Sheila, Gabriele

Attached is a plot and script that we can use to get an idea of our recycling gain during the CARM offset reduction.  The plot shows the reflected power (normalized by the reflected power at the start of the locking attempt) vs TR_X_NORM, which is the TMS X red QPD normalized so that the build up of a single arm lock is 1.  The slope of this line is determined by the recycling gain (Gabriele has some plots he will attach).  There are two historical traces to show what this looks like when we had better recycling gains.  

We can have difficulty locking when we have a poor recycling gain (due to a bad initial alignment or alignment references which haven't been set), since the relationship between sqrt(TR_X +TR_Y) (TR_CARM offset) and the actual CARM offset changes.  This can fool us into thinking that we need to retune the CARM offset reduction sequence (ie, loops go unstable or transitions seem to be happening to early or late causing locklosses), when what we really need to do is align the interferometer.  The traces from the last week on the attached plot illustrate this, the gold trace shows a lock when we first relocked and the CARM offset sequence was working well (the recycling gain is worse than O2).  The purple and green traces are from times when we had difficulty locking, Jenne and Georgia did an initial alignment and which restored our slope (baby blue) to more similar to the gold reference.  

There may be a way of using this information to make our CARM offset sequence less sensitive to misalignments.