We went to the Yend for the power cycle of the coil driver, and it seemed to get recovered. However, when I started the IFO work, I found that still, ALS DARM was difficult to be locked. The first figure shows the glitch like oscillation happened when the signal is fed back to the suspension. I started to investigate it again.
The inspection details are as follows:
- The second figure shows the ETMY response when we went to the Yend, oplev yaw signal (purple) with COILOUTF signal for H1 (green) and H2 (cyan).
- We turned the coil driver off between -10 min and -8 min. The first excitation after the power cycle (at around -8 min) kicked the ETMY in yaw, but the second (-5 min) and the third (-2 min) did not kick. So we thought it was fixed.
- As shown in the first figure, the ETMY suddenly started to move in yaw without any strange signal when I tried to lock the ALS DARM.
- Then, I tried to put the excitation signal into the ETMY again, and the ETMY was kicked. The third figure shows the time-series of the same signal as in the second figure. ~20 seconds after I started to excite, the ETMY was kicked. However, the different point from before, the oscillation got small after a while, this seems to be a different situation from the case of klog9996. That indicates this is not the nonlinearity of the circuit but a glitch-ish kick.
- The excitation signal was injected into K1:VIS-ETMY_MN_LOCK_EXC, and the amplitude and the frequency were 10000 ~ 15000 cnts and 0.2 Hz, eventually.
- When I put larger excitation such like 30000 cnts in amplitude, the oscillation did not calm down, and once I cut it off, the oscillation got small as shown in the fourth figure. The oplev signal spectrum is shown in the fifth figure. This seems similar to the behavior we saw in klog9996.
- We need to find which is the evil, the digital system, the analog circuit, or the suspension. I could not tell which is the one so far, but I hope it is not the suspension problem
