The cupola of the filter cavity input mirror vacuum tower was opened since 8:30am yesterday for fixing the vertical motor of the payload. After the motor was recovered, we closed the vacuum tower again and worked on the characterization of the inverted pendulum since the electronics was changed a while ago. 

Since we do not have access to the LVDTs, we first calibrated the LVDTs signals with the trillium by injecting lines to each DoF of the LVDTs (x, z, ty). Then by checking the transfer function between the actuation and the trillium readings(x, z,y), we get the error matrix and feed this back to the old driving matrix. In this way, through a few iteration, we aligned the LVDTs axes to the trillium axes.

Using the new driving matrix, we updated the sensing calibration factor for converting LVDTs voltage to um, including also the y direction.

Then we diagonalized the LVDTs DoFs again by injecting lines to x, z and ty, and measured the transfer function between the LVDTs signal and the actuations. The driving matrix is then updated accordingly. The final driving matrix was well diagonalized in z and ty DoFs, while in x direction, we had about 30% of coupling into the ty. 

The position loop was successfully closed with lower gains then before and was stable for hours even the tower was still in air (attached figure shows the loop closed for half an hour). Then we also closed the optical lever loops with the usual gain (dithering lines were on in TX and TY).

We informed the vacuum team to start the pumping of the tower around 4:40pm.