This is a follow up to this entry about the excess noise in the EX TR QPD signals.  The OSEM fix did not help this, so they seem to be unrelated.  I looked at the PIT and YAW signals, and found that they are very coherent with SUM in the 5-100 Hz band where the excess noise appears (plot 1), and that the phase of the transfer function in that region is 0 (e.g., SUM, PIT and YAW all have the same sign and phase).  This appears to point to a problem in segment 1, though I cannot confirm that with offline data.  I looked at the OUT16 channels for all segment and didn't see anything obviously wrong (plot 2).  Interestingly, the same signal appears in QPD A and B, so this is not a single channel problem.

To do: investige the individual segment signal in the online data... is this just sement 1?