Yesterday, I took transfer function measurements on ZM4 in HAM7 and found several issues related to cross coupling (between Pitch and Yaw DoF) - please see the results from yesterday attached below in the pdf file.
After discussing this issue with Betsy and Lee, today I inspected ZM4 in HAM7 chamber to rule out any rubbing or touching of EQ stops (and it looked all fine). Next, I inspected the PSAMS cable (which is connected from bottom mass to the Intermediate mass and then to the Structure as shown here). I found that the cable at the intermediate mass is a little bit under tension, hence I gave it some relief and then re-took the TF (this took few iterations between adjustment and measurements). I found a strong co-relation between the cable routing and the cross coupling issue in ZM4. The plot here shows comparison (red and orange traces are the measurements, black is for reference)) between damping ON and OFF for Y Dof. When damping is OFF, cross coupling is still there but it looks much better from the initial results. This is further improved when damping is switched ON (although this is not a normal procedure) while taking the TF. Also attaching a plot which gives a comparison of all 3 scenarios for Y Dof (blue trace - strong coupling damping OFF, red trace - cable adjusted and damping is ON, orange trace - cable adjusted damping OFF, black is reference line).
Although, some amount of cross coupling is still observed in Pitch DoF (for both damping ON and OFF measurements). Perhaps the cable is still stressed and is pushing the IM, hence on Monday I will re-adjust it once again and then re-take these measurements.
The TF templates are stored at the following location,
cd /ligo/svncommon/SusSVN/sus/trunk/HXDS/H1/ZM4/SAGM1/Data/
2021-10-20_1400_H1SUSZM4_M1_WhiteNoise_L_0p01to50Hz.xml
2021-10-20_1400_H1SUSZM4_M1_WhiteNoise_P_0p01to50Hz.xml
2021-10-20_1400_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
2021-10-20_1450_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
2021-10-21_1400_H1SUSZM4_M1_WhiteNoise_L_0p01to50Hz.xml
2021-10-21_1400_H1SUSZM4_M1_WhiteNoise_P_0p01to50Hz.xml
2021-10-21_1400_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
2021-10-21_1405_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
2021-10-21_1410_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
2021-10-21_1415_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
2021-10-21_1420_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
I can also confirm (looking by eye) that the pitch and yaw for both intermediate and bottom mass looks fine on ZM4. I forgot to measure the height of the bottom mass (after adding add-on mass during alignment with VOPO team) - hence will measure it and post the numbers later on.
J. Kissel
Heading back to this data as I start to export it and add it to the collection of measurements for matlab processing.
In doing so, I found I needed a bit more translation of the many yaw data sets that Rahul gathered on 2021-10-21, i.e.
/ligo/svncommon/SusSVN/sus/trunk/HXDS/H1/ZM4/SAGM1/Data/
2021-10-21_1400_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
2021-10-21_1405_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
2021-10-21_1410_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
2021-10-21_1415_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
2021-10-21_1420_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml
Thankfully he saved most of these measurements with *filename* *labeled* 5 minutes apart in the last template, labeled 2021-10-21_1420, and the measurement time for each of the reference traces helped me better understand. As such, I've recreated his ZM4_Yaw_all_measurements.png where he compares several different yaw traces, but I've trended the TEST bank's yaw excitation channel and the DAMP bank's output for all DOFs to confirm which data sets had damping ON vs. OFF.
In my re-creation, I've arranged them in chronological order, put the actual measurement time and the filename's reported measurement time in the legend, as well as indicating whether damping loops are ON or OFF and the different cabling configurations.
Note that I do not show the 2021-10-21_1410 data taken at 2021-Oct-22 18:11 UTC, the damped version of the "1405" cable arrangement #2 configuration simply because that reference wasn't captured.
Also note that these damping loops are still very crude damping loops which are copied over from the old HTTS design, without any effort to adjust the gain or loop shape, so it doesn't suprise me that comparing damping ON vs. OFF does not do much for yaw.
Since it is useful to save what the transfer functions look like during the worst example of the PZT cable rubbing, and Rahul later has found a configuration that he likes (see the complete 2021-10-25 data set in LHO aLOG 60400), I'm going to export the 2021-10-21_1400_H1SUSZM4_M1_WhiteNoise_Y_0p01to50Hz.xml data set that has corresponding L and P data in the same cable configuration.
