Kiwamu, Dan

For a while now the OMC transmitted light has been flickering with an amplitude of ~20% of the full light level; this makes it look like the OMC is about to lose lock and generally puts everyone on edge.  The first image is a ten second trend of OMC-DCPD_SUM to give a sense of the size of the excursions.

The RIN is dominated by a broad plateau from 1-4Hz with an ampltiude of about 0.01 RIN / rt{Hz}. see the second plot attached.  Tonight we made some studies of this noise to pin down the source:

• The OMC TRANS intensity noise is much quieter when the OMC is locked to a single bounce beam, or to the 45MHz sidebands with the IFO fully locked.  (See the black trace in Fig. 2.)  So, the problem is not from the OMC LSC loop, and it's probably not from the DRMI.
• With the OMC locked on the carrier light, but DARM still on AS45_Q, the intensity noise is just as bad.  Compare the blue and green traces in Fig. 2.  This means that when DARM is handed off to DC, the intensity noise is not suppressed -- apparently this is from real DARM motion that is already being suppressed as much as the loop can manage in RF readout.  This appears to rule out AS port beam jitter that would couple to OMC TRANS but not AS45_Q.  Probably we could improve things with more DARM gain, but we'd prefer to fix whatever is moving.
• There's precious little coherence with anything.  The 3Hz peak is coherent with DHARD_Y and the ETMX_OPLEV_YAW.  There is a feature in the DHARD_Y plant inversion at 3Hz, and we thought maybe it was a little mis-tuned or tuned for lower power, but we measured the loop (Fig. 3) and it looks ok.  As a test we reduced the DHARD_Y loop gain by 6dB and the noise did not improve.  Anyways the 3Hz noise is just the tip of the iceberg.

One thing to do is run BRUCO and see what turns up in the 1-4Hz band.