According to the outcome of the yesterday daily meeting, this morning I started to change the DAS powers both for NI and WI to match the values (see VIR-0698A-22) reported in the following table:

I encountered a problem with the WI DAS IN wave plate (LP2 in figure 1) as the corresponding calibrated channel (V1:TCS_WI_DAS_POWER_IN_CALI) did not reach values ​​below 450 mW. Around 7.45 UTC, Suzanne and I went to the WI CO2 bench to verify that the power meter was working properly and whether the DAS calibration was still valid or not. After verifying that the DAS IN power meter was working well, we moved the Pickoff PM in front of the viewport and we switched on only the WI DAS IN to see if the value on the Pickoff corresponded to what reported by the calibrated channel: it was ok. Then we decided to check all the WI actuators and the NI actuators too, to be sure that the calibration was still valid for all the channels. All these checks went well. In the meanwhile we noticed that the V1:TCS_WI_DAS_POWER_IN_CALI decreased to 100 mW (due to the temperature decrease of the WI bench, as it was open - see figure 2).

When we came back to the control room, I tried to move again the WI DAS IN wave plate but the power could only increase, it would not go below around 130 mW.

We first investigated the possibility of a problem with the process TCSAgilisRot (even if it had low probability, since the other rotators were moving without problem).  I stopped the process but I couldn’t get it to restart. I asked Nicola to try (see the operator entry 56364) and, at the end, the problem was fixed by Rhys Poulton and Franco Carbognani.

However, this did not solve the problem of the WI DAS IN, which, in the meanwhile, had increased to 450 mW, due to the bench thermalization after closing.

The other possibility is that 450 mW is really the minimum reachable power, even if on June 28th the power was set to 300 mW (as reported in logentry 56314). If this is the case, the solution is to move LP1 (see figure 1) which changes the power ratio between the two DAS arms. As a consequence, the WI DASes would need to be re-calibrated.

As decided at the daily meeting, tomorrow morning we will proceed with this option.

The goal of this shift was to implement the loop that centers the beam on the ITMs by adjusting their transversal position. The error signals fed to the loops are obtained by demodulating B7/B8 on the LF lines injected on the ITMs.

We sent to DPS the following channels:

• ASC_NI_TX_LF_B7_I_CAL_FS
• ASC_NI_TY_LF_B7_I_CAL_FS
• ASC_WI_TX_LF_B8_I_CAL_FS
• ASC_WI_TY_LF_B8_I_CAL_FS

 We adjusted the demodulation phases and checked the sign on the calibrations, then we engaged the loops in the DSP.

Their action is very slow, but after some gain tuning to cure an high-gain oscillation, they seemed to work as the error signals slowly converged to zero.

Fig.1 shows the behaviour of the loops over a longer period of time, compared with the local controls. A few interesting points:

• UTC:20:30: we set a ramp on the horizontal position of NI,WI to observe the reaction of the loops. WI reacted well and slowly converged to the original position. The signal of the NI did not react significantly, but the local control shows the mirror converging to the original position as well. Interstingly, this ramp triggers a slow oscillation on the TX component of the signal, that is then applied on the mirror Y coordinate by the loop.
• UTC:20:50: same test on WI Y. It reacts well and converges back to the original position.
• UTC:21:00: same test on NI Y. It also reacts well, but this movement produces a reaction identical in shape on the X error signal for the NI, which is then applied on the mirror by the loop. It seems that the TX/TY components of the error signal for the NI displacement are strongly coupled. To be further investigated.

Overall, these loops appear to be behaving adequately, since despite the aforementioned issues, after each test all the error signals went slowly back to 0 and the local controls show a corresponding return to the initial positions. While they could definitely benefit from additional work, like a better calibration, this test was successful.

We disengaged the loops and left the arms locked on the IR.

The planned activity on IM centering proceeded without issues; still in progress.

No other relevant operations to be reported.

DAQ

ISC_Fb restarted after a crash at 16:46 UTC.

No DMS alerts in the last 8h.

The change of the PLL setpoint was necessary due to the change of the reference laser frequency. We notice the jumps of the reference laser since June the 10th. The value changes by ~657 temperature steps which corresponds to ~500 MHz in the laser frequency.

PLL Main:

pll_TH0 = 26364   # till 2022 06 10
pll_TH0 = 27032   # till 2022 06 13
pll_TH0 = 26400   # till 2022 06 28
pll_TH0 = 27029   # till 2022 07 04
pll_TH0 = 26331   # current

The current value is the typical that we had since 2021.

Today the CC loop was not engaging correctly.

The stray light loop was engaging correctly and correcting the effect of the bump (with CC open).

Then after several trials and by doing reset and select of the CC filter, CC loop was closing, despite the usual problem of M1 and M2 saturation (coarse).

The seism did not seem worst than the previous days.

We added an actautor on the CC coarse loop to have more range: M4_Z. This actuator was used for the stray light loop.

We swap the correction sent on the mirror on python:

EQB1_HD_AA.M4_Z_cmd.M4_Z_CORR = 0 (SLL correction off)

EQB1_HD_AA.M4_Z_cmd.HD_CC_COARSE_OUT = 1 (M4 for CC coarse).

Now the CC coarse is acting on LO_M1, LO_M2 and M4_Z. The CC loop is now closed in a stable way.

NB. after the work we should remeber to put back the initial configuaration.

Today's shift task was to attempt the locking of the interferometer after the new DAS power settings.

However at the beginning of the shift we encountered some troubles on setting the powers on the WI DAS inner. On this actuator, it was impossible to set the power below the value of 450 mW, so Claudia and Suzanne went to the bench to investigate the issue in order to find out if it was either a mechanical problem or something else. At the end the problem was not solved and the power on WI DAS IN has been set on a different value from the scheduled one. For more details, see dedicated entry.

From ~ 13.00 UTC we started to lock the ITF anyway. We obtained several few-minutes-long locks. The general monitored figures of merit were not good: B4_DC, 12MHz and 112MHz sidebands had lower values with respect to the usual ones, and also both B4cam and the EPRB phase cameras showed great asymmetric shape, in both directions, especially in TY, see fig.1.

During the locks the only tests we performed were to adjust the PR angular setpoints, and the TY DoF has been adjusted accordingly, TX didn't need any adjustment.

During the last lock, at 14.07 UTC we experienced a jump, not sure if triggered while we were adjusting the PR setpoint. In fig.2 are shown the trend of the main figures of merit during the occurrence.

In Fig.3 and Fig.4 are reported the images of the B4 camera, before and after the jump, respectively.

In the footage attached, for more clarity, is reported the animation of the EPRB phase cameras during the jump.

The planned ISC activity on TCS locking (Pinto, Casanueva, Taranto) started at 07:00 UTC and went on without major problems for the whole shift.
 

Parallel activity:
SQZ - measurements (Ding, De Laurentis, Vahlbruch)
DAQ - 12:05 UTC - All the Fbs servers were restarted.

Automation
At 09:30 UTC DET_MAIN Metatron node got stuck. I put it on Init,  FORCE_CHECK_CLOSE, and then on SHUTTER_CLOSED.

DAQ
TCSAgilisRot VPM process crashed, I tried to restart it without success. Experts fixed the problem.
The crash of the TCSAgilisRot process highlighted a pre-existing DAQ problem that was remotely fixed by Rhys Poulton and by Carbognani in TCS Room (replugging the USB port).

DET
At 12:53 UTC the spot of B1p was in the lower part of the screen. I fixed the position via VPM (SDB1_LC):
- Used fixed setpoints (spot again in the center of the screen)
- Open B5 beam drift
- Close B5 beam drift
- (after few minutes) Use floating setpoints
- Open B5 beam drift

TCS
The bug on TCSAgilisRot has been fixed but the optical problem with the rotor that moves a wave plate is still present and must be solved.

DAQ
(06-07-2022 10:00 - 06-07-2022 11:00) Other
Status: Ended
Description: TCSAgilisRot VPM process crashed, I tried to restart it without success. Experts fixed the problem.

Noise assessment of maintenance works of CEB Hall AHU

Attached is a comparison between the predictive accelerometers (ENV_CEB_AHU_RET_xxx) that were operational during maintenance works on the CEB HALL SUPPLY AHU system.

(Purple = before maintenance, Blue = after maintenance)

No major changes are visible, apart from a slight increase in frequency of some lines. The overall amplitude remain the same.

All the Fbs servers were restarted with the version v8r23p2 using the lastest Fd version v8r43p7.

Operations performed between 2022-07-06-12h04m22-UTC and 2022-07-06-12h05m37

WE PCal photodiodes has a frequency response which is not flat, because of the Audio channel.

They have been replaced, the 05/07/2022 morning, with photodiodes with preamplifiers without audio output channel, made by François Frappez.

A new bigger shutter has been placed in front of the viewport on Both injection and reflection benches.

Photodiodes calibration

The ADC port dedicated to WSV has been calibrated, but the gain and offset hasn't been changed. This calibration has been done only to monitor the stability of the calibration gain of the ADC through time.

The laser is switched off, O3-WSV has been placed on injection bench. Then the PCal box is closed and the offsets of the channels (WSV Tx_PD1 Tx_PD2) is removed.

The laser power is set to 1.3 W, without control loop. The calibration data start from 9h48 to 10h14 UTC.

Rx photodiode has a high noise level, and hasn't been calibrated.

The new gains has been applied the 06/07 at 8h53 UTC

ITF State: ITF Locked on IR
ITF Mode: Commissioning
Quick Summary: IMC and RFC locked; all Suspensions loops closed, all SBE loops closed

Activities ongoing since this morning: No activities ongoing

DAQ
06:00 UTC - Fds_Img process found down. Restarted via VPM

Triggered by the observation of a slow drift of the beam on the ITMs seen by the new demodulated B7/B8 AA signals (mainly tx) that seems to correlate with a thermal drift on SIB1 (see fig. 1) we wanted to study whether this correlation is real and, in case, if this is generated by a " wrong" parking position of the PR (during for exemple the LOCKED_ARM_IR state) that would send the backreflected beam by PR touching somewhere IB bench (or some optics on it).

This effet, indeed, has already been observed few times, and was taken into account to choose the actual PR parking position several months ago (delta_ty = -60 urad).

Anyway, since then, the PR has been recently re-centered and the parking position has not been checked again.

At the beginning, we pre-aligned the PR with the usual procedure and live it in aligned position (ACQUIRE_DRMI state) for ~45' in order to cool down (thermal transient ~1h, see fig. 2 for exemple). We then changed the PR parking position at delta_ty = -20, much closer the the aligned optical path. Indeed, in principle the closer to the aligned path do we are, the less chances do we have to impact somewhere on SIB1.
As a figure of merit, we use thermal probes on SIB1, RFC_TRA_DC and the arms demodulated signals. A summary of the performed test are shown in fig. 3.
In a first part, we moved relatevely fast the PR parking ty from -20 to -60 urad in order to see if there was some macroscopic effect on SIB1/SIB2 cameras, but we didn't spot anything in particular.

Because of the time needed to study the thermal transient, we thus decided to focus on just one position for the rest of the shift and try to repeat the duty cicle "thermalisation/heating up" few times for the delta_ty= -20urad position.

A deeper analysis of the collected data will be performed offline. Nevertheless, it is worth to notice two point (fig. 4):

• we observe a strange behaviour of the transmitted power for some parking position.
• No easy correlation with the IB temperatures can be observed.

At the end of the shift we leave the arms locked and the PR parking position reverted to -60urad.

PS: at the beginning of the shift we did a measure of the arms mismatch, the results are shown in fig. 5

The first hours of the shift were dedicated to complete the NCal NE installation.

At around 15:00 UTC started the planned activity on ISC-INJ SIB1 thermal drifts (Spinicelli, De Rossi); the activity proceeded without issues and is still in progress at the end of the shift.

In parallel SQZ activity: alignment optimization (De Laurentis); activity concluded.

SBE

SNEB loop closed at 15:16 UTC.

DAQ

The DMS was reporting a red flag for Fast_DAC - WEB_PCAL_FAST_DAC_ssfs_flags; I informed the experts and they fixed the problem.

During today's maintenance we complete the installation of the NCal system in the North End building. More specifically:

• We brought R4-06 and R4-31 to the site and installed them respectively on the East Middle slot and the South Near slot.
• We moved R4-04 to the East Near slot (see picture East_setup), R4-05 to the South Middle slot, R4-02 to the North Far slot.
• We performed the zeroing of the three setups using new 2.5 mm calibrated spacers (see first picture). The residual offset was below 0.01 mm.
• We replaced the 2.5 mm spacers by 1 mm spacers, inducing a 1.5 mm offset, and observed the following offsets on the axial position: 1.56 mm for NSF, 1.53 mm for NSN, 1.54 mm for NEF, 1.47 mm for NEN, 1.58 mm for NNF and 1.52 mm for NNN.
• The setups have then been released and we roughly balanced and aligned them, the three figures (North, East and South) are showing the typical motion around their rest position.
• The cabling and configuration of the DAQ box 119 has been updated to include the NCal box temperature readout. For this purpose the two TCS readout have been moved to the Far end of this DAQ box (see photo of the new layout).

We left the power supplies on for the three setups.

The wind speed sensor of the external weather station was not working since June 10 12:26:33 UTC (Fig 1). After a few plug in and out we declared the sensor not working and decided to replace it. We borrowed the full wind block from a new weather station waiting to be mounted on a terminal building and installed it to replace the broken one. From today 13:54:00 UTC we can consider the wind-speed sensor back in operation

Today I tried to characterize FDS with the stray light loop closed.

The system was not aligned for the IR and Jacques recovered the alignment reaching 11V in transmission for the IR.

1. Bump check on the audio diff channel

I tried to optimize the filter of the stray light loop (SLL) to suppress the bump in the best possible way.

Fig. 1 the shot noise with the loop open (magenta) and closed (blue).

ACL_FILTER_SET        "HD_M4_Z_flt_boost"        1       250    1    8   
ACL_FILTER_POLES    "HD_M4_Z_flt_boost"        0          0
ACL_FILTER_POLES    "HD_M4_Z_flt_boost"        0.01     0
ACL_FILTER_ZEROS    "HD_M4_Z_flt_boost"        0.5      0

ACL_FILTER_POLES    "HD_M4_Z_flt_boost"        15         0

We tried to move the second pole position lowering it (from 20 to 15 Hz). Then to increase the Q factors of poles and zeros, but it didn't improve the situation. We also changed the gain at 1Hz and we decided that 250 it was the best solution.

2. Compensation line on LO_M1 : ATTEMPT FAILED

I tried to introduce the compensation line at 500 Hz on LO_M1 but the CC loop was not stable enough to be able to tune it properly.

I tried with the parameters found by Fiodor and Jacques in the past:

line M4_Z = 30 mV and line LO_M1 = 0.19 V -> to have more or less these values for LO_M1 line, I set on the vpm 2.5 as line weigth, phi = -0.3.

Since it seemed too high as modulation amplitude, I went back to:

line M4_Z = 15 mV and line LO_M1 ~ 0.1 V -> to have more or less these values for LO_M1 line, I set on the vpm 1.25 as line weigth, phi = -0.3.

It seemed that the line was compensated, but the CC was diverging (LO_M1 and M2), making impossible to really appreciate the compensation and to optimize it.

3. Normalization of the channel EQB1_HD_DIFF_AUDIO_RMS_1_3kHz_dB/norm

By doing these studies, I realized that the channel that we usual use to check the squeezing level, was not normalized. Infact the channel EQB1_HD_DIFF_AUDIO_RMS_1_3kHz_dB was ~0.6 dB only with the LO (shot noise) and the channel EQB1_HD_DIFF_AUDIO_RMS_1_3kHz_norm was around 1.06 V instead of 1. I changed the calibration factor first on python and then on EQB1_HD_AA config file:

ACL_SUM_CH    HD_DIFF_AUDIO_RMS_1_3kHz_norm    ""    1    270270      HD_DIFF_AUDIO_RMS_1_3kHz

I just did save and not reload config.

4. FDS at 100 Hz characterization with SLL open/closed

I decided to go on without compensating the line and to acquire a full characterization of the system detuned at 100 Hz, i.e. sqz/asqz/shot measurements with SLL open/closed.

NB. the CC was regularly unlocking and LO_M1,M2 were saturating. It means that we should be careful in analyzing data offline.

4.1 ASQZ : phi CC = 0.6, level = 6.7 dB

12h55m00 UTC : SLL open

13h06m13 UTC  : SLL closed

4.2 SQZ : phi CC = 1.9, level = 3 dB

13h08m40 UTC : SLL open

13h17m30 UTC  : SLL closed

4.3 Shot noise

13h20m00 UTC : SLL open

13h43m05 UTC  : SLL closed

In Fig. 2, these spectra are shown.

-----------------------------------------------------------------------------

I felt the system with CC open and I turned off the line on M4_Z.

Once the system will be further optimized, we will characterize FDS with the SLL open/closed.

The shift was dedicated to maintenance. Here is a list of the activities reported to the Control Room:

• standard Vacuum Refill from 6:00UTC to 9:30UTC (VAC Team);
• cleaning of CEB, NEB buildings, and FCEM (Ciardelli, Menzione with external firm: from 6:00UTC to 10:00UTC);
• DET - scan of SPRB in single bounce (Gouaty)
• revision of the NEB and WEB crane (Baldocchi with the external firm)
• CAL - replacement of WE PD on Tx bench (+ alignment and re-calibration) (Lagabbe)
• grass cutting around CEB
• Installation of 2 cables from TCS room to WI and NI TCS benches (Nocera, Montanari)

all the maintenance activities listed before were completed at 12:00UTC.

NE CAL activities are still ongoing.

Yesterday (Mon July 4th), I prepared the system for the SQZ measurement campaign of this week.

First there was a change in the PLLs setpoints (slow) during the week-end, for unknown reasons : 

Old one:

• MAIN: 27029 
• CC: 34477
• SC: 33220

New one:

• MAIN: 26329
• CC: 33864
• SC: 32035

Then the FC was aligned with both GR and IR. Once this was done I did some quick measurements of FIS with RR :

• before 10:47 UTC : CC contrast optimization on RR
• 10:48 UTC : scan of phase from 0 to 3.5 rad in 120s. ASQZ at 0.86 rad, sqz at 2.18 rad.
• 10:54:00 to 11:01:00 : ASQZ phase.
• 11:01:10 to 11:06:10 : SQZ phase. Fast unlock at 11:05:45.
• 11:07:20 to 11:12:20 : Shot noise.

After this, the RR was removed and the system was left untouched. This morning it was found that the IR_AA was probably left on when the RR was moved in, causing IR_M3 and IR_M5 to change positions. This means that the measurements of FIS on RR had probably more losses in alignment. 

Attached is the plot of the FIS on RR. I estimate : 

The losses are quite high (14%), correlating with the un-optimal alignment of the IR_M3 and IR_M5.

The measured OGs are expressed in mW/urad, not W/urad.

We reverted the changes in the DRMI.ini file. The parameters are the same used before the re-alignment of the TCS actuators.

This morning we scanned the position of the SPRB bench and checked the image of the B4_Cam1 at the same time.

ITF put in single bounce by the operator at ~07h50 utc.

We put the integration time of the camera B4_Cam1 at maximum to better visualize the image. Snapshots of the camera image taken during the following scans are shown in the attached file.

Initial position of the bench: TX = 155 urad, TY = -604 urad, X = 900 um, Y = -70 um,  Image on Page 2.

Scan in TY (TX = 155 urad):

08h24m10 utc (2 min) TY = -559 urad: Fig.4

08h27m10 utc (2 min), TY = -569 urad: Fig.5

08h29m40 utc (2 min), TY = -579 urad: Fig.6

08h32m00 utc (1min30), TY = -589 urad: Fig.7

08h34m00 utc (3min), TY = -599 urad: Fig.8

08h37m50 utc (1min30), TY = -609: Fig.9

08h39m40 utc (1min30), TY = -619: Fig.10

08h41m30 utc (1min30), TY = -629: Fig.11

08h43m50 utc (1min30), TY = -639: Fig.12

08h45m40 utc (1min), TY = -649: Fig.13

08h47m10 utc (1min), TY = -659: Fig.14

Scan in TX (TY = -609):

08h57m00 utc (1 min), TX = 180: Fig.16

09h02m00 utc (1 min), TX = 170: Fig.17

09h03m40 utc (1 min), TX = 160: Fig.18

09h05m20 utc (1 min), TX = 150, Fig.19

09h07m20 utc (1min), TX = 140, Fig.20

09h10m40 utc (1min), TX = 135, Fig.21

Scan in Y (X = 900 um):

09h48m00 utc (2 min), Y = 250 um, Fig.23

09h54m00 utc (2 min), Y = 150 um, Fig.24

10h00m00 utc (2 min), Y = 50 um, Fig.25

10h08m00 utc (2 min), Y = -50 um, Fig.26

10h14m00 utc (2 min), Y = -100 um, Fig.27

10h18m00 utc (2 min), Y = -150 um, Fig.28

10h22m00 utc (2 min), Y = -200 um, Fig.29

Scan in X (Y = -70 um)

10h58m00 utc (1 min), X = 1400 um, Fig.31

11h01m00 utc (30s), X = 1200 um, Fig.32

11h03m00 utc (30s), X = 1000 um, Fig.33

Qualitatively, scanning the bench position does not seem to have a clear effect on the B4 camera image. A detailed analysis remains to be done.

To ease the cabling for some NCAL temperature channels , the TCS_HWS_NE_TE1 and TCS_HWS_NE_TE2 channels are now connected on the ADC2378_SN276 channels 6 and 7 .

The ones related for the  NCAL are connected on the ADC2378_SN276 channels 0 and 1.

First I realigned the PSDs on SQB1 which were misaligned (mainly vertically) since the recovery after the sorbothane replacement of EQB1:
PSD1:
 -- GM12_V ==> -10040 steps.
 -- GM12_H ==> 160 steps.
 
 PSD2:
  -- M_PSD2_V ==> -24300 steps
  -- M_PSD2_H ==>  -5870 steps
 
 PSD3:
  -- M_PSD3_V ==> 15500 steps
  -- M_PSD3_H ==>  1800 steps

Then I realgined the one on SQB2 :

PSD2:
 -- GM11_V ==> 127 steps.
 -- GM11_H ==> 55 steps.
 
PSD1:
  -- GM_13_V ==> 60 steps
  -- GM_13_H ==> 30 steps

Then I scanned the GR_M5 and GR_M7 mirrors both in X (horizontal and vertical).
Below are the most interesting scans

Scan M5 _Y :
2022-07-05 07h54m34 UTC    bonnand 'GR M5:Y ramp - GM5 [Frequency=0.05,ampl=10,bias=-2.2]' sent to AFC_Ctrl
2022-07-05 07h55m43 UTC    bonnand 'GR M5:Y ramp - GM5 [Frequency=0.05,ampl=0,bias=3.8]' sent to AFC_Ctrl

Scan M5_X:
2022-07-05-08h02m54-UTC    info bonnand      'GR M5:X ramp - GM5 [Frequency=0.05,ampl=10,bias=-5]' sent to AFC_Ctrl
2022-07-05-08h03m48-UTC    info bonnand      'GR M5:X ramp - GM5 [Frequency=0.05,ampl=0,bias=0]' sent to AFC_Ctrl

Scan M7_Y:
2022-07-05-08h19m59-UTC    info bonnand      'GR M7:Y ramp - GM7 [Frequency=0.05,ampl=16,bias=-8.67745,ramp duration=1.]' sent to AFC_Ctrl
2022-07-05-08h21m07-UTC    info bonnand      'GR M7:Y ramp - GM7 [Frequency=0.05,ampl=0,bias=0.67745,ramp duration=1.]' sent to AFC_Ctrl

Scan M7_X:
2022-07-05-08h15m14-UTC    info bonnand      'GR M7:X ramp - GM7 [Frequency=0.05,ampl=10,bias=-7.4]' sent to AFC_Ctrl
2022-07-05-08h16m30-UTC    info bonnand      'GR M7:X ramp - GM7 [Frequency=0.05,ampl=0,bias=-2.4]' sent to AFC_Ctrl

Analysis will follow later.

Observations :

During the scans the FC locked in a wrongly aligned position.
The rigth alignment was recovered by adding and offset on M5_X to get the FC transmission high, with GR pointing ON.
Then the offset was removed steps by steps allowing the GR pointing loop to compensate the M5_X offset which was set back to 0 at the end.

I left the system back to its original settings.

ITF State: LOCKED_ARMS_IR
ITF Mode: Commissioning
Quick Summary: IMC and RFC locked; all Suspensions loops closed, all SBE loops closed.
Activities ongoing since this morning: None