1. OPTION 1: We set the M1_DAMP_Y filters to 20% gain, leaving all other M1_DAMP filters at 100% nominal gain. Measure all of the Yaw estimator transfer functions. Then, set M1_DAMP_L and M1_DAMP_P to 20% gain, leaving everthing else at 100% nominal gain. Measure all the Length/Pitch estimator transfer functions.
2. OPTION 2: We set all three of the M1_DAMP_{L,P,Y} to 20% gain and get all measurements for L, P , and Y transfer functions. This method is faster for measurement, but needs more fitting to deal with cross-couplings to Yaw.

Here are the thougths

• The risk of measuring all with 20% damping simultaneously is that we might have to fit  L/P-to-Y and Y-to-L/P transfer functions because the cross-couplings will peak at the lightly damped resonances. 
  Two cases to consider:
  1. In the worst case scenario, this is 4 extra transfer functions per stage to fit, added to the 5 we already fit. So for the HSTSs it will be 16 extra models, for a total of 36 fits.
  2. If the cross-coupling resonances are small with light-damping, or if the estimator to limits the drive to the cross-coupled plant enough that we can ignore the cross-coupling effects, then we can get away with only doing 20 fits.
• However, we can use the HLTSs as an example: if a 20% LPY damping measurement does not change the Y2Y transfer function for the PR3/SR3 compared to 20% Y and 100% L/P, then the model we are running on the HLTSs is identical to case (b) above, the best case scenario. 
  • Since the estimator on SR3/PR3 seems to work fine without L2Y/P2Y couplings, I see no loss whatsoever in doing 20% damping for all three.  [This is the thought that convinced me it will be fine to do OPTION 2] 
     
• Also, we run the plant with 20% LPY damping with the estimator already. It makes sense that we measure the exact plant that will be available when the estimator is running. Plus, It will be cleaner, only one set of measurements per estimator fit.
• If we wish to add a L to Y plant, the procedure is lengthier in the OPTION 2 (20% LPY) case, we have to add fits for L2Y for all stages measured. In OPTION 1, we can skip adding the M1 stage, as the damping is already folded into the measurements.
  • However, for suspensions with ISC drives, we risk double-counting the effects of the Suspoint motion into the ISC drives with OPTION 1. SInce the ISC drives were not included in the original measurement.