Kiwamu, Stefan
We looked at the optical gain change during the 80min lock at 50W (alog 29556) as measuired by the 331.9Hz cal line, see plot 1. The reduction between the beginning of the lock (8:10 UTC, blue) and the end (9:15 UTC, red) is about 22%.
Plot 2 plots ASAIR_45_Q for the the period from 8:10 UTC to 9:15 UTC, since it should be a measure of effective fringe offset. Its value drops by 41% (from -526 to -373 microcounts). Also it shows a significantly bigger variation in the DARM offset towards the end of the lock.
While qualitatively this would be consistent with a decreased contrast defect feeding carrier light through the OMC, the magnitude of the gain reduction doesn't quite add up - the percent reduction ought to agree since both optical gain and AS_Q DC value are proportional to the DARM offset. Also, the AS_45_I DC offset remains rougly constant (not plotted).
Here are also plots of the DC values of the POP45 levels, together with AS_45_I and Q. Plot 1 is the long 80min lock, plot 2 the previous TCS tuning lock.
Note that the only non-servo'd POP_45 signal is POP_9_Q. In principle it should be a measure of side-band inbalance (of the 9MHz, but if the 45MHz had an offset, it would result in a MICH offset, which again should show up in the 9MHz.).
In both locks POP_9_Q driftsa toward 2900cts - unimpressed by the different TCS state.
Also shown here are AS_45 I & Q - note that I is about constant in both locks, although the TCS assisted 1st lock shows a light upwards drift of AS_I. Not sure this is consistent with the above theory that we have an increased contrast defect.
