Recently, Y arm transmission power after initial alignment is lower than before, so I checked ADS signals during initial alignment.

First, I checked dithering signals can be seen in the corresponding signals like GRX trans, IRX trans, and so on.
Since, some of Type-A suspensions dithering frequencies are close to the mechanical resonance or peaks in the signals without dithering, I changed their dither frequencies.
Also, demodulation phase was measured in the same way as PRCL BPC (klog24539).

All signals during initial alignment was checked and I confirmed PRFPMI can be locked with new setting.
Since it takes long time to perform initial alignment because signal check was done step by step, so I don't know this work makes  the alignment better.
So, please use the new setup for a while and if it was not good, please restore all the SDFs of ASC_ADS (fig1 and fig2).

Abstract:

I tried to close BPC of PRCL but it didn't work well.
Probably, it is better to use IMMT1 but it is necessary to modify the real-time model to do so.

Detail:

I tested BPC at PR suspensions, especially PRM and PR3 by using PR3 and PR2, respectively.
Frequencies of excitations are temporary decided as 70, 75, 100, and 105 Hz.
Figure 1 shows the setting of BPC.

Phase of the demodulation was decided as following procedure.
1. Set bubd-pass filters at demod sig filter banks. frequency band is +/- 1 Hz from each excitation frequency.
2. Send excitation to the susoension.
3. If signals after band-pass filter is too low, intentionally misalign the mirrors to obtain good S/N ratio.
4. Measure the relative phase (phi0) between LO signals and signals after band-pass fiter like fig2.
5. Set demod phase as -phi0.

After setting demodulation phase, I tried to close the BPC loops but it didn't work well (drifting too large and lock was lost).
It probably due to the BPC and ADS are fighting each other because if we turned off ADS, BPC itself seems wirking fine.
Currently, beam spot on IMMT2 was fixed, so it is difficult to move the beam spot on PRM while keeping good alignment of PRMI.
So, feedback to IMMT1 seems necessary but real-time model modification is necessary to do so...

The reduction of the 60 Hz power-line-like peak and its harmonics might be due/thaks to ISS on, though we need to check when they disappear step by step to confirm this theory definitely. Anyway, checking each state of Guardian would be worth doing for the other questions.

Following 24187, 24529.

Summary

As already known, the spectral bumpy structure in DARM at around 117 Hz would be due to the angular jitter of the input beam. What is strange is that the corresponding bumps at IMMT1T seems only grown during PRFPMI locked, and not found when it is down. Considering the control topology, this should not happen. This post is to show the points by creating the attached figures.

Details

Fig. 1 shows projection from IMMT1T QPDs (K1:IMC-IMMT1_TRANS_QPDA{1,2}_DC_PIT_OUT_DQ) to DARM (K1:CAL-CS_PROC_DARM_DISPLACEMENT_DQ) made with the same transfer function (-4*x^2; in other words, I reused this same transfer function) introduced in 24187 in the silent run on Mar. 21 00:00:00 UTC (see 24475). In fact the S/N of the QPD output is better for QPD2 than that of QPD1, but the transfer function was of QPD1, so I plot both; these QPDs show slightly different gain seemingly. Anyway, It is obvious, at least the 117 Hz bumpy structure, the 128 Hz stable peak, and the 163 Hz bump would be related to the angular jitter of the input beam that can be detected with these IMMT1 transmission QPDs. This has been already known since 24187.

The strange thnig is related with the fact already reported in 24529. To repeat the point, Fig. 2 shows that the bumpy structures around 117 Hz and 163 Hz get larger when PRFPMI locked (or, during the silent run, of Mar. 21 00:00:00 UTC), while they are less when the interferometer is down (of Mar. 21 02:00:00 UTC) but IO Gurdian is in PROVIDING_STABLE_LIGHT. So, the question is why these bumpy structures appear in IMMT1T QPDs when the interferometer locked? Understanding the mechanism how the bumps appear in the QPDs will make DARM better, I guess.

The first scenario might be due to some coupling from MCL feedaround during PRFPMI locked. Fig. 3 shows the relevant signals, but there are no coherence with the 117 Hz bump in IMMT1T QPD and MCE input signal, while there is strong coherence around 127 Hz, which is we know a stable peak found in DARM, IMMT1T, and CARM-related signals (but may be not so related to the current issue, although we love to damp it...). Then, whar is the next candidate of the scenario...???

Anyway, the other things I noticed in Fig. 3 are

• Why the 60 Hz power line (maybe) and its hramonics are only seen when down?
• Strong coherence below 20 Hz. Do we need more diagonalization of P/L for the MCE suspension? Is there possibility that this unwanted pitch input may have upconversion to 117 Hz? Indeed there are several other bumps found in IMMT1T when PRFPMI locked. For reference Fig. 4 is wider version of Fig. 3. Do we already have a measured transfer function from MCL excitation to DARM?

By the way the other peaks in IMMT1T that are not correlated with DARM seems to be due to the fluctuation of the pylon; they have strong coherence with IMMT1T oplev signals at these peaks.

[Ushiba, Yuzurihara]

I implemented the shell script to reset ASC feedback related PRFPMI. The script is at `/users/Commissioning/scripts/run_reset_ASC_feedback_all.sh`.

This script will be used before starting initial alignment or when relocking is difficult.

• For "IMMT2" "BS" "PRM", after reseting the gain, the script asign the good oplev values for PRFPMI.
• For "ITMX" "ETMX" "ITMY" "ETMY" "IMMT1" "PR2" "PR3" "SR3",  after reseting the gain, the script asign the good oplev values for FPMI.
• These process will be run at parallel. So, I expect it finish quick.

The operation test will be done on Monday.

[Ushiba, YamaT]
 

Abstract

We tried to modify anti-dewhitning filters at COILOUTF by estimating zero/pole frequencies of measured circuit responses of dewhitning filters reported as klog#23733 and klog#24162.
But measured circuit response seem to be something strange (affected by the saturation, response of measurement system, etc.?)
We need to know the measurement situation in more detail and may need to measure them again...
 

Details

Because engaging pairs of an analog dewhitening filter and an digital anti-dewhitening filter of the MN stage is not so stable,
we considered the possibility that dewhitening and anti-dewhitening filters did not fully cancel each other.
So we planned to modify the anti-dewhitening filters at COILOUTF by using the measured zero/pole of dewhitening filters circuits reported as klog#23733 and klog#24162.

According to the above two klogs, transfer functions of coil driver circuits were measured. But zero/pole haven't seemed to be estimated yet.
At first, we checked the measured data stored in
- /users/CAL/GitLab/cal-measurements-raw-data/pre_o4/common/2023-01-27_ETMX_coil_driver_TF
- /users/CAL/GitLab/cal-measurements-raw-data/pre_o4/common/2023-01-31_ETMX_coil_driver_TF
- /users/CAL/GitLab/cal-measurements-raw-data/pre_o4/common/2023-02-24_ETMY_coil_driver_TF
- /users/CAL/GitLab/cal-measurements-raw-data/pre_o4/common/2023-03-06_ETMY_coil_driver_TF

Then we found measured transfer functions have unexpected behavior as shown in Fig.1.
Figure 1 shows measured dewhitening filter responses for the MN H1 channel of ETMX with one, two, and three stages of z10p1 filters.
Though coil driver circuits have high frequency poles in order to avoid the oscillation, there should be no high frequency zeros.
So behavior above 10Hz in each measurement is unexpected one.

Measured TF only with measurement system has roughly flat response as shown in Fig.2.
There is gap around 1Hz and gain increases around 100kHz. But they are ~0.01dB and ignorable.

Other channels (MNH2-3, MNV1-3, IMH1-3, IMV1-3) also have a same strange behavior.
I show some of them as Fig.3-Fig.6.
Anyway, it's difficult to believe measured TFs now.
We need to know the measurement situation and we may need to re-measure all channels again...

I noticed that several network cameras(In the PSL room, on the PSL room, near IFI, and near PRM) and the wifi access point cannot be accessed via a network.

The common network switch for them was rebooted by Hayakawa-kun and the connection was recovered.

I did additional measurements to assess the performance of the sensor correction filter in IP-L. Analysis is pending.

Analysis of the previous measurements, that has not been reported in the klog, shows that the sensor correction filter is not performing as expected. At 0.2 Hz, attenuation of the seismic noise in the corrected sensor by a factor of 2 was achieved, whereas attenuation by a factor of 5 was expected. Therefore, I tried to find the origin of the discrepancy.

I did the necessary measurements to calculate again the intercalibration factors between the seismometer and IP-L and IP-T (per klogs 20873 and 22229). Although a proper calculation is pending, a cursory inspection of the data suggests the previously calculated values are just fine (see klog 22442). The data is in the directory /kagra/Dropbox/Subsystems/VIS/vis_commissioning/bs/inter_calibration/ in the file bs_ip_lt_intercalibration_with_seis.xml.

Then, I repeated the measurements referred to in klog 24444, but without any diagonalization in the LVDTALIGN matrix. The channel T was used for the uncorrected LVDT-L. The matrix was

LVDTALIGN =
[[ 1   , 0 ,  0  ],
 [ 1  ,  0 ,  0   ],
 [ 0  ,  0  ,  1  ]]

The data is in the file /kagra/Dropbox/Subsystems/VIS/vis_commissioning/bs/data/diaggui bs_ip_twr_float_05.xml.

For the last measurement, I used the filter banks for the blending filters in order to subtract the motion of the IP table measured by the geophones, from the corrected LVDT readout. In principle, this is the corrected LVDT noise. This quantity was at the channel  K1:VIS-BS_IP_IDAMP_L_IN1. The data is in file /kagra/Dropbox/Subsystems/VIS/vis_commissioning/bs/data/diaggui bs_ip_twr_float_06.xml.

Once I finished, I set all the LVDTALIGN matrix to whatever it was before (see klog 22376).

The 128Hz bump, which is suspected to be an environmental noise, is not changed.

Attached are graphs of residual gas measurements for IXC and EXC.
The pressure gauge nearest the cryostat is a Q-mass attached to the cryostat.
The tendency of fluctuation of the total pressure is small, and I consider it to be within the margin of error.

I assume that the pressure fluctuation in EXA (EXC?) is due to the effect of stopping the vacuum pump system of EXT.

I showed the spectrogram.
There are so many glitches above 100 Hz.
I may select accidentally not good timing for before IFI work(9:00) and accidentally good timing for after IFI work(12:00), but its spectrum is strongly depend on the glitch.
So, the effect of sound protect sheet is still unknown.

Anyway, please don't believe the spectrum of klog24526

Following the Yokozawa-san's original post, I just checked the spectra by my self and compared some, and found strange points.

In Fig. 1,

PRFPMI locked

• Blue 3/22 04:47 UTC
• Brown 3/23 6:29:00 UTC

PRFPMI Unlocked (only IMC locked with IMC ASC)

• Green 3/23 15:00 UTC
• Magenta: 3/24 0:00 UTC
• Red 3/24 03:00 UTC

Fig. 2 is the one that the red curve is removed from Fig. 1 for clearity.

What I found are:

• 60Hz peak and its harmonics can be seen when PRFPMI unlocked. I am not sure if this would be occasional or not, but might be related to electric grounding issue.
• Bumps at 20-30 Hz and 120Hz and 180? Hz can be seen when PRFPMI locked, but not seen when unlocked. Does this fact suggest that these bumpy structure might be introduced via MCL feed around path from CARM???
• Diaggui time stamp bug? Against the original post by Yokozawa-san, the time stamp assignment for the red and magenta curves are opposite. If this would be true, locating the sheets on the IFI chamber bellows legs might not be so effective (or at least this instance the noise floor got larger, but in my observation, this seems occasional... I am actually puzzeld. -> due to differet choice of frequeny resolution, according to Yokozawa-san. So breathing...

I have registered k1ctr29 to the DNS.

DGS Regular maintenance day(3/24)

Stepper motor
 [Checked]
  Check if the power supply can be controlled by BIO
  Check if the script can be started
  Checked that the limit switches installed are green.
 [Check result]
  No problem

Release Sensor
 [Checked the Release Sensor.]
 Checked that Release Sensor is 0, 1: Seated.
 [Check result]
 ETMX, ETMY, ITMX, ITMY : No problem

Picomotor
 [Confirmation]
  PingTest is performed.
 [Result]
  Only known

 NG:
  MCI,STM1,
  POS_POM,POM1
   => Confirmation only, no recovery operation was performed.

 OK:
  ETMX,ETMY,ITMX,ITMY
  BS(IM,BF),SRM(IM,BF),SR2(IM,BF),SR3(IM,BF,STM)
  PRM(IM,BF),PR2(IM,BF),PR3(IM,BF)
  MCE,IMMT1,IMMT2,OMMT1,OMMT2,OSTM
  REFL_WFS,POP,POP2,POS,POS2,POP_WFS,AS_WFS
  POP_POM, MCF(MCO,MCI)
  PCAL(EX1,EX2,EY1,EY2):Not checked because it is ON only when used

Temperature controller
 [Checked]
 Is the temperature acquisition working properly?

 [Check result] 
  Is the temperature acquisition working properly?
  => No problem

Precision air Processer
 [Check Contents]
 Is the temperature acquisition working properly?

 [Result]
  No problem

FunctionGenerator
 (Confirmation)
 Is there a ping response?
 Are all EPICS channels alive?
  [sitemap]-[Commissioning]-[ALS FG]-[Xarm],[Yarm].
 [Check result]
 X and Y PLLs are working properly.

HWP control PC
 (Confirmation)
 Check if it is possible to login to the control PC.
 (Result) 
 Both k1hwp0, 1 , 2, 3 and 4 are OK.
 

Ushiba, Tamaki

ref: klog24290

We have put notch in some suspension controls to achieve the goal of eliminating the peaks that appear in the DARM sensitivity.
To start with, I put in an OLDAMP in SR2 to suppress the 47.75 Hz peak that appears in the OMC DCPD signal (K1:OMC-TRANS_DC_{A,B}_IN1_DQ), and an OLDAMP notch in SR3 to suppress the 45.75 Hz peak and the 39.5 Hz peak.
(The cause of each peak was confirmed by Bruco.)

As a result, the peaks that appeared in the OMCPD disappeared (green to red) as shown in Fig.1, and the peaks in DARM sensitivity was also suppressed.

Ushiba, K. Tanaka

### Abstract

We implemented the ASC for almost degrees of freedom of PRFPMI (OMC -> Beacon ASC, COMMON ETM -> WFS, ITMs -> BPC, PRMI -> ADS) except for DIFFERENTIAL ETM. We tweaked ETMY to move DIFFERENTIAL ETM degrees of freedom, and the fluctuation of the DARM error value seems to improve. So this indicated we can increase the input power to the interferometer if we implement ASC for DIFFERENTIAL ETM. Moreover, the alignment seems improved, the optical gain slightly improved in the high-frequency region.

### What we did

#### Implementation of OMC Beacon ASC

• We implemented OMC Beacon ASC with PRFPMI RF lock. At this time, only ADS for PRMI is engaged. The Beacon ASC setup is almost identical to the one I did with the FPMI in December. The fluctuation of the transmitted power is still significant because the instability of the interferometer's output is still significant even though Beacon ASC turns on (Fig. 1).
• After then, we engaged DC readout. As for the sensitivity curve, the excitation for the beacon signal does not affect the sensitivity. Moreover, the optical gain is increased slightly from the previous gain (Fig. 2), and the sensitivity is improved slightly at the high-frequency region ( >1 kHz) (Fig. 3).

#### Engage COM ETM control with WFS and BPC on ETM

• We engaged ASC with WFSs for COMMON ETM {PIT, YAW} and BPC on ETM {PIT, YAW}. Therefore we controlled almost the alignment degrees of freedom of PRFPMI except for DIFFERENTIAL ETM {PIT, YAW}.
• The beam shape on the AS camera split horizontally, so we tweak ETMY in the YAW direction to move DIFFERENTIAL ETM YAW. Fig. 4 shows the time series around this tweaking. The time indicated by the left cursor is when we turned on ASC by WFS, and the time indicated by the right cursor is when we turned on BPC. The DARM error signal in the upper right of the figure is improved slightly. This result suggests that controlling the degrees of freedom of the DIFFERENTIAL ETM allows a large input power to the interferometer because of the more significant margin of error signal offset.
  We are looking forward to implementing ASC for DIFFERENTIAL ETM.

Just for visual inspection; checked energy conservational viewpoint.

In terms of REFL power, the brighter is the better. Even while slight reduction of the REFL power, the arm transmission, PRC-inside carrier power, and AS contrast seem tolerable. But, if the REFL power reduced further, drastical reduction of the arm transmission, PRC-inside carrier power, and AS contrast quality simultaneously happens.

For example, if just only looking at the REFL power, and if it reduces like the one indicated as "2" in the attached figure, one may think it would be obvious that the contrast of the interferometer would get worse, but the fact is not so simple; seemingly not so "entangled". The phenomenon is not so linear. Where the energy went...? As if there would be two different causes of the REFL reduction. One cause worsen of the contrast, while the other would not do so.

[K.Tanaka, Ushiba]

We made several states in ASC_LOCK and OMC_ASC guardians.
Since we haven't decided how to use these states during the lock acquisition, we haven't changed LSC_LOCK guardian.
Followings are the summary of the new states.

ASC_LOCK guardian:

*ENABLE_REFL_WFS:
Open beam shutter for REFL WFS, which is closed during lock acquisition for reducing glitches in ALS DARM signals.
This state also open K1:ASC-WFS_SW and set the gain of 1 to K1:ASC-WFS_GAIN.

*CENTERING_REFL_WFS_QPD:
Close QPD centering loop for REFL WFSs.

*REFL_WFS_QPD_CENTERED (requestable):
A state just for stopping guardian after QPD centering loop engagement.

*SET_COMMON_ETM_WFS_OFFSET:
Average current WFS signals for 1 seconds and set new offset according to the averaged values not to move the suspension too much by engaging WFS loop.

*ENGAGE_COMMON_ETM_WFS:
Close WFS loops for ETM common motion.
Currently only yaw loop is closed during the state not to conflict ADS of pitch DoFs.

*COMMN_ETM_WFS_ENGAGED (requestable):
A state just for stopping guardian after engagement of WFS for common ETM motion.

New edges can be seen in fig1.

OMC_ASC guardian:

*PREP_OMC_BEACON:
Waiting for OMC lock and open the excitation path for beacon.

*ENGAGE_OMC_BEACON:
Average current beacon signals for 1 seconds and set the offsets values.
Then, close the beacon loops for OMMT2 and OSTM.

*OMC_BEACON_ENGAGED (requestable):
A state just for stopping guardian after engagement of WFS for common ETM motion.

New edges can be seen in fig2.

Date: 2023/3/23

With Satoru Ikeda, Shingo Fujii

We removed some cables which were not used and hanging in the air.

Because we could trace the rest ones, we did not touch them.