Here is a prediction of the OMC length noise with the configuration we have been using for the last few weeks (dither amplitude750 counts, length servo digital gain 24, resulting in 6Hz ugf, before the addition of the boost described in 33989)

The length noise is larger than it was the first time I checked this, because of the reduced dither amplitude, but not really a problem as is. 

I described how I made these measurements in detail in alog 30510, which includes links to other meaurements. This time I relied on 12 Hz excitations to increase the OMC length noise to be visible in DARM. To briefly summarize, at high frequencies I estiamted the length noise by locking the OMC on the sideband with the ISS second loop gain increased, at low and mid frequencies using the in loop error signal, and the excitation amplitide was estimated using the height of the second harmonic.  Then the resulting RIN in transmission of the OMC is estimated, and the DARM loop supression corrected for.  

The three attachments show:

• the resulting estimate of the OMC length noise (including some checks on the way that I estimated the noise using the in loop signal),
• comparisons of the predicted DARM noise with an exitation on to the measured DARM noise (they pretty much agree)
• comparison of the RIN caused by OMC length noise in normal conditions to the normal DCPD RIN.  
• A noise budget with OMC length noise included