ITF in UPGRADING - DOWN

Today's activities:

Arms recovery (Bersanetti, Boldrini, Majorana, Ruggi, Spinicelli, Was)

TCS auxiliary cooling system cabling - new data channels (Cavalieri)

ALS recovery (Lagabbe, Spinicelli)

The attached shows the excitation produced by opening and closing the doors of the SAS and piscina. The excited bump looks well matched to the 18.50 Hz one. Some small excitation of the 12Hz bump is noticeable as well.

We had a look to the effect of the sound wave pulses produced by repeated door opening and closing.

In the morning, while the HVAC was on, we performed a set of pulses with the door between the laser bench room and the minitower room. Figure 1 shows ASD of the microphones inside the laser bench room. They all see a bump at approximately 20Hz growing above the noise backround noise. We remind that the "LLR" microphone was positioned in the N-W corner of the room, at aproximately the beam height.

In the afternoon we repeated the same test, but the HVAC was off (also those of CEB) and we acted with the door between the Atrium and the laser bench. In Figure 2 (and Fig.3) the purple spectra correspond to the quiet condition. Two bumpy structures at approximately 12 Hz and 18.5 Hz are visible, in particular in the LLR microphone. In the excited condition (blue) microphones see again a peak at about 20Hz and also, less pronounced, growing bumps at roughly 40, 45, 55, 85,135 Hz (Figure 2).

Figure 4 (also 5 and 6) is a spectrogram of LLR mic over the time of this test. After the test is finished (about 13:13:30 UTC) additional pulses around 20Hz are seen (actually here the excited frequency is a bit lower, it looks more around 18.5Hz). These are due to Irene exiting the lab and opening/closing the doors of the atrium, the SAS and the piscina.

We would interpret these frequencies as acoustic modes of the room. We are also tempted to think that the 18.5 Hz bump seen in quiet condition is the same 20Hz mode excited by the door.  We have excluded already it to be a resonant mode of the air ducts ( https://logbook.virgo-gw.eu/virgo/?r=69148). The frequency mismatch (20Hz instead of 18.5Hz) however needs an explanation.

In this logbook entry we will explain how the alignment of the North End Mirror has been done, using data from the Instrumented baffle.

The baffle interface shows the power and radial angle positions of the 120 sensors present. The baffle has 6 different sections, each of them with 4 columns and 5 rows of sensors.

The procedure followed was to:

• Center the beam in one column of sensors and maximize the total power in that column, to make sure that the beam is directy hitting the sensors, not around them (Fig.1)
• Take note of the power of each sensor in that column, and of the note of TX and TY coordinates of the North End 
• Redirect the beam into the other side of the baffle until you hit the opposite column and repeat the procedure, trying to match the detected powers to approximate the same (specular) position of the beam on the new column. Take note of the NE TX and TY coordinates
• Move the NE at the halfway points on TX and TY

Note that the power when you are above a sensor is of approximately 30 micro watts.

The current reference NTP time for some linux machines, including RTPCs, is taken from the local windows domain servers, backed by standard ntp references on the Internet.
It turns out that their pace when the remote references are unreachable is misbehaving.

There is no need to add yet another GPS receiver, for the purposes of the standard linux machines clock time (not the Virgo data frames time) It should be enough to use properly local hardware clocks.

We are going to check the windows domain servers time synchronization configuration in order to handle this corner case correctly.

This morning we continued the sanity check of the ALS beam at the ends. We started checking the status of the ALS fibers at the WE,without finding any particular issue. However, the IR power coming from the fiber was still too low to seed the amplifier. 

We then went in the DAQ room were the pickoff fiber arriving from EIB is then 50:50 splitted in two beam for north and west end. At the input of the splitter, there was another fiber splitter (90:10, see fig. 1), used in the past to reduce the power without affecting the phase camera and squeezing pickoffs. As the ALS pickoff is now indipendent from the others, we decided to remove it. As a result, the infrared beam was powerfull enough to close both WARM and NARM pstab and to seed both amplifier.

Moreover, we noticed that the ALS CEB green power was also quite low (<10mW), indicating a misalignment of the beam on EIB, most probably due to the work on the BPL/BPC loops of the last weeks. 

A realignment of the green beam on EIB will be done in the next days.

ITF DOWN in UPGRADING mode.

Planned activities:
- 06:30 UTC - NI Cryotrap valve open. Laser beam on north arm.
- 14:30 UTC - WI Cryotrap valve open. Laser beam on west arm.
- North arm alignment (still in progress...)

06:30 UTC - NI Cryotrap valve open. Laser beam on north arm.
14:30 UTC - WI Cryotrap valve open. Laser beam on west arm.

Similar tests for studying acoustic modes of the laser bench room were made back in 2024 but with the HVAC on and no much effect was seen:  https://logbook.virgo-gw.eu/virgo/?r=65126

The microphone added externally of the laser bench room is named ENV_CEB_LLR_MIC, the microphone insided the is named ENV_LLR_MIC.

In order to see the main laser beam on NE mirror and to re-align it, the pcal NE has been switched off on July 29 at 12:19 utc.

Before attempting the relock of the arms foreseen next days today I tried to asses the actuation balancing of the WI and NI payload at the level of the MAR, due to the new mounted payloads.

The coupling between Z corr and TY corr have been evaluated through some DC steps applied in Z corr, in order to compute the correct driving weigths to minimize such coupling. Both for NI and WI the misbalancing was quite high.

For WI the found value was 0.0046 (from 0.0002), directly applied in the branch that feeds part of the Z corr in TY MAR corr, see Fig.1. Further fine tuning are not appreciable since we are covered by corr signal noise.

Once we have a lock, by measuring some TF between the corrections, it will be easier to estimate the finer weights.

For NI the found value was quite high with respect to the current one, but I didn't implemented it since for NI is not possible to see at a first sights the effect of the driving adjustment due to the way the corr probes are implemented in the DSP card. So I will leave it for tomorrow.

Same decoupling procedure has been applied also to minimize z2roll coupling z-> tz corr, fig.2. (WI weight: from 0.0115 to 0.0104; NI weigth: from 0.00374 to -0.01326)

N.B. MIR coil balancing both for NI and WI need yet to be done. 

Motivation: the first scope is to study the sound transmission between the CEB hall and the two labs: in order to quantify the contribution from the CEB HVACs this time we first switch off the INJ and DET labs HVAC and THEN the hall and CleanRooms HVACs. Also, we perform a white noise injection with one of the two sub-woofer positioned in the DET terrace. The second purpose is to study the origin of the low frequency peaks at approximately 12 Hz and 19 Hz which are clearly seen by microphones when the INJ HVAC is off (previous investigations: elog https://logbook.virgo-gw.eu/virgo/?r=69148).

Preparation (this morning):

• we moved the GRAS microphone inside the laser bench room (the microphone was laid onto the fan coil at the left of the entrance). We positioned it on a tripod in the corner of the bench room at the right side of the entrance.
• we installed a second GRAS AZ46 microphone on a tripod inside the hall on the west side of the INJ tower (nearby the EEroom). In the attached picture.

Action times:

• 12:35 UTC - DET OFF
• 12:38 UTC - INJ OFF 
• 12:45 UTC - CEB CLEAN ROOM OFF 
• 12:50 UTC - CEB HALL OFF
• 12:58 UTC - acoustic injection in CEB hall (acoustic injection file: /virgoData/NoiseInjections/AcousticInjectionsO4/output/AcousticColored_NOISE_CEB_DetTerrace-1469278756.txt)
• 13:05:50 to 13:11:30 UTC - open (and kept open) the acoustic door between Atrium and LaserLab
• 13:11:30 to 13:13:00 UTC - open and close repeatedly the atrium door to generate sound pulses inside the INJ bench room
• 13:16 UTC - clean room ON 
• 13:18 UTC - CEB hall ON
• 13:21 UTC  - DET AHU ON
• 13:24 UTC - INJ AHU ON

The laser lab temperature increased by 1deg overall as shown in Figure 2. The weather conditions were good, with low wind (also in Fig.2). Analyses will follow.

As rotors were moved, the hInj values change.

All values have been updated in NCalMoni.cfg,  /virgoData/NCal/set_NEB_default_lines.sh and /virgoData/NCal/set_WEB_default_lines.sh. after new fromage simulation.

Values will need some more tuning as:

-WEB mirror position needs to be determined

-NEB rotor survey was not complete

On Friday, July 24, we performed an environmental noise characterization of one of the two chillers of the auxiliary CO2 laser cooling system. Since the installation had not yet been completed, the chiller was connected only to the tank, with both components located in the TCS chiller room.
A microphone was installed on the right-hand side of the room, immediately to the right of the entrance. An accelerometer and a triaxial magnetometer were instead installed, in turn, first on the floor next to the chiller and then directly on top of the chiller, Figure 1-6. The sequence and timing of the different measurement configurations are reported in the attached file.

Noise inside the TCS chiller room

*** Seismic ***

Figure 7 compares the vibration spectra measured under the different operating conditions. The highest vibration levels are measured when the accelerometer is mounted directly on top of the chiller (yellow and violet curves). Even when the cooling fan is not operating (violet curve), the vibration level measured on top of the chiller remains significantly higher than that measured on the floor, indicating that other components of the chiller continue to generate vibrations. Only after the chiller is completely switched off does the vibration level measured on top decrease by several orders of magnitude, becoming comparable to the floor measurements.

When the cooling fan is operating, spectral lines are observed at approximately ~22.5 Hz and ~112.5 Hz. The 22.5 Hz component is consistent with the fan rotational frequency, while the 112.5 Hz line corresponds to the blade-passing frequency  (Fbpf=Npale*Frot) of a five-blade fan. These two spectral features are consistently observed in both the acoustic and magnetic measurements, Figure 8, 9.

Additional spectral lines are observed at approximately ~47.8 Hz and ~48.9 Hz. The 47.8 Hz line is present both with the cooling fan on and off and disappears only when the chiller is completely switched off, indicating that it is associated with a component operating continuously while the chiller is powered.
Following the inspection carried out by Cecilia, the circulation pump nameplate was found to report a nominal rotational speed of 2850 rpm, corresponding to a rotational frequency of 47.5 Hz, in good agreement with the measured line. This supports the attribution of the 47.8 Hz feature to the circulation pump. The 47.8 Hz spectral line was also observed when the accelerometer was mounted on the supply water pipe connecting the chiller to the tank, Figure 22.

The ~48.9 Hz line is observed only during the fan-on periods (blue and yellow curves in Figure 10 ), suggesting that it is associated with a component operating only during the cooling phase. Both spectral features are also identified in the acoustic and magnetic measurements, Figure 11.
 

*** Acoustic ***
The acoustic spectra show no significant variation below approximately 40 Hz. During the cooling fan-on condition (blue and yellow curve), a clear increase in acoustic noise is observed mainly above about ~40 Hz, with both a broadband contribution and several narrow spectral features, Figure 12.
 

***Magnetic ***
The three components of the triaxial magnetometer exhibit a similar overall behaviour, Figure 13,14,15.. When the cooling fan is operating and the magnetometer is placed directly on top of the chiller (yellow curve), a pronounced spectral comb with an approximately 2.5 Hz spacing is observed over a broad frequency range. A similar spectral structure is also visible when the magnetometer is placed on the floor (blue curves), although it is less regular and has significantly lower amplitudes.

In the ~1–3 Hz range, an increase in magnetic noise is observed both during the cooling fan-on and fan-off conditions compared with the completely switched-off chiller. The low-frequency magnetic contribution could be generated by components that remain active while the chiller is powered.

Noise outside the TCS chiller room

The following observations summarize the environmental noise signatures associated with the operation of the auxiliary CO2 laser chiller outside the TCS chiller room.

• No signatures correlated with the chiller operation are observed in the UPS/IPS current and voltage monitors, Figure 16.
• An increase in acoustic noise is observed only inside the TCS room (NN array: microphones #216 and #217). No corresponding signatures are observed in the microphones installed in the experimental halls, Figure 17 and 18.
• Clear signatures associated with the cooling fan operation are observed by the NN geophones installed inside the TCS room (#218–#221). A weaker but still identifiable signature is detected by the NN geophones located around the NI (#204, #205) and WI (#212, #213) towers.  Apart from these sensors, no significant increase in vibration is observed elsewhere. Figure 19 and 20.
• While the magnetometer installed inside the TCS chiller room shows several spectral features associated with the chiller operation, only a single narrow spectral line at approximately 27.5 Hz (11th harmonic of the 2.5 Hz spectral comb) is detected outside the room by the NI, WI, BS and CEB magnetometers. No evidence of the spectral comb or of the other characteristic magnetic signatures observed locally is found in the magnetometers, Figure 21

Today we moved both BS and WI using the old references of the optical levers, and we could find also the WI-reflected beam on the B4 camera. Both reflected beams are now visible (the third one is the second reflection of the NI one).

Actvities communicated in control room:

• maintenance activities in the morning;
• NEB+WEB bridge crane maintenance;
• AHU switch off test in LLab, DET and CEB ;

At the end of the shift Diego started to work at NI/WI mirror alignment.

SBE

SWEB and SNEB opened by the guardian; properly closed at 9:42 UTC.

We tried to recover the alignment of the NI using as a reference the camera image on B4 and the DC power signal. We explored several position both of NI and WI, by moving the local controls (and PR and BS for checks). We maybe were able to find the main beam on B4 cam. Tomorrow we will continue.

The shift was covered first by Lunghini and then by Zaza

Lunghini's report:

ITF in UPGRADING Mode and DOWN State.
All times are UTC.
Below the activities communicated to the Control Room:
07:35 - 07:46 TCS: Update NE/WE RH Powers and Turning ON SR RH (Lumaca, Corubolo, #69444);
08:49 VAC: Re opening of NI-BS, NI-PR, and NI Cryotrap Large Valves (Vacuum Team, #69446);
10:36 NEB/WEB safety laser inspections (Galimberti);
10:36 ALS: NEB turning ON amplifier (Lagabbe);
12:55 Re opening of BS-SR, BS-WI Large Valves (Vacuum Team, #69450);

Zaza's report:

13:45 UTC SDB1 camera realignment (Boldrini, Lagabbe)
14:00 UTC WE, NE RH V flags thresholds adjusted (Zaza, Corubolo, Lumaca)
6:00 UTC on: old TCS auxiliary cooling system removed, new system connected to the lasers (Ciardelli, Menzione, Zaza). 
15:20 ALS: WEB turning ON amplifier (Lagabbe)

On July 26, at 15:08 UTC, the IMC unclocked, which causes the BPC error signals to go away since they follow the IMC reflected signal.

After investigation, it appears that the PMC transmitted power signal started oscillating around 15 s before the IMC unlock. Which causes both the IMC and PMC to unlock. The PMC relock automatically a few second after the unlock, but the IMC could not because the BPC misaligned the laser beam with the IMC.

We don't know yet what causes these oscillations. The PCM error signal was stable before the unlock, as well as the slave laser cavity. The master laser power signal and frequency correction signals seem stable enough. The PSTAB tried to compensate for these oscillations, but at some point, the AOM started to saturate.

We don't know neither why the BPC did not come back to its previous position after the IMC unlock.

On July 27 at 8h08 utc, the BPC loop filters have been reset and the IMC and RFC have been relocked.

The ipcamz3 camera was re-oriented in order to better view the sdb1 bench.

To access the camera view:  http://ipcamz3/ 

login: root

password: tocaxis

The Vacuum Team it's going to reopen the following Valves as planned:
- BS-NI Large Valve;
- BS-PR Large Valve;
- NI Cryotrap Valve;

The beam will be able to reach NI Tower.

The Vacuum Team it's going to reopen the following Valves as planned:
- BS-SR Large Valve;
- BS-WI Large Valve;

The beam will be able to reach WI and SR Tower.

Retune the offset of the NEB north and West NCal tremerature sensor.

After NCal maintenance operation (see lgbook https://logbook.virgo-gw.eu/virgo/?r=69385), the temperature sensors of NNN and NWN changed their calibration (see https://vim-online.virgo-gw.eu/resources/2026-07-20/resources/20260720_cal_ncal_temp.png).
Between 8:40 and 9:03 UTC this morning, I modified offsets in /virgoData/VirgoOnline/SNEB_dbox_rack.cfg to recover a 22.5° += 0.1° on all sensors when rotors are off.

New values are :

July 23rd entry, published on July 27th due to network issues on the 23rd

ITF in UPGRADING, DOWN

Activities communicated to the control room:

CEB laser safety inspection (Galimberti)
TCS aux cooling system final assembly (coarse control side) in preparation for tomorrow's ENV tests (Ciardelli, Gherardini, Menzione, Zaza)

At 7.40 UTC, the SR RH was switched ON and the voltages of the RHs on the terminal mirrors were updated as follows: 

• NE RH set to updated value (from 16.60 V to 18.32 V)
• WE RH set to updated value (from 17.40 V to 19.10 V)
• SR RH set to standard value (23.13 V)