The first part of the shit was dedicated to the LNFS replacement.

After that the commissioning team worked at the CARM NULL recovery; activity concluded at around 19:00 UTC.

At 19:33 UTC the ITF was set in NI single bounce for 5 minutes.

ITF left in LOCKED ARMS IR.

SUSP

SR LC opened by the guardian at 16:00 UTC; properly closed.

The thresholds used for the definition of the QPD_B1s flag have been update in DetMoni. The checks on the B1s centering and the status of the galvo loop will be performed only after the state "SHUTTER OPEN".

Still digging in the data of July 28^th switch off, we further characterize acoustic noise entering the INJ room from outside. This investigation profits for the very low noise condition while the INJ HVAC was off.

A first point is the noise that enters from the Atrium. Figure 1 shows the effect of opening the Atrium door. There is a broadband noise increase of sound and vibration noise above 100Hz, particularly enhanced between 500 and 600 Hz (a factor 10) and the amplification of some narrow lines (105, 134 ... 290 ... 1010, 1048 Hz) which are indicated in Figure 2 with red markers.  For accelerometers see Figure 9. Inspecting this morning with Suzanne, we verified that the main (and seems only) sound source in the Atrium is the cooling fan of the ALS amplifier module (see Picture).   

The second  point is shown in Figure 4: note that structures between roughly 100 Hz and 1kHz in the acoustic and bench vibrations have large coherence with the IB tower vibration (sensor ENV_IB_ACC_X). We think this noise is related with the Turbo pump, as reported in these old tests (5 yrs ago.. and it seems yesterday!): elog 52500, this Figure  

Figure 4 also shows coherence with IB tower vibration in the region 1-3 kHz, that highligted in red circle. This is most probably noise from racks. Although we have not able to identify them: also the microphone placed on the platform near SIB1 tower does not show enhanced and coherent noise at these frequencies. And no evidence from EEroom microphone either.

Last point concerns the noise entering from EEroom: it is a triplet of acoustic lines (also with seismic counterpart) between 44 and 48 Hz  (Figure 5) and one important peak at 60Hz (Figure 6): all of them more intense in EEEroom mic.

The quadrant photodiode acquiring previously the B5_QD2 beam is now used to acquire the B1s_QD1 beam .

To be compliant with the new name of the beam, we  proceeded with the renaming of the B5_QD2 channels to B1s_QD1 channels .

The operations were performed between 2026-09-08-06h39m28-UTC and 2026-09-08-09h18m52-UTC 

Below the details of the operations performed:

• SDB2_dbox_bench server
  • DBox SN106 - demodulation mezzanine SN87
    • rename  as SDB2_DEMOD_B1s_QD1 instead of  SDB2_DEMOD_B5_QD2 previously
    • add the demodulation of the B1s_QD1 beam at  6MHz, 50MHz, 56MHz and 112MHz  for the 4 quadrant sectors
  • DBox SN007 - ADC2378 mezzanine SN43
    • renamed as  SDB2_ADC_B5_B1s_QD instead of SDB2_ADC_B5_QD previously
    • add the readout of the B1s_QD1_{H,V,SUM} channels  instead of the B5_QD2 ones
• SDB2_Tpro server and SDB2_ Fb server
  • Replace the old Tolm packets' name by the new ones
• SDB2_Quadrants  server 
  • Remove all the functionalities related to the B5_QD2 quadrants channels 
  • Add the last part of the demodulation of the B1s_QD1_{6,50,56,112}MHz channels
  • Add the B1s_QD1 galvo loops part
• SDB2_readout server
  • swap the B5_QD2 channels with the B1s_QD1 ones
• QPDMoni server
  • change the name associated to the interface ethmt7-2011 as B1s_QD1 instead of B5_QD2
• ASC_Acl server
  • the internal channels labelled B1s_QD1 are build using the SDB2_B1s_QD1 input channels
  • the internal channels labelled B1s_QD2 are build using the EDB_B1s_QD2 input channels
  • the B5_QD2 channels used in the TX_SENSING matrix and in the TY_SENSING matrix have been replaced by the B1s_QD1 respective channels
• Automation servers. ARMS_LOCK, DRMI_LOCK,
  • The related python config files have been updated by commenting the parts related to the SDB2_B5_QD2 quadrants 
• DetMoni server
  • all the parts related to the B5_QD2 quadrant photodiode have been commented
  • the monotoring part related to the B1s_QD1  quadrant photodiode have been added and remains to be tuned
  • A new flag B1s_QD1 has been in the DMS interface : the DMSserver server has been restarted to take in account this modification

6:00 UTC ITF found in LOCKED_ARMS_IR, COMMISSIONING, and set to MAINTENANCE.

Operations communicated to the control room:

• Nitrogen refill (VAC)
• Scheduled Tuesday cleaning (external company)
• TCS chillers refill (Ciardelli)
• Laser Lab measurements (Fiori, Tringali)
• Glycol addition to TCS AUX cooling system (Zaza)
• TCS cameras checks (Nardecchia, operator)
• TCS CH power checks (Lumaca)
• LNFS troubleshooting (Nocera, Spinicelli)
• SDB2_B5_QD2 renamed to B1s_QD1 in vpm and automation #69707 (Masserot, Bersanetti)
• Thermocamera Installation at NI #69708 (Melo, Spinicelli)

Taking advantage of the maintenance morning, between 09:20 and 09:38 UTC I set the CO2 DAS powers to the values identified by simulations as the starting working point for the interferometer with 12 W input power.

The CH powers were not set during the morning activity because glycol had just been added to the chillers (69706) while I was adjusting the DAS powers, making the measured powers unstable. Their setting was therefore postponed until after the daily meeting, when it was confirmed that the activity could proceed, and was completed at 12:25 UTC.

Since the CH powers had not yet fully stabilized, the reported values should be checked again when the CH flip mirrors need to be opened.

The resulting powers are reported below:

All flip mirrors were left CLOSED at the end of the activities.

Today we have installed the thermocamera Optris 640i on the HR face of the NI mirror, using the same support used for the infrared NI_cam. We adjusted the image focus and connected it to the online streaming, accessible only via EGO Wifi, at the address : http://172.31.10.192:8080/.

We also had to slightly adjust the focus of the NI_cam after touching the support to install the Optris thermocamera.

In the first part of the shift the injection team worked on the input power reduction (#69702) then the ITF lock recovery started; the work went on without any major problem with the lock recovered up to CARM_OFFSET_STEP_3_OF_3 state (#69704), the work stopped at 18:30UTC, we left the cavities locked on the infrared.

DAQ
Another event of data loss form suspensions happened at around 17:30UTC I recovered the data with at stop and restart of SUSP_Fb process.

The lock has been recovered up to step 3/3

the gains have been changed accordingly to the change of power (the ones that are not normalized)

the demodulation phases have been tuned (apart from Diff +)

Today afternoon, around 13h UTC we started the procedure to decrease the power towards the ITF. The power was lowered from 18W on the IMC transmission to 12 W. We followed the steps described below:

• We measured the TF of the IMC. The UGF was 110 kHz and the phase margin was 23 degrees;
• We moved the IPC1 and IPC8 rotators to adjust the power of the tranmission and reflection of IMC, respectively. The overall steps are summarized below:
  • cm_send('EIBAgilisRot','MOVEREL', 1, 1, 10600, 50) #IPC1
  • cm_send('EIBAgilisRot','MOVEREL', 4, 1, 3500, 50) #IPC8
• We changed the thresholds of some INJ signals:
  • RFC threshold: from 2.2 mW to 1.5 mW (MCMir DSP card);
  • IMC threshold: from 18 W to 12 W (IBMar DSP card) and also changed the variables 'LOCK' and 'IMC_trigger' on the VPM ISYS_Acl (lines 231 and 235);
  • FmodErr Z_Calib: from -5e-9 to -1e-8 (on metatron)
• Then, we engaged all loops and everything locked again. 
• Finally, we remeasured the TF of the IMC (see Fig. 1). The UGF is now 93 kHz and 26 degrees of phase margin.
• Fig. 2 shows the FFT of the main INJ signals and no significant difference can be seen.

To be noted that we didn't change the B2 signal, which now is 8.5 mW instead of 12.2 mW.

ITF found in locked arms.

At around 7:00 with Piernicola we went to NE to reactivated the green laser then the commissioning team recovered the DRMI state, see 69700.

Once recovered the ISYS team started working to reduce the input power.

ISC

Etalon setpoints changed and loops enabled:

2026-09-07 09h02m42 UTC    berni 'WI:WI Set point [WI_RH_SET=19.15,rampTime=1]' sent to LSC_Etalon_Acl
2026-09-07 09h02m45 UTC    berni 'WI:Enable servo' sent to LSC_Etalon_Acl
2026-09-07 09h02m52 UTC    berni 'NI:NI Set point [NI_RH_SET=19.53,rampTime=1]' sent to LSC_Etalon_Acl
2026-09-07 09h02m53 UTC    berni 'NI:Enable servo' sent to LSC_Etalon_Acl

TCS

In the first part of the shift Cecilia worked at  TCS TCS AuxCooling System, see entry 69697.

DAQ

At around 1:20 UTC we had a general loss of data: the data from Suspensions and TCS did not recover. At 13:44 UTC I contacted the DAQ on call; investigation in progress.

This afternoon we have updated the following thresholds on the detection system, in accordance with the reduction of the input power from 18 to 12 W:

• we updated the thresholds on SDB2_Photodiodes triggering the fast shutter for B1p_PD1, B1p_PD2 and B1s. All thresholds multiplied by a factor 2/3 in the configuration:

2026-09-07-13h18m00-UTC>INFO...-AcRelayChSet> B1p_PD1_DC_saturation - Trg01 0.55(3)- Trg10 0.39(1.5)
2026-09-07-13h18m00-UTC>INFO...-AcRelayChSet> B1p_PD2_DC_saturation - Trg01 0.78(1.17)- Trg10 0.19(0.29)
2026-09-07-13h18m00-UTC>INFO...-AcRelayChSet> B1s_PD1_DC_saturation - Trg01 30(45)- Trg10 15(22)

• Thresholds in ITF_LOCK.ini for B1p_PD1 have been reduced accordingly at 13h22 utc:
  • in the section SHUTTER: b1p_pd1_thr_lo = 0.3 > 0.2 ; b1p_pd1_thr_hi = 0.5 > 0.33 ; b1p_dc_thr = 0.04 > 0.027
  • in the section ITF_LOCK: b1p_pd1_thr_lo = 1.5 > 1.0 ; b1p_pd1_thr_hi = 3.0 > 2.0

• Thresholds in DET_MAIN.ini have also been reduced by the same factor at 13h24 utc, namely:

max_power_B1p = 0.04 > 0.027, max_power_B1p_PD1 = 0.44 > 0.29, max_power_B1s = 48 > 32

Both DET_MAIN and ITF_LOCK were reloaded at 13h26 utc

We recovered the DRMI state to allow the INJ crew to proceed with the reduction of the injected power. 

We needed to change a few phases that were preventing the lock:

DRMI_1F:

• MICH_phi0: -1.95 -> -2.35 
• PR_TX_GAIN 0.75 -> 0.55

DRMI 3F:

• MICH_phi0: -0.45 -> 3.7

Noticeably, the image of B1p is different from before, as it now looks like there are two beams impinging on the photodiode (Fig.1).

Piernicola mentioned that this double spot shape of B1p in DRMI was already present in August. Figure 1 shows an example showing that.

Today I added insulation to the pipes (pic3, pic 4) and I added gradually to the tank half of the maximum allowed glycol according to the manufacturer (5% in volume, maximum allowed is 10%). 

As in the following plots, a step in temperature is present as a consequence of the fact that the glycol was at room temperature, as well as turbulence, which was expected since the higher viscosity of the glycol with respect to ultra pure water's. I will continue monitoring the system as the solution gets homogeneous.

The additive is needed to slow corrosion and stop biological contamination as summarized in this document and was chosen consulting one of the technicians working at the company producing it, taking into account virgo's history with corrosion (ARCHA analysis). More chemical analysis will follow.

The shift started with the TCS activity - HWS INJ check - in progress.

Tha activity was concluded at around 18:00 UTC, see entry 69694

The ITF was left in LOCKED_ARMS_IR.

Yesterday evening, the last measurement performed with the CH ON showed the CH almost completely outside the HWS map, as shown in Fig. 1. This result was also confirmed by the measurement performed with the RH and reported in entry 69692.

After several movements of EIB_HAR_M0 performed this morning, we reached a configuration in which, after switching on the CH, the HWS map appeared as shown in Fig. 2.

Before re-suspending EIB, we checked the alignment between the HWS beam and the green flashes and found that they were not superimposed, with a clear shift between them.

At the moment, we do not understand why superimposing the HWS beam on the green beam does not result in the HWS beam being centered on the WI mirror.

After re-suspending EIB, we started a new acquisition with the WI RH. The ITF was in single bounce, with the PR aligned and the control on WI MIR engaged.

The RH effect was clearly visible in the HWS map (Fig. 3).

We then waited for about 1.5 hours before switching on the CH, to allow the RH effect to partially decay.

At this point, we acted on the hot mirror of the WI CO2 bench in order to align the CO2 actuators with the RH.

We left the situation as shown in Fig. 4. The CO2 actuators are aligned with the WI mirror within approximately 0.5 cm.

Then, we switched OFF the HWS-INJ SLED.

==================

To be noted: to move the hot mirror, we initially used the new VPM interface developed by Erin. It worked well at first, but at some point the message command failed appeared. We therefore continued the alignment using the old interface on ctrl1.

The TCS work: inj hartman beam alignment, work done inside the laser lab, and WI CO2 centering with the ITF in WI single bounce configuration went on all the morning without any major problem and it is still in progress... 

Air Conditioning
from 7:30UTC to 13:10UTC I set the injection area air conditioning system in "Portata Nominale" to allow the work on hatman beam alignment.

The correct Fig. 4 is the following one.

This morning, as anticipated in the entry 69688, we started with a new HWS-INJ measurement to check the WI CH position after the movements performed yesterday evening using the hot mirror on the WI CO2 bench.

Francesco set the ITF in WI single bounce with PR aligned. 

The HWS-INJ illumination is shown in Fig.1 

07.34 UTC: START HWS ACQUISITION (20260903T0934)

07.38 UTC: WI CH ON (WAVEFRONT 7)

07.48 UTC: WI CH OFF (WAVEFRONT 25)

This measurement confirmed that the CO2 actuators were reasonably well aligned with the RH.

At this point, we went to the Laser Lab with Suzanne and Piernicola to try to improve the centering of the HWS-INJ beam on the WI mirror.

As a first attempt, we performed a pure shift using EIB_HAR_M0 (see Fig. 3). The vertical alignment was easily corrected by a small counterclockwise rotation of the mirror (see Fig. 4).

The horizontal alignment was more challenging. We tried moving the horizontal knob in both directions, but we were not able to obtain the required displacement.

We then tried to recover the alignment using the usual procedure, by superimposing the HWS beam on the green beam. Although we were able to achieve a good overlap, the beam was still clipped on the CCD and no further horizontal adjustment was possible, as the corresponding knob had reached the end of its travel range.

At this point, we rotated the mirror around its axis to move the adjustment away from the end of the travel range. We then realigned the HWS beam with the green beam and also checked the beam reflected back from the WE. The returning beam is shown in Fig. 4.

At this stage, the alignment appeared to be recovered. However, a new HWS acquisition with the WI RH showed that the RH signature was still approximately in the same position as at the beginning (see Fig. 5).

Tomorrow morning we will try to optimize the HWS beam alignment around this position.

To be noted: in order to ease the alignment of the HWS beam on the green, we temporarily removed the automatic check of the presence of the CEB GREEN BEAM in ARM_LOCK automation node (lines commented at #311 and #828). To be reverted once this work will be finished.

14:00 UTC ITF in DOWN state with commissioning group still working in Laser Lab (Corubolo, Melo, Nardecchia, Spinicelli)

20:00 UTC WI RH switched on. Additional control loop on the WI MIR. EIB left on ground, INJ locked.

20:05 UTC INJ Lab air flow set to Portata Ridotta

21:00 UTC ITF left in SINGLE_BOUNCE_WI with commissioning group still working on the lock of the arms.

ITF found in LOCKED_ARMS_IR

At 7:30 UTC  ITF in SINGLE_BOUNCE_WI  with PR aligned.

At around 8:30 UTC the commissioning team went in the Laser Lab; actvity still in progress.

INF

At 12:30 UTC UTA INJ in "Portata Nominale" under request of the commissioning crew.

The /virgoLog disk area was at more than 80% since severals weeks . The investigations made the 2026-08-28 gives the following report 

Filesystem          1K-blocks      Used Available Use% Mounted on
virgodata:/virgoLog 361112576 300837888  41928704  88% /virgoLog

Scan /virgoLog/VirgoOnline directory
191608888       /virgoLog/VirgoOnline
33953916        /virgoLog/VirgoOnline/NNC_NEB
33939604        /virgoLog/VirgoOnline/NNC_CEB_LEV1
21681556        /virgoLog/VirgoOnline/BaDAQ

8734112    /virgoLog/VirgoOnline/FFLMoni55

All the files related to theses 4 servers with a size greater than 1GB were moved from /virgoLog/VirgoOnline to /data/dev/logs/VirgoOnline/ .

The attached  plot shows  the virgoLog rate and usage since the beginning of the year(2026) upto today