Note:

According to check all PT100 in PSL room, some temperature monitors detected a temperature change.

Figure 1 shows the state number PLL guardians (top panels), temperature feedback of PLLs (the 2nd top panels) and 6 temperature monitors. t-cursor represents a time when we recovered PLL lock (klog#30266) after finishing the stress test of DAQ (klog#30267). Though PT05 which was located near the HPL didn't detected a change in temperature, PT01-04 detected a temperature change larger than each own RMS (each location can be seen on the attachment of klog#19826). A temperature change around that time on PT01, 03, and 04 was only 0.02degC. On the other hand, PT02 which was located near Green laser fibers detected a change as ~0.12degC.

Interested things are that 1) these temperature monitor showed local minimum values when we recovered PLL lock after the maintenance work and 2) a drift of the feedback signal to the temperature was one direction after lock recovering. I'm not sure a detailed temperature gradient on the PSL table (there might be more large change around Prometheus body) and which temperature monitor is the most sensitive to PLL stability. Anyway, we may not able to conclude that temperature is not related to this issue yet.

[Yamamoto, Ushiba]

We found the source of the TEMP BIAS jump of PLLX/Y observed yesterday.

Figure 1 shows the PLL guardian state and temperature feedback signals for PLL X/Y and main laser durng recovery work of PLL X/Y yesterday.
When lock loss happened (left T cursor), temperature feedback to the main laser was cleared by guardian (IMC guardian has a function to clear temperature feedback if the value is greater than 0.1 at DOWN state).
Due to this feedback clear, the main IR laser frequency is changed very quickly: that results in changing the good TEMP BIAS of PLL X/Y.

In terms of long-term stability, it is better to manage it at some point but I leave it as it is at this moment.