I carried out the sensor and actuator diagonalization in the IP. However, the resonant frequencies in L and T are very close, making it difficult to distinguish sensor coupling from actuator coupling. Both are at 66 mHz using 5 mHz resolution in the TF measurement, values that are consistent with klog  17208.

At a first glance, the results don't seem encouraging, as large coupling between L and T remain (see the final result in the figures). I'll analyze the data more carefully and explain.

The matrices that I calculated are:

LVDTALIGN =

[[ 1.          0.          0.        ][ 0.          1.          0.        ] [-0.0151811  -0.00525001  1.        ]]

ACTALIGN=

[[ 9.77986387e-01 -1.13989532e-01  5.24470303e-02] [ 1.89263099e-01  9.78481134e-01 -5.07918172e-02] [-9.46772968e-04 -8.76784587e-03  1.00048999e+00]]

I haven't commited this values in the SDF.

----------------------------

The files I used live in this directory: /kagra/Dropbox/Subsystems/VIS/TypeBData/SR2/TF/Measurements/20221011/

------------------------------------------------------------

This is a back up of the old ACC2EUL matrix that I removed before begining the work.

####### SR2 IP ACC2EUL
### JGW-E2012144-v11 != JGW-D1707269-v2 (R=0.5915, H1=-75.0, ccw=True)
[[ 0.64395055, -0.47140452, -0.17254603],
 [ 0.17254603,  0.47140452, -0.64395055],
 [ 0.56353903,  0.56353903,  0.56353903]]

### current EPICS
[[ 0.70717917, -0.4365051 , -0.24289589],
 [ 0.18548488,  0.4903748 , -0.6430353 ],
 [ 0.63563217,  0.61924465,  0.59764597]]

### diff
[[ 0.06322862,  0.03489942, -0.07034986],
 [ 0.01293885,  0.01897028,  0.00091525],
 [ 0.07209314,  0.05570562,  0.03410694]]

### diag = current * theory^{-1}
[[ 1.05460442,  0.10899513,  0.01643079],
 [-0.00115293,  1.01587877,  0.01941562],
 [ 0.02141928,  0.02510413,  1.09576723]]