[Shuta Tanaka, Ide, Obayashi (Aogaku), Yuzurihara]
We performed the lockloss investigation for the lockloss which happened between 2025/06/18~06/24. The previous lockloss investigation was posted in klog34527. During this period, there were 13 lockloss from the OBSERVATION state. The plots are listed in DAC wiki.

Quick saturation of OMC DCPD

For 1 lockloss (i=37, 2025-06-18 01:29:31.0625 JST), the quick saturation of OMC DCPD happened. This phenomena looks similar with the lockloss reported in klog34527. Compared with the usual saturation of OMC DCPD, the quickness (or the time scale) of the saturation is different. Figure 1 shows the timeseries.

Slowly saturation of OMC DCPD

For 1 lockloss (i=45 2025-06-22 14:35:29.0625 JST), the slowly saturation of OMC DCPD happened. This phenomena is same as regular report one. The oscillation just before the saturation seems to be ~18 Hz, as shown in Figure 2.

For 1 lockloss (i=49 2025-06-22 20:43:59.3125 JST), the slowly saturation of OMC DCPD happened. The oscillation just before the saturation seems to be ~37 Hz, as shown in Figure 3.

Earthquake

For 3 lockloss , the EQ label was asgned. For 1 lockloss (i=44, 2025-06-21 23:17:07.9375 JST) which was not asigned with the EQ label, we found that the earthquake happened coincidently with the lockloss, even though the EQ label was not asigned.  Figure 4 shows the timeseries around this lockloss.

Excess of 1~10Hz seismic motion

For 2 lockloss , the 1Hz label was asgned (one lockloss of them was also asignmed with EQ label.). For 1 lockloss  (i=38) which was not asigned with the 1Hz label, we found the excess of the 1Hz seismic motion. Figure 5 shows the timeseries around this lockloss.

Gradual drop of IR trans power

For 1 lockloss  (i=43), at -5 s, we can see the gradual change of CHARD and DHARD feedback. Coincidently, the IR trans powers started to decrease.
Figures 6 and 7 show the timeseries around the lockloss.

For 1 lockloss  (i=46), the gradual drop of IR trans power was observed. Figure 8 shows the timeseries around the lockloss.
We are still not sure why the gradual drops happened.

Lockloss by the human request

For 2 lockloss , the REQ label was asigned.