YamamotoT, Darkhan,
We took a set of DARM transfer function measurements on the April 2-3 calibration day.
Actuation levels of DARM, XARM, YARM excitations were adjusted (lowered) to avoid saturating the TM coil driver actuators. Actuation levels of PCALX and PCALY excitations were lowered to avoid breaking the OFS loops.
The anaysis of the measurements will be posted at a later time.
Picture with ITM TCam image
The initial alignment with which we can obtain highest transmitting power for both arms started decreasing just after the locking and seemed to go some constant level as attached files.
ETMY Yaw drift was about 5 micro radian.
ETMY Pitch drift was about 3 micro radian.
SR3 Yaw drift was about 2 micro radian.
PR3 Yaw drift was about 2 micro radian.
Around 5:30 (TMSXY values are max), just after locking about 460 kpc was indicated.
Around 7:15 (TMSX values are about 105/125 for X, 85/100 for Y, compared with the max values just after locking), 595kpc was indicated, however, we should consider the reduction of the optical gain.
Picture with ITM TCam image
The ADS yaw loop was closed. Dither frequency is 4.6 Hz for ETMX and 11.3 Hz for ETMY. It seems to be working well.
The ISCWD is working well. The functionality of this watchdog is as follows:
- This watchdog is monitoring the input to the ISC INF from the LSC block and the ASC block including the ADS and BPC. If these signals get beyond the threshold, immediately the feedback signal of the ISC will be held by the new RAMPHOLD block.
- If the reset pulse is sent, the RAMPHOLD block will release the output with the ramp time as shown in klog13937
- The VIS guardian will turn off the ISCINF just after the watchdog tripped, and reset the watchdog.
- It working well, and the kick of the suspension got small thanks for this WD.
A beam dumper for a beam reflected on QPD was also put btw KG5 and a window (IR filter).
Nakano, Miyo
We add ISCRAMPHOLD block in Type-A suspensions.
We modified TM, IM, and MN master model of Type-A in userapps/vis/common/models .
The new block is added in userapps/sys/common/models/KAGRA_UTIL.mdl.
Now, SVN version is 2937.
Detail information about this modification will poseted by Nakano-san later.
The ISCWD is working well. The functionality of this watchdog is as follows:
- This watchdog is monitoring the input to the ISC INF from the LSC block and the ASC block including the ADS and BPC. If these signals get beyond the threshold, immediately the feedback signal of the ISC will be held by the new RAMPHOLD block.
- If the reset pulse is sent, the RAMPHOLD block will release the output with the ramp time as shown in klog13937
- The VIS guardian will turn off the ISCINF just after the watchdog tripped, and reset the watchdog.
- It working well, and the kick of the suspension got small thanks for this WD.
[Akutsu, Nakano, Yamamoto, Ushiba]
We modified the alignment recording guardian.
If we requested PRFPMI_RECORDING to ALIGNMENT_RECODER guardian, IMMT2, all PRs, all test masses, SR3, and SR2 good OpLev value is recorded only when alignment is good.
Restore OL set point, OFFLOAD_PRFPMI state, also implemented but not tested yet.
I will test it tomorrow morning.
[Lucia, Ushiba]
In this morning, we tried to close the ADS loop for YAW of ETMX and ETMY.
First, we saw the signals of OpLevs, TRX, and TRY.
Then, we found that ETMY YAW excitation excited ETMY PIT at the same time and PIT motion at this frequency is larger than YAW.
So, we started from optimization of drive align matrix.
However, even though we scaned enogh large range for Y2P matrix value, we couldn't find good values for diagonalization.
We don't know exact reason but maybe due to the resonance or damping filter, the phase of between PIT and YAW motion is not 0 or 180 degrees.
So, we gave up drive align matrix optimization.
What we can do is as follows:
1.Using higher frequenies for dither excitation.
2.Using TM OpLev set point to excite the Yaw motion.
We tok over he work to Nakano-kun.
I changed the FM1 so that the single arm TRANS got 1 at the FB output when the main laser power is 3.6 W.
non
[Miyoki, Akutsu, Kokeyama]
Because of the VIS offload campaign in daytime, the alignment was completely off for Y arm. This log may be useful when you completely lose an arm alignment.
- Last night X arm and Vertex were restored without any problem. Y arm was completely lost.
- For Y arm, the cavity flash did not recovered even oplev values restored for SR3, IY, and EY. First, to find the GR flash, we used IY telephoto camera to find the single spot from SR3.
- Note that EY camera was not useful for the initial aliugnment: it was too slow and too dark to see a single bounce. The time stamp is not right to the shown image, either.
- Then, align EY so that the reflection spot from EY can be found on the IY camera.
- Adjust EY, IY, SR3 by looking at the GR TR level and also the spot positions on the IY camera. EY camera can be used when flashes are good.
- Request Y arm guardian to "IR LOCK." It should lock IR.
- Align IR beam by BS, looking at the IR TRANS level.
- Check the spot positions for IR. Also again GR spots.
- Repeat the procedure if necessary.
After this, Yokozawa-san and Yuzu-san kindly took over (for the initial alignment lecture) at 3am.
non
Here are cartoon summaries:
Attach 1 (top) - When the initial installation (probably)
Attach 1 (bottom) - On Apr 1, 2020, before the reapir (while entire commissioning so far)
Attach 2 - After repair.
Yokozawa-Sama fixed it.
I analyzed the measured data.
The results are following:
https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=11587
Betelgeuse V band
YamamotoT, TomaruT, Darkhan,
Today (April 1) we measured the TM stage coil driver transfer functions at the X-End and Y-End.
Transfer functions of four channels of the TM chassis were measured in three different states:
- no low-pass filters (LP) switched on
- only the first stage LP turned on
- only the third stage LP turned on
When analyzing the TFs the corresponding reference measurement must be divided out. Reference TFs were measured by connecting the output from the BNC-to-DB9 interface board to the input of DB9-to-BNC interface board. So, reference TFs take into account the 2x gain factor and possible non-flat responses of the interface boards and the spectrum analyzer.
Measurement info and data format:
Excitation level : 0.5 V
Number of averages: 5
Frequency vector : same for all measurements (in files *.X): log-spaced 0.1 Hz - 51.2 kHz
States:
State 0 (CH#ST0A.TXT, CH#ST0P.TXT, CH#ST0V1.TXT, CH#ST0V2.TXT): all low-pass filters (LP) switched off
State 2 (CH#ST2A.TXT, etc.): only the first stage LP is turned on
State 6 (CH#ST6A.TXT, etc.): only the third stage LP is turned on
Data format:
Magnitude (REF#A.TXT and CH#ST0A.TXT) : dbMag
Phase (REF#P.TXT and CH#ST0P.TXT) : degrees
CH1 (input) Norm Var (REF#V1.TXT and CH#ST0V1.TXT) : LinMag
CH2 (output) Norm Var (REF#V2.TXT and CH#ST0V2.TXT) : LinMag
Measurement data was saved in
- JGW-T2011585 - ETMY TM coil driver TF
- JGW-T2011586 - ETMX TM coil driver TF
I analyzed the measured data.
The results are following:
https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=11587
[Akutsu, Miyoki, Kokeyama]
50/50 BS was placed in front of the TMSX IR PD. The power was reduced for a factor of 0.43.
In MEDM, which gain to change, FM1 (name "1.7W") or GAIN (now 3.16) of K1:AOS-TMSX_IR_PDA1??
I changed the FM1 so that the single arm TRANS got 1 at the FB output when the main laser power is 3.6 W.
Related to 13942; a QPD assembly for IMMT2T (IR transmission from IMMT2) was detached and stocked in a black strage for AOS located between the PR3 chamber and PR2 chamber.
By the way, the QPD circuit was not fully fixed to the support structure!! The four M4 screws to fix the QPD circuit were loosen somehow... OMG!
