This morning I measured IM transfer functions with the horizontal OSEMs to make sure the IM was not touching the IRM. The DoF involved are L, Y and T. As can be seen from the pictures, all them look fine, however, the IM-Y transfer function deserves more explanations:

• Although the amount of actuation applied was 1000 counts in the original diaggui file, the actual actuation seen at the output of the test channel was only about 200 counts in amplitude. The coherence was relatively low at some frequencies (see klog 16207).
• I increased the actuation to 2000 counts, giving an effective 400 counts in the test channel. As seen in the corresponding picture attached the coherence improved and the peak 562 mHz dissapear and became a notch.
• A picture shows the time series with the actuation signal with the two different amplitudes.
• It seems suitable to review the reference measurements and compare them with the SUMCOM model.

About the transfer functions:

• The ones for L, T and Y look good, suggesting the IM is hanging free.
• The ones for R and P, which rely in the vertical OSEMs, have an offset. Effect that likely arises from the OSEMs being out of linear range. Notice the coherences are good.
• The SF is locked so I didn't measure the IM-V TF.
• During the measurements the BF keystone was at nominal position given by the setpoint (reached by using about -14,000 counts of actuation). Strangely, this position does not favor the vertical OSEMs; V2 and V3 tend to be out of range.