I did a few more things tonight

• Turned off LP80 in SRCL, this reduces the gain peaking in that loop from almost 5dB to 2dB.  The second attached screen shot shows the control signal, the orginal configuration (with LP80) is the blue reference, the green reference is with the LP120 that I added to SRCL2, and the red trace is with only the 200Hz cheby low pass on.  Since SRCL noise doesn't bother us at 100Hz, it seems OK to add some noise there to reduce the control signal at 40 Hz.  (I've removed LP80 from the guardian)
• I also did a little bit of SRCL FF tuning, I used the old FF filter which is not as good as Jenne's JFF from september, but lets us split the FF into high and low frequency parts to adjust the gains seprately.  The third attachement shows the marginal improvement this made, but Jenne's new and improved filter will probably be better. After this I ran A2L, there was a bit of coherence (0.2 ish) from 20-30 Hz with CHARD Y, so we can probably puch the gain in that loop down and make the low passing more aggresive. 
• The remaining noise in DARM from 40-20 Hz was not very stationary after these small changes, but slightly better than where it was at the start of the lock. 
• Next I tried a little bit of driving the compensation plate while watching the camera.  I ended up drive 600 counts into RO DAMP L at 0.1 Hz for ITMX, the second to last attachment shows the 4 fringe wrapping shelves that show up with our nominal alingment of no offsets.  I could change offsets and make some shelves larger while other shelves got smaller.  A compromise position was -93 urad P, no Y offset, which is shown in the red trace. I wasn't able to see anything convincing in the SRM/SR3 camera.
• After Cheryl relocked the IFO (I caused the lockloss), I did a similar excitation on ITMY CP, and different noise in DARM.  For CPY only 100 counts excitation was needed to see noise in DARM up to 100Hz.  Cheryl and I watched the cameras, and might have seen something that was related to moving the CP.  We were able to steer the CP to reduce the noise in DARM while we were driving it, but it seems to make no difference to DARM when we aren't driving.