As written in the operator reports (Operation report - Afternoon shift) the automatic tuning of B1 demodulation phases was not working properly since the photodiode displacements (B1 setup moved from EDB to the small bench attached to the tower).

As expected the optimal demodulation phase for d2 changed of 0.5m/24m * 360 degrees = 7.5 degrees. Instead d3 phase changed much more (why?). The optimal one was positive and this is not possible with Lo boards. To recover the missing 15 degrees of dynamics I tried the following procedure: switch to B1_d1, change in Pr_ITF the sign of all d3 signals, rotate the phase of Lo3 of -180 degrees. Unfortunately I forgot that Lo3 controls also B1_d1 demodulation phase (which is a known bad configuration since a lot of time) and so caused the ITF to unlock.

Before relocking I changed the sign of all B1_d1 signals. Unfortunately now the demodulation phase of B1_d1 at step 9 is very mistuned. The drawback is a much reduced DARM optical gain for this diode, which caused unlocks when switching to B1 at step 9. To cope with this I had to increase, in Pr_ITF, the B1_d1 gain from 200 to 400 (lines 69 and 73).

I also had to change the sign of the slow B1_Lo3 servo in Alp_Moni.

In this configuration everything seems to be properly working and phases to be well tuned.