It seems to be because of the oscillation of NBDAMP_R (fig1).
I found that NBDAMP_R was not tuned from 300K, so I tuned it.
I didn't measure OLTF but it seems to wrk well according to the time series of the feedback signals (fig2).
It seems to be because of the oscillation of NBDAMP_R (fig1).
I found that NBDAMP_R was not tuned from 300K, so I tuned it.
I didn't measure OLTF but it seems to wrk well according to the time series of the feedback signals (fig2).
ARM : Y
TEMP ITM : 87.3
TEMP ETM : 85.8
NUM : 10
IFO REFL HWP : 160.0
PSL HWP : 30.0
IMC POWER : 7.4
VALUE : 1320.7
ERROR : 8.8
Measured data is stored to /users/Commissioning/data/Finesse/Yarm/20250302-233827
https://gwdet.icrr.u-tokyo.ac.jp/~controls/trendMon/finesse.html
The temp at EYC1F and maybe 2F increased by 0.3C because of the two more cryocooler operations.
I adjusted the Sunrise and Delonghi heaters' power around 9:00 to reduce the temp at EYC area.
The temp at EXC1F and maybe 2F increased by 0.3C because of the operation of two more cryocoolers.
Unfortunately, there is no tool for temperature reduction, even if I turned off 1 FFU. I need more reduction of FFU.
ARM : X
TEMP ITM : 87.7
TEMP ETM : 83.8
NUM : 10
IFO REFL HWP : 150.0
PSL HWP : 30.0
IMC POWER : 7.4
VALUE : 1437.0
ERROR : 6.0
Measured data is stored to /users/Commissioning/data/Finesse/Xarm/20250302-231240
Anyway, the present design of duct shield cooling has no redundancy because there is only one cryocooler for each. We need to consider to add one more cryocooler for each for the reliable commissioning and observation. Otherwise, we will keep losing so much time for the frosting trouble and its recovery.
Or can we recover the frosting on mirrors and windows by only using the IM heaters and window heaters under vacuum conditions?
The BRT4 side of the duct shield is near the mirror. In addition, the duct shield was inserted "inside" the structure of the outer shield of the cryostat itself. So the outgas from the BRT4 area can easily reach the mirror AR side.
The figure 1 shows the past data when the trouble of the cryocooler for the DUCT SHIELD BRT at EX happened.
In this event, the temperature increase continued at the cryocooler head for the DUCT SHIELD BRT, and a sudden enhancement was observed from 78K to 113K. However, the vacuum pressure enhancement seemed to start ~ 2 hours later when the BRT4 temperature reached ~ 157K.
We guess we are now facing the same event.
At that time, the cooling performance seemed to be recovered after ~ 1 day as shown in Figure 2. However, slight frosting was detected on the AR surface of the mirror and windows. Also the temp enhancement at the head of the cryocooler could be the trigger for the cooling performance recovery as observed in the cryocooler for the DUCT SHIELD ARM in EY.
Although we turned on the cryocoolers for the payload, the temp at DUCT SHIELD BRT1/4 kept gradually increasing, according to the increase of the temp at the 50K BRT HEAD. On the other hand DUCT SHIELD ARM4 started slightly decreasing due to the radiation reduction from the 80K shield.
Actually, we have no way to stop the temperature increase of the 50K REFBRT HEAD.
The final trial is to stop the cryocooler for the DUSCT SHIELD BRT until its HEAD temperature reaches 100K, then to restart. This method was tried for the cryocooler for the DUCT SHIELD ARM in "EY", then the cooling performance was recovered and reported as klog#32861.
Because the temp at DUCT SHIELD BRT4 is approaching the critical temp of 157K, we should perform this in the morning of March 3rd in parallel with the OMC vacuum area recovery.
----
Miyoki, Uchiyama, Ushiba, and Kimura discussed the cryocooler's troubles for the DUCT SHIELD ARM in EY and BRT in EX on Saturday and Sunday morning, then we concluded these strategies, considering trying to take any "renewed" sensitivity data using 2 DCPDs and 10W laser power as early as possible.
The cooling performance for the ARM side duct shield seemed to be recovered. Then the temps at Duct SHIELD ARM1/4 started decreasing before the critical temp of ~ 157K. The timing of the temp reverse at ARM4 seemed to be delayed by one day from ARM1.
No serious vacuum pressure level enhancement at the GV_eya position, etc has not been detected up to now.
If a common outlet is used:
A common outlet was used for an AC source of IO chassis, an function generator for LVDT, etc. in the past and it was stably enough for driving them. If these power suppliers are particularly weak compared to IO chassis or function generators, then we should prepare a recovery method in addition to improvement of environment (only what we can do is installing an UPS? But from the view point of noise in DGS rack, UPS had better to be installed to a rack outside of VIS room). And also I wonder another devices connected to this common outlet (if remains) detected a voltage drop or not.
If dedicated one is used:
An environment should be already much better than the common outlet. In this case what we can do for improving is less than the case of using the common outlet.
That should be a severe case...
> Because the power suppliers for EY IM heater seemed to be in error mode, I restarted them.
This error is caused by a voltage drop in the input AC power supply.
Since a minimum of 85 Vac is available, this means that the voltage has dropped below this level.
I recommend that you first check the quality of the power supply and the environment.
Slides used in a commissioning meeting for this topic: JGW-T2516543
So I recovered it with following the recovery manual.
----
By the way, some system services are not running on camera servers and also they cannot be updated because of expiration. It might be a good time to consider the system upgrade.
We found that one of the cryocooler's cooling performances was so strange.
Fig.1 shows the REF4 4K/50K HEAD cooling performance is not smooth,c compared with REF2 4K/50K HEAD.
I, Ushiba-kun, and Kimura-san discussed it, and we decided to monitor it until the next morning.
Around 17:30, two EY Payload Cryocoolers started.
Because the power suppliers for EY IM heater seemed to be in error mode, I restarted them. Ushiba-kun checked the healthy status of the power supplier for the IM heater. According to Ikeda-san, the error reset seemed to be possible via LAN. So, we asked him to prepare the reset function for all power suppliers for all Type-A suspensions.
We found that one of the cryocooler's cooling performances was so strange.
Fig.1 shows the REF4 4K/50K HEAD cooling performance is not smooth,c compared with REF2 4K/50K HEAD.
I, Ushiba-kun, and Kimura-san discussed it, and we decided to monitor it until the next morning.
> Because the power suppliers for EY IM heater seemed to be in error mode, I restarted them.
This error is caused by a voltage drop in the input AC power supply.
Since a minimum of 85 Vac is available, this means that the voltage has dropped below this level.
I recommend that you first check the quality of the power supply and the environment.
That should be a severe case...
If a common outlet is used:
A common outlet was used for an AC source of IO chassis, an function generator for LVDT, etc. in the past and it was stably enough for driving them. If these power suppliers are particularly weak compared to IO chassis or function generators, then we should prepare a recovery method in addition to improvement of environment (only what we can do is installing an UPS? But from the view point of noise in DGS rack, UPS had better to be installed to a rack outside of VIS room). And also I wonder another devices connected to this common outlet (if remains) detected a voltage drop or not.
If dedicated one is used:
An environment should be already much better than the common outlet. In this case what we can do for improving is less than the case of using the common outlet.
The cooling performance for the ARM side duct shield seemed to be recovered. Then the temps at Duct SHIELD ARM1/4 started decreasing before the critical temp of ~ 157K. The timing of the temp reverse at ARM4 seemed to be delayed by one day from ARM1.
No serious vacuum pressure level enhancement at the GV_eya position, etc has not been detected up to now.
The temp at EYC1F and maybe 2F increased by 0.3C because of the two more cryocooler operations.
I adjusted the Sunrise and Delonghi heaters' power around 9:00 to reduce the temp at EYC area.
There is no dependency updates, so it's no impact for another packages used in daily commissioning works on workstations.
I haven't checked stability of latest zoon on Debian. But even if it's not so stable, it's better than one which cannot be launched.
[Kimura, Yasui, Nakagaki]
We installed an automatic closing device on the IFI large gate valve.
The vacuum gauge and GV controller are not yet connected.
Connection and operation test will be done separately.
Kimura, Yasui
They closed 5 flanges of the OMC chamber.
The 5 flanges are 2-2, 2-3, 2-4, 2-6, and 2-8.
We used a new window flange at 2-3.
We put metal covers on the window flanges (2-2 and 2-3).
Uchiyama
I started pumping down OMC and OMMT by a dry pump from 12:34. Since the vacuum pressure reached 20 Pa at 14:30, I started TMP at SRM, and the TMP was normal at 14:50.
We studied the lock duration of the locked events from 2025-01-19 UTC to 2025-02-23 UTC. A total of 104 events were divided into two groups: 22 events before 2025-02-04 and 82 events after that date. A frequency distribution of lock durations was created for each group Additionally, only the events during 10 W operation were extracted to generate a separate frequency distribution. The tables and their corresponding graphs are attached. Pie charts were created to visualize the proportions of each category. No indication was found that event duration were shorter during 10 W operation compared to 1 W one. Before 2025-02-04, half of the events had durations of 1,024 sec. or less, but after that date, this proportion decreased to about 40%.
Pictures: link
