Naoki, Vicky, Daniel
We worked on PI damping of 10428 Hz mode. We used ETMY ESD for the damping. The bottom plot of the attached dtt shows the OMC DCPD spectra with (red) and without (green) PI damping. The 10428 Hz peak is reduced by a factor of ~6.
Detail:
Since we have been having troubles due to PI ring up around 10400 Hz, we worked on PI damping. As shown in the attached figure, the significant PI modes are 10424, 10425, 10428, 10431 Hz. We tried PI damping of 10428 and 10431 Hz today.
We used the OMC DCPD channel 7 for the error signal, which is down converted by 10kHz. The down converted signal is sent to mode24 for 10431 Hz damping, and mode31 for 10428 Hz damping.
For mode31, we made a new bandpass filter named 10428. The bandpass filter for 10428 Hz has the same function as the original 10430 bandpass filter, but we adjusted the frequency of the bandpass to pass the 10428 Hz mode and cut the 10425 and 10430 Hz modes.
Then we adjusted the gain and phase. The gain and phase setting of PI damping for 10428Hz mode is as follows.
BP gain (H1:SUS-PI_PROC_COMPUTE_MODE31_BP_GAIN): 10000
damp gain (H1:SUS-PI_PROC_COMPUTE_MODE31_DAMP_GAIN): 1000
PLL phase (H1:SUS-PI_PROC_COMPUTE_MODE31_PLL_PHASE): -45
We used the channel 7 of ETMY for up conversion by 10kHz. With these setting, the 10428 Hz peak could be damped as shown in the attached figure.
For 10431 Hz mode, Daniel could ring up this mode by actuating on ETMY ESD. So both of 10428 Hz and 10431 Hz modes come from ETMY. The damping of 10431 Hz mode is on going.
I tested this damping again and confirmed that the 10428.1 Hz peak can be damped by a factor of 5 as shown in Fig. 1.
The dtt template is stored in userapps/sus/h1/templates/diaggui/PI_mode_monitor_10428.xml.
I also noticed that this damping somehow damps also the 10431.4 Hz peak by a factor of 4. Fig. 2 shows a bandpass filter for 10428 Hz damping. The gain of this bandpass filter for 10428 Hz is 30dB while the gain for 10431 Hz is -22dB, so this damping should not affect the 10431 Hz peak.
FYI, here are some alogs about PI ring up I could find.
10.4276 kHz: alog64440
10.4314 kHz: alog67327
It seems that the 10431 Hz peak changes a lot by itself. The 10431 Hz seems reduced in the second test, but it is not in the first test. So it is difficult to tell whether the 10431 Hz peak changes by itself or by the 10428 Hz damping.