Abstract:

Linear range of PRC2_Y seems very small and REFL_QPDA1_RF45_I signals seem no more sentive than REFL_QPDA2_RF45_I.
For stability tests, I changed PR3 ASC signals from REFL_QPDA1 to REFL_QPDA2, and then IFO could reach OBSERVATION state for a short time.
Since the large earthquakes occures, the stability should be checked later.

Detail:

Since PRC2 yaw seems a problematic DoF, I checked the signals of REFL QPDs while moving PR3.
Figure 1 shows the signals when PRFPMI was locked with ASC except for PRC2_Y with 1W.
At that time, PR3 was about 0.4 urad apart from the good value in +yaw direction (-43.2 of K1:VIS-PR3_TM_WIT_Y_DQ was good for IFO).
Since PRC2 flter sign is negative for closing loop, the error signals should be positive if the good values are -yaw direction, however the DC values of K1:ASC-PRC2_Y_IN1_DQ seems negative.
It implies that PRC2 might not measure PR3 motions well, so I moved PR3 intentionally and see the QPD signals.

Figure 2 and figure 3 show the QPD signals when PR3 moved in plus and minus yaw direction around good values, respectively.
According to the data shown in fig2 and fig3, REFL_QPDA1 signals seems less sensitive than REFL QPDA2 signals though REFL_QPDA1 signals are used for the current PRC2_Y signals.
Though I'm not so sure why REFL QPDA1 doesn't measure PR3 motion now, I anyway tried to use REFL_QPDA2 signals instead of REFL_QPDA1 signals instead.
Figure 4 shows the time-series data of POP90, PR3 OpLev, and QPD signals when PR3 yaw was controled with REFL_QPDA2_RF45 signals.

Snce IFO seems stable with PRC ASC with REFL QPDA2 RF45 signals, I requested OBSERVATION state to LSC_LOCK guardian and confirmed that IFO can reach OBSERVATION state.
Since the big earthquake happened, I cannot check the stability, so it is necessary to check it.

What I changed:

To test PRC ASC with REFL QPDA2 signals, I modified following.
1. K1:ASC-INMTRX_Y_3_18: from 1 to 0.
2. K1:ASC-INMTRX_Y_3_11: from 0 to 1
3. K1:ASC-OUTMTRX_Y_1_3: from 1 to 0
4. K1:ASC-OUTMTRX_Y_3_3: from 0 to 1.
5. Turn off INPUT of K1:ASC-PRC2_Y filter bank.
6. Setting gain filters (-3dB, -6dB, -12dB, and -24dB) at K1:ASC-PRC1_Y filter bank (fig5).

Note:

To test the PRC ASC with REFL QPDA2 RF45 signals, it can be engaged just turning off FM8 (null) at K1:ASC-PRC1_Y filter bank after reaching PRFPMI_RF_LOCKED state.
Since I didn't modified the guardian, resetting ASC signals and turn on/off ASC should be manualy done at this moment.
No check of UGF, so loop gain should be checked and adjusted.