After the work on the DSP card of the BPC,  we started to rise the input power up to 15W.

We decided to increase the power on IPC0 (roughly 35% more) and then act on IPC1 to fine tune in order to reduce the power dissipated on the beam dump on the Laser Bench,in front of IPC0. Indeed, this one is not water cooled and it was receiving more than 40W (its temperature reached 75⁰, now  ~10⁰ less) . 

We then slightly realigned the PMC, checked its error signal and consequently adjusted the attenuation (see fig.1). We changed the trigger threshold on BPC DSP card in order to take into account the increased PCM  and INJ power.
With the IMC locked and manually realigned, we decreased the IMC_REFL  DC and PSTAB_PDs  in order to reach the previous values (fig. 2).

Finally, we updated the RFC threshold  on the dsp card and IMC thresholds on ACL and tried to close all the loop. Since we didn't expect any huge difference, we directly tested the  Automatic Alignment. All the loops closed correctly, included fmoderr. To be done: re-verification of the AA sensing matrix. 
Additionally, we verified the different working points (angular and longitudinal):
IMC WP: 0.08 
NF_H -> 0.05, now: 0.15 
NF_V -> -0.15, now: -0.13
FF_H -> 0.0115, now: -0.1
FF_V -> 0.57, now: 0.55
All the values are now stored on ACL. 

Once verified that the arm cavities could relock, we aligned the PR and recovered the correct value for SIB2_B2_DC. The OLFT of the IMC is similar to what we had before (see fig. 3)

The injection system is now in its nominal state with the exception of a slightly increased noise at very low frequency still present on the PSTAB/PMC (compared to before the swap IPC0/IPC1).