This morning I came into the control room as the magnetic injections were finishing, and ran the a2l script while the charge measurements ran.
I did this because after Jennie Wright ran A2L yesterday afternoon it was not well tuned: 77495
When Jennie ran it yesterday, the coherence was over the threshold for all 8 gains, so all 8 were adjusted. This morning I ran it and the coherence was below threshold for ITMX P2L and ETMY Y2L, so those were not updated. I increased their amplitudes in the run_all_a2L.sh script, but as I was re-running the IFO unlocked due to the charge measurements.
The attached screenshots show how it changed the optics that it did run for, this is a large difference from the values that Jennie got yesterday. Jennie W ran the script 20 hours into the lock, while I ran it 37 hours into the same lock, so these were both with a well thermalized IFO.
It seems that there may be two problems we are facing with tuning A2L, one that our process doesn't seem to always work well to minimize A2L, and second that it seems that the values may be changing over the course of a lock.