- the first unlock behaves as a fast unlock with the EOM_CORR saturating a +1, quite unsual

- the second unlock 21 mn later fetaures also a fast unlock with EOM stuck at +1. In the second plot the ML_AC signal has been rescaled by a factor "-6" and we guess the correpsondance with EOM_CORR signal which has lead us to think many times that the EOM_CORR voltage induces some beam jitter on the transmitted beam (through the crystals at Brewstr angle) or some RAM.

- the third unlock is due to the IMC alignement drift (third plot)