I tried to engage WFSf3 control to IMMT2 PIT and YAW, these could be installed these. But these are need some offsets.
Detail:
- First, I considered an ASC scheme using WFSf3 with reference to klog31196. Although the QPD3 RF11 signal has a large coupling with CHARD and other factors, I thought that it could be suppressed because these degrees of freedom are currently controlled by PRFPMI ASC. The IMMT2 signal is orthogonal to the other degrees of freedom. Also, when moved to IMMT2 in the case of klog31196, it is clearly visible as the QPD3 RF11Q signal, so we decided to feed back it to IMMT2.
- The demodulation phase of IMMT2YAW in QPD 3 RF11 is 105deg. (Calculated from file /users/Commissioning/data/ASC/2024/sensingmatrix/0910/{IMMT2data}.) However, when I rotated it 105deg and measured, the values of the I and Q signals were slightly close, so when I rotated it to -10deg (95deg), the I signal was 1.06 and the Q signal was 0.4, and they were separated. (Demodulation phase before: FIG1, after: FIG2)
- Next, the OPLEV/QPD3 RF11I signal was measured and input into the input matrix to be the same as the units between the INP1 error signal and OPLEV. (before:FIG3, after:FIG4)
- Next, I measured the PLANT. IMMT2 has poles at 8.1Hz and 1.8Hz. They were measured correctly.(PIT:FIG5, YAW:FIG6)
- I set the control filter so that UGF is 0.1Hz.Also, I put -1 in the GAIN on the MEDM screen for negative feedback. In the case of other ASCs, when the signal is merged with the OPLEV control, a gain of -1 is added, so the ASC does not need a negative feedback of -1. This time, it is returned directly to IMMT2TM, but to distinguish it from that case, I set -1 to GAIN.(FIG7)
- The control is closed. The control is stable. However, POP RF90 has dropped slightly compared to IMMT2 control of klog31196.(FIG8)
- To investigate the offset, we measured whether the error signal changed when we applied an offset to the PRCL, just like in klog30877. However, the error signal did not seem to change much.(FIG9)
The reasons for this offset are as follows:- RFAM → In that case, if the PSL temperature is stable, the offset may also be stable.
PRCL offset origin → As mentioned above, it seems to be quite low. If this is resolved, we will need to install an RF90centering loop.
- RFAM → In that case, if the PSL temperature is stable, the offset may also be stable.
- I added an offset to the error signal based on when IMMT2 control in klog31196 is on.
- I measured the OLTF. Because it takes time to measure, I was unable to measure the UGF at Gain 0, but both PIT and YAW control have a UGF of 0.1Hz or less, and a phase margin of 40deg or more. PIT: FIG10, YAW:FIG11
- We also measured the power spectrum. (FIG12) The IMMT2 control in klog31196 is the INP2 filter, and this is the reference data. The filter this time is INP1. The power spectrum of the ARM TRANS, POP RF90, and error signal is almost the same or slightly better in some areas. However, when measured powerspec with this control, there was coherence in the ARM TRANS and POP RF90, perhaps because the alignment was not good. And the IMMT2 control in klog31196 seems to be around UGF 0.05Hz, and since the UGF is different from this time, a strict comparison could not be made.
Future Plan：
- It is necessary to investigate whether the offset of the error signal is repeatable. Since the transmition power did not decrease during the two hours of the work after the offset was set once, the offset is not likely to change over that span.
