We performed 2 different steps:
1) 12.35 UTC: 19°C->18.90°C
2) 13.18 UTC: 18.9°C->18.86°C
By acting on the chiller, the laser temperature decreased by 0.1°C restoring the 2% of CO2 power loss (fig. 1), and probably also the polarization state.
During these tests, no changes in B1p and B4_112MHz have been observed (fig. 1).
After the CO2 laser reached the steady state, Julia closed the control of BS in full bandwidth without any problems.
Then the ITF unlocked due to an earthquake.
We profited of the unlock to verify the proper behaviour of the BS control loop: during the relocking no issues have been observed on this side.
The behaviour of the 112 MHz has also been monitored and compared with that of the first day of run (April 1st-TCS well tuned) and with that of April 24th (TCS not tuned).
In presence of a TCS not well tuned the 112 MHz experiences an overshoot. With the today intervention the overshoot has been reduced (fig. 2).
Yesterday morning TCS_NI_C02_PWR started to have a bistable behavior, with a periodicity which is relaxing in time. Observing hrec at high frequency, where the effect of bad or good CMRF is well visible, one can see a similar behavior. At the same time, a long term evolution of the 'noise in the bucket' has been observed. For that, I have no idea about possible correlation with TCS tuning. In any case, I would be in favour to spend more commissioning time in some trial of TCS tuning, including NI in the business.
The oscillation of the NI laser was triggered by the switching off of the fan coil in the chiller room (reported by the operator in #45787). Since it seems to be fading away, I would wait some more time before touching the NI. In any case, tomorrow we'll look at the last lock and decide the next actions.
