Naoki, Sheila
After the OM2 change, the output alignment seems to change shown in alog70886 and this might have caused the SQZ ASC running away yesterday in alog70890. The last time we measured the SQZ ASC sensing matrix was 75W, cold OM2 in alog70308. So we measured the SQZ ASC sensing matrix again with 60W, hot OM2.
We kept the SQZ ASC WFS switch ON and turned off the input switch of the SQZ ASC filter bank. We measured the sensing matrix by stepping ZM5/6 P/Y by 100 count as shown in the first attached figure.
| ZM5 P | ZM6 P | |
| AS A RF42 P | -0.4 | 0.49 |
| AS B RF42 P | -0.23 |
| ZM5 Y | ZM6 Y | |
| AS A RF42 Y | 0.37 | 0.46 |
| AS B RF42 Y | 0.3 |
The sensing matrix changed a lot compared to previous sensing matrix. We calculated the inverse matrix and decided the new input matrix by multiplying the inverse matrix by -10 as follows.
| AS A RF42 P | AS B RF42 P | |
| ZM5 P | 43 | |
| ZM6 P | -20 | 35 |
| AS A RF42 Y | AS B RF42 Y | |
| ZM5 Y | -33 | |
| ZM6 Y | -22 | 27 |
The SQZ ASC works well with new input matrix. We accepted the new input matrix in SDF as shown in the second attached figure.
