In the past days I have restored the local control on the BS mirror (used as steering mirror after the telescope). The optical scheme (similar to that described here) is shown in picture 1. We have noticed that there are two reflected beams (from the first and the second surface) which are very close. In this configuration the two reflections cannot be separated and they arrive almost superposed on the PSD. Moreover the beam is impinging at a distance of about 4.5 cm from the mirror center. According to my computation (reported here), it should induce an error on the yaw measurament of about 3.5%.

I had some trouble in diagonalizing the driving. After many tries, I have found that the best diagonalization is achieved without using the two lower coils (see coils disposition in the last attached picture). This fact is very strange and should be better investigated. With the following driving matrix 

I found the mechanical TFs shown in picture 2 and 3 (when exiciting yaw and pitch respectively). They look similar to those measured last november.

The open loop transfer functions are shown in picture 4. Due to the second narrow resoance in pitch at about 10 Hz, the UGF is crossed two times. I'm not sure about the phase of the pitch open loop TF for the second crossing point. Anyway I was able to close the two loops and they look stable. We also added offsets and we were able to observe the change in the beam position on the target at 290 (both in vertical and orizontal direction according to the loop to which we add the offset) 

The comparison between the open and closed loop spectra are shown in picture 5. Maybe the UGF can be increased a bit.

The calibration is  0.37 mrad /V. It has been computed as done here. where V_SUM of the PSD is 13.5 V and the lever arm is about 0.7 m

Some remarks

1) In order to investiate the driving issue, I have injected a line at 5 Hz with the same amplitude in each coil (one by one). The spectra in the four cases are shown in the attached pdf and seem pretty much the same.

2) I have observed that while measuring mechanical TFs, at the begining I was not able to find a good coherence at low frequency (below 1 Hz) and the resonances both in pitch and yaw where always very excited. I found out that this was due to the air flow inside the cleanbooth. I have temporarily disconected it to make the measurements. 

3) I have observed an oscillation at 50 Hz in the four signals sent to the coil driver. The amplitude is quite high (about 600 mV pp) and it is not present if i look at the signals just at the output of DAC. I'm not sure about how to get rid of this. 

4) While glueing the magnets we checked that the north-south polarization was in agreement with the convention reported in the last attached picture. Now, observing the driving matrix it seems that the sign should be inverted. I remember that this happened also for the dummy BS.