I checked the drift of the GAS output for the Type-A and Type-B suspension.
Type-A, ITMX F3 and ITMY F2 and F3 drift is larger.
Type-B BS GAS drift larger. (Maybe, the temperature drift is larger around BS area.)
- The demodulation phase for the CARM signal (RF45) rotated by 90 degree. I cabled the Q-MON of the I/Q demod board to the CARM servo. This is because the limitation of the rotation angle of the analog phase shifter as reported in klog12668.
- Now the CARM loop UGF is 50 kHz, and seems much more stable. I left the IFO locked, and we will see the long term stability.
- I also tried to transit 3f MICH signal to 1f, but it still cannot be stable enough. Actually the RF17Q did not have coherence with RF51Q. I will try again tomorrow.
- DARM OLG was measured again.
[Kamiizumi, Shimode, Tomura]
Shimode-san modified one board (S1706801) of the LVDT Combiner (S1706809).
Coils for Ch.4 and Ch.6 were laid down to reduce the crosstalks from adjacent channels.
This chassis was installed to EXV2 rack as a replacement and former LVDT Combiner for ETMX (S1706805) was brought back to Mozumi.
I heat up a little bit in the corner station, however not enough at all. we need to put more heater in the corner station.
(Corner)
Delonghi Heater
C-HT01:25C/L4(~1200W) -> L5 (~1500W)
C-HT02:26C/L4(~1200W) -> L5 (~1500W)
C-HT03:26C/L5(~1500W)
SUNRISE Heater
C-HT04:26C/L3 (~1200W)
C-HT05:26C/L2 (~700W)
[Miyakawa, Kamiizumi, Tomura, Shimode, Okutomi]
-
We replaced the LVDT combiner for the ETMX BF LVDTs to a modified one.
- The modified chassis has rotated inductors by 90 deg on one of the two boards.
- This modified board is assigned to the horizontal LVDTs, while the verticals have the initial board-mounting configuration.
-
We measured the transfer function in the longitudinal DoF of the BF.
- The result was similar to the previous measurement, nothing improved. (see the figure below)
- It means that the inter-channel couplings in the LVDT combiner was not the main reason of the spurious coupling at higher frequencies.
-
I suggest some scenarios to explain this spurious coupling.
- The main coupling path could be the LVDT-actuator unit itself where the receiver coil of sensing and the actuator coil are shared.
- The mechanical transfer function has poor overall gain due to the position the actuation force applies on.
-
Geometrical configuration of the sensing and actuation coils can cause the coupling.
- Because the IP and GAS coils also have the coupling with similar characteristics even without sharing the sensing and actuation coils.
-
A better way to understand this spurious coupling is to measure the pure actuator-to-sensor coupling and investigate the setup to reduce it.
- A prototype BF LVDT unit (could be in NAOJ Mitaka) sounds useful to test it.
I additionaly checked ETMX IM damping channel and they are affected by PCal lines.
More information is available here.
https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=11426
Following link is list of channels which is affected by DARM DoF. It maybe better to avoid to use these channel in noise source study.
https://github.com/gw-detchar/tools/blob/master/GlitchPlot/Script/DARMaffected.dat
Attached are the current OPLTFs of the OMC QPD centering servo.
Yaw loops are OK.
The DOF1 Pitch is marginally stable because I added more agressive roll-off filters only to this loop.
The DOF2 Pitch is more problematic as the notch at 9Hz is deep. Therefore, I had to increase the UGF to 50Hz.
[YokozaWashimi]
For the check of the air compressor with new configuration, we placed two microphones and one accelerometer.
K1:PEM-SENSOR_RACK_IX0_BNC2 : Microphone around IX0 rack
K1:PEM-SENSOR_RACK_IX0_BNC3 : Microphone around the air compressor
K1:PEM-SENSOR_RACK_IX0_BNC4 : Accelerometer on the air compressor
[Yokozawa, Washimi]
We removed the IXC seismometer for R&D
The temperature drift was small, and GAS output drift would not so large drifting, I decided to stop the vacuum pumps during mid night.
If you find the serious temperature drift, please let me know.
Hi all,
My original thought was the following:
- as the DAC output is finally output by a Dsub-BNC converter, it will be useful if the output range (seen in MEDM) is from -5 to +5 V.
- as the analog phase shifter only receives positive voltage, it is safer to set limits.
- though it is not very useful, I thought it is fine if we can shift the phase by 90 deg. at max.
Two hour trend of the ondotori and Type-A and Type-B GAS filters
0.75Hz peak in IYC OpLev is not observed in IYC infrasound microphone (red)
We may need some modifications on analog part
because DAC output range is not enough for changing from 0deg. to 360deg.
- Current setting of demod. phase is 8.362 Volts (see fig1)
- We can set demod. phase as between 0V and 5V (see fig2)
- We can do as between -5V and 5V by removing saturation block from k1alspll.
- We need to set as between 0V to 12V for changing from 0deg. to 360deg.
For evaluating the drift of the temperature, we stopped the turbo and dry pumps of the 3,5,6,13 (pumps between BS and IXC/IYC)
10:05 started the turned off the turbo pumps
10:20 stopped the dry pumps and chillers
Two hour trend of the ondotori and Type-A and Type-B GAS filters
The temperature drift was small, and GAS output drift would not so large drifting, I decided to stop the vacuum pumps during mid night.
If you find the serious temperature drift, please let me know.
I heat up a little bit in the corner station, however not enough at all. we need to put more heater in the corner station.
(Corner)
Delonghi Heater
C-HT01:25C/L4(~1200W) -> L5 (~1500W)
C-HT02:26C/L4(~1200W) -> L5 (~1500W)
C-HT03:26C/L5(~1500W)
SUNRISE Heater
C-HT04:26C/L3 (~1200W)
C-HT05:26C/L2 (~700W)
I checked the drift of the GAS output for the Type-A and Type-B suspension.
Type-A, ITMX F3 and ITMY F2 and F3 drift is larger.
Type-B BS GAS drift larger. (Maybe, the temperature drift is larger around BS area.)
I modified the k1omc model so that the QPD signal is sent to k1asc via Dolphin from this model, rather than from k1aso.
To do so, I removed IPC blocks named K1:OMC_ASC_QPDV{1,2}_{PIT,YAW} from the k1aso model and added them to the k1omc model.
I recompiled and installed k1omc, k1aso and k1asc. Then rebooted them and k1dc0.
The new k1omc model uses a modified QPD block, with which users can specify the the saturation thresholds of the sum output with EPICS PVs.
Details of this QPD block can be found in 12665.
I created a modified QPD block called QPD_WITH_SATURATION_CTRL, which is available in /opt/rtcds/userapps/trunk/isc/common/model/QPD.mdl.
This block has a custom made saturation block after the SUM output, before dividing the Pitch/Yaw signals with SUM.
The saturation thresholds can be specified through EPICS PVs. This EPICS controllable saturation block is available in
/opt/rtcds/userapps/trunk/isc/common/models/C_Code.mdl
With help from YamaT-san
OMC guardian ASC part has been commissioned. It closes DOF1 and DOF2 loops to center the OMC input QPD loops by actuating OMMT2 and OSTM. We still (absolutely) need the pico motor centering in advance because the ranges are not enough large. Since the sensors/actuators are not well diagonalized (or I took wrong numbers from Aso-san), all the four loops must be closed to center the two QPDs.
Next is the LSC part.
Finally, I could recover without the Dolphin glitch and applied burt snapshot at 16:40 for all related models.
= What I did =
- Dolphin for k1lsc0 was disabled. (Dolphin for k1als0 was accidentally disabled at the same time.)
- k1lsc0 was reset from the BMC web interface.
- k1als0 was rebooted.
- Burt snapshots were applied for all related models.
Dolphin glitch can be avoid by using reset on the BMC web interface even when the SSH connection is not reachable.