Following alog 47217 we set the DCPD input matrix.

- Added a gain of 0.977 to FM10 on DCPD_A. This matches the shot noise level in PDA and PDB

- Then the signal ratio g=PDA/PDB was measured to be g=1.021. Base on that, using the script in alog 47217, the matrix was calculated to be

  1.01039   0.98961
  0.98961  -1.01039
 

This matrix was loaded and committed to SDF, as was FM10.

I verified that the coherence of PDA-PDB agrees with DELTA_A-DELTA_B, i.e. the mixing and un-mixing of the signals indeed does not introduce a error.

Attached is a x-power measurement. Note that there is evidence for correlated noise at higher frequencies. I suspect that comes from glitches.

(The stuff at lower frequencies was an injection.)