J. Garcia, J. Kissel, S. Steplewski

Tuesday's measurements showed severe cross coupling on the Main Chain, and only two vertical modes (where there should be three) on the reaction chain. Brett and Mark subsequently modeled that the R0 UIM and R0 PUM were moving as one, and Travis went in and found (a) that the M0 BOSEMs were all wonky, and (2) that the R0 UIM blade stops were locked. 

Since then, Travis has unlocked the R0 UIM blades, and re-aligned the M0 BOSEMs, and Jeff has remeasured all TFs. The results are attached. In summary, M0 (though the measurement was taken with damping loops ON*) looks good**; R0 still has some rubbing (likely at or near the top stage*** -- maybe the UIM blade stop was backed off too far and is contacting TOP to UIM SUS wires?), and its cables appear to be adding stiffness****.

Jeff, Szymon, and Travis are going in to investigate further as I write the aLOG update...  

Details / Further Explanation
-----------------------------
* The M0 TFs were taken with damping loops ON, simply because of a failed automated hand off between HEPI measurements and SUS Measurements. Nominally, we want to set up a system similar to the BSC-ISI and SUS, where there's an EPICs variable reserved for indicating "measurement ON" and "measurement OFF" that automated Matlab TF scripts would turn on and off appropriately. Unfortunately, for the HEPI / QUAD trade off that Vincent / Garcia tried last night, either the variable wasn't there yet, HEPI's automated scripts were set up to turn the EPICs variable or the SUS damping loops on/off, the QUAD script wasn't watching it to take the appropriate action, or something (I couldn't quite tell what happened from Garcia's description) because things were set up a little bit too quickly. So, the damping loops for the QUAD (required during the HEPI TF) were left ON after the HEPI TF had finished during the M0 TFs, but were turned OFF by the automated script by the time R0 TFs rolled around. No big deal, there's a first time for everything with these automated scripts, and it never works out the first time. Garcia's gunna make sure that either he gets the chamber for the evening, or the automation works better this time.

** Looking at pages 1 through 6 of the attachment comparing current to previous should-be-the-same measurements, you can see 
- If you by-eye undamp the transfer functions, you can see that "everything seems to be in the right place" compared with the previous measurements
- The cross coupling that was in the 2012-05-07 measurement looks to be gone
- We're inconsistently damping each degree of freedom; some way overdamped (P), and some poorly damped (e.g. L)
So, we say we think the dynamics look good from this eyeballing, but we know that damping signifcantly reduced cross-coupling, so we really need an undamped measurement to really confirm the dynamics are OK.

*** Pgs 7 to 12 show the reaction chain. As far as "rubbing at the top stage" I see that the data for low frequency modes of R, T, and V are low Q and noisy which we've found from experience usually means rubbing. Because the rubbing appears to be affecting only the low frequency modes, we know that the rubbing should be localized to the upper stages. (Remember: for low frequency modes, the QUAD masses are swinging as one, so the "pivot" for these mode will be high up in the suspension; for high frequency modes, the dynamics are more complicated, involving differential motion between the masses, so the dynamics involve the lower stages).

**** Pg 11 shows the P to P TF, and specifically that the large, second Pitch mode is stiffer that previous measurements. From past experience, we know that this mode is easily affected by cable stiffness (the higher the frequency of the mode, the stiff the cables are). We also know from past experience that we can arrange the cables in such a way that we can get the Pitch TF to be consistent between suspensions. We also also know that something has happened to the cable arrangement between the 2012-05-07 measurement and last night's measurement that has stiffened up the cabling. So, we should look at the cable arrangement again.