Over the last couple weeks I have been fitting in SEI to DARM injections on the input side HAM chambers, like the series of injections Huyen did here. I generally followed the excitations that Huyen put in that alog, but for the HAM2&3 ISI injections I used slightly smaller drives in places, because I wasn't sure how hard I could push without breaking the lock. For each chamber I drove in the cavity length, vertical, table pitch and yaw, or X,Z,RY & RZ for each of the input hams. For each excitation, I used the feedforward paths available , like HEPI to ISI for HAM2&3  or the never yet used 3d l4c FF path for HAM1, after turning off the inputs and disengaging any filters. This bypasses the ISO filters and meant I didn't have to do any compensating for loop gain.

I then talked with Sheila about how the DARM projections were done, because it was faster for me reinvent the wheel for myself in matlab than it would be for me to figure it out in python. After getting calibrated asds for each of my excitation I used: 

 coupling = sqrt ( darm excited ^2 - darm quiet^2) / (witness excited ^2 - witness quiet ^2)
 estimated ambient = witness quiet * coupling

where the "witness" was the GS13 or L4C in the direction of the excitation. One thing I'm not 100% sure of is the calibration for the CAL_DELTAL channel in matlab.

Three attached plots are the results for HAM1. First compares the injections for each DOF (X,Z,RY & RZ) for HAM1 as seen in DARM and the witness, to a quiet time on Jul 5. The second attachment shows the estimated ambient coupling to DARM for each of the DOFS. Third attached plot shows the estimated ambient coupling for RX on top and the noise injection vs quiet on bottom, just couldn't fit it in with the other DOFs.

Results seem mostly similar to LLO. But it does seem like our HAM1 is maybe a little closer to DARM than LLO. I would like to try the Z excitation again, I think that I probably wasn't driving hard enough 10-20hz and that is making it look like HAM1 HEPI Z is closer to darm here that it really is. I should also try including the tabletop L4Cs, but that will take a little more time, and the HEPI excitations I did won't exactly line up with the tabletop motion.

I'll attach the HAM2&3 couplings as comments to this alog.