed a little weird, we checked to see what value of SRCL input offset gives us a minimum value for POP 90.  This is the same method from alog 14531 (Oct 2014) and repeated in June of this year in alog 42640 where we move the SRCL offset around and find the location where POP90 is minimized.  The offset value in both Oct 2014 and June 2018 was -800, but today we find that the minimum value of POP 90 happens with an input offset of about -1225.  (In the attached figure, I plot the SRCL_INMON value, so the offset that was applied is the negative of the plotted value). 

For now, however, we are just leaving the SRCL offset commented out and we'll see how locking goes for the next few attempts.

Alog 14550 (a comment to the Oct 2014 measurement) gives a SRCL calibration of 1.65e11 cts/m.  If we say that this is at least still roughly true, then the 800 count offset we had been using corresponds to 4.8nm, while the new (potential, untested in practice) number corresponds to 7.4nm. 

For this test, we moved the SRCL offset by hand, and then I use the attached script to plot the result.