Figure 1. The ring heater temperature gives us with a little more precision the time of the opening of the cryo trap valves. It was between 8:25 UTC and 8:26 UTC.

ITF found in UPGRADING Mode with Arm FSR scan of cryotrap transient commissioning activity ongoing (Cjacquet). Activity concluded at 17:07 UTC. Logbook entry #68417.
Under request of Pasqualetti I added MOBILE3, under Particles, set of flags alarmed in the DMS.
ITF left in UPGRADING Mode with the arms locked in the IR.

This logbook presents the scans made during the thermal transient of the west arm cryotrap. At 8:30 UTC the vavles have been opened (68410). Scans started at 9:29UTC, below is the list of scans done. After each scan the cavity are locked on IR, then on green before the start of the next serie of scans.

Observations during the shift

The first series of scans (from 9:29UTC to 11:06UTC) stopped before the end due to an unlock of the green. In the scan series done until the end between 11:09UTC and 13:50UTC, the first order mode in the west cavity seems to increase a lot over time, suggesting that the cavity was misaligning during the scans. This produced an unexpected result concerning the mode spacing in scan. The resonance position was shifted in the opposite direction compare to the expectation regarding the transient produced by the cryotrap (See figure 1 showing an example for the second order mode). It is possible to remove the divergent points by keeping only the data where P_HG01<0.05mW. This give the figure 2. After 13:53 and until the end of the shift, I stopped manually the scan when the observed amplitude of the HG01 was reaching 0.04mW or more. 

These figures illustrates details of the peaks that are excited during tappings on SDB1 chamber South flange (data of Nov 28, 9:28:50, 120s - purple curve) and during one acoustic injection in CEB (data of Nov 29, 12:19:10, 180s - green curve).

Note that some of the peaks are also present (small) in the quiet hrec.

In the txt file is a list of the center frequency of these peaks. The "*" indicates that the peak is also seen in quiet hrec, "T" if it is excited by tapping SDB1, "A" if it is excited by acoustic noise in CEB.

The conclusion we draw in https://logbook.virgo-gw.eu/virgo/?r=67936 is not correct. Instead, the correct conclusion is that the family of peaks that arise in hrec when injecting acoustic noise in the CEB hall are associated to vibration of SDB1 chamber.

The attached figures show the effect in hrec and noise levels in microphones of CEB hall and inside DT lab (EDB) and vibration of SDB1 chamber during:  one acoustic injection inside the CEB hall (Figure 1), one acoustic injection inside DT lab (Figure 2) and tappings of SDB1 chamber (Figure 3). 

Similar peaks are excited in hrec. Comparing env probes is evident that the driver of the peaks in hrec is the level of vibration of SDB1 chamber.

At NE, on Wednesday 17 December 2025

Worked on the NE Pcal reflection bench, to investigate possible sources of optical losses (expected at the level of <0.9%) from the steering mirror first, and from the viewport then.

• 13:05.00 UTC start activity, open reflection bench cover.
• 13:11.14 -> reduced the angle of incidence of the beam on the steering mirror (and moving the sphere to keep the beam centered)
• 13.15.55 -> put back to initial positions.
• 13:21 -> put back the cover on the bench
• 13:24.10 -> switch off pcal NE lines
• 13.26 -> reopen the bench cover
• 13.29 -> reduced again the aoi, even a bit more than first time (a bit smaller, cannot do more otherwise the input beam is clipped) 
• 13.31.24 -> increased the distance between the sphere and the input beam by ~2 mm (without changing the aoi on the steering mirror, hence beam less well centered on the sphere)
• 13.34.15 -> slightly increased the AOI to center again the beam in the sphere input 
• 13.37.40 -> went back to the initial positions.
• 13.44 -> increased the AOI, moved the sphere by about 2 cm (increasing it distance to the viewport)
• 13.48.45 -> increased the aperture of the diaphragm installed on the viewport.

• 14h25 UTC ->  stop laser, visual inspection of the viewport; presence of scratches at the bottom at least, and of dust.
• 14h30 UTC -> blow viewport with compressed air (during this action, the sphere was not attached on the bench, it may have been moved a bit, even if we do not think so). Note that due to limited access, the air could not be blown from the side of the viewport, but mainly from the front, with a small angle. Also the mount to support the diaphragm may prevent the dust to go away easily when doing this action.
• 14h32 UTC -> switch on laser and loop
• 14h35 UTC -> stop laser
• 14.40 UTC -> cleaned viewport with tissue  (during this action, the sphere was touched, and then recentered on the beam). Again, difficult access, so cleaning doing some rotation movements around the center.
• 14h45 UTC -> switch on laser and loop
• 15h10 UTC -> laser stopped to dismount the sphere and mirror, and bring them back to LAPP.
• 15h15 UTC -> took some pictures (se figures 3 and 4)

The two first figures show the trend of the two photodiodes (Tx_PD2 in loop) and of the sphere during this period.

• there was no significant change of the power measured by the sphere when changing the aoi on the steering mirror.
• after blowing air on the viewport, the power measured by the sphere was reduced by 2% to 3% (...but we have a doubt to have possibly misaligned the sphere when blowing the air without noticing).
• after passing a clean tissue on the viewport, the power measured by the sphere may have increased by about 0.1% (see the last part of the figure 2).
• note that the viewport was not clean even after these actions, as at least dust was still visible and also some thin layer, even if possibly less around the center itself.

Visual inspection of the viewport (also see the pictures 3 and 4): there are scratches at least at the bottom of the viewport, quite some dust. Before cleaning with the tissue,  we could see a layer of something on the viewport, which was less visible around the center after the tentative cleaning. However, the viewport is difficult to access for cleaning, and to observe, because of the supporting structure of the diaphram in front of it. 

In the Rx calibration done for the previous logbook entry, figure 7 is correct for the top plot (Rx / Tx_PD1) but not for the bottom plot (Rx / Tx_PD2): in the analysis code, the correction to be applied to take into account the new calibration of the phototiode was applied only on the Tx_PD1 photodiode, not on the Tx_PD2 one.

The code has been modified to correct for both now. The initial "old gain" of the Rx sphere is the one being set in March 2024. The update figure is given below, and the results summarized in the table:

For NE PCal, the photodiode Tx_PD1 has a slight dependence on humidity, hence the reference calibration used for long term monitoring is based on Tx_PD2, in bold in the table.

ITF found in UPGRADING Mode and North Arm locked.

At around 8:30 UTC the WEST valves have been opened; I manually realigned the cavity and then started the planned activity of arm FSR scan.

Activity still in progress.

WE CALIBRATION ON 2025_12_17
On Wednesday, 17th, we did calibration activities on the WE PCal.
We first performed the installation of WSV at WE at approximately 7h45 UTC, switched off the laser and started the installation of WSV on the reflection bench.

ADC channel calibration
- Calibration for the temperature channel 5 ADC 0 (8h30 UTC)
with a multifunction calibrator Time series connected to the input of the follower circuit for the PCAL_WE_WSV_temp channel. Here the list of actions performed:
         • Changing the channel gain to 200 and offset to 0. in the SWEB_dbox_rack config
         • Signal injection from 0 to -6V with 2V steps with the Time Series
         • Computing the calibration coefficient (gain and offset) with a linear regression
         • updating the coefficient in the config at approximately 8h50  UTC

- Calibration for the power channel ch6 ADC 0 (8h55 UTC):
with a multifunction calibrator Time series connected to the input of the follower circuit for the PCAL_WE_WSV_DC channel. Here the list of the actions performed:

• Changing the channel gain to 2. and offset to 0. in the SNEB_dbox_rack config
• Signal injection from 0 to -6V with 1V steps with the Time Series
• Computing the calibration coefficient (gain and offset) with a linear regression
• updating the coefficient in the config at approximately 9h10 UTC

 
The following table contains the old/new gains and offset for each calibration performed:

See figures 4 & 5 for the linear regression plots (POWER and TEMP)

 
 
WE Tx_PD1/2 vs WSV calibration measurement -

WSV gain used as reference : -2.617869 V/W

• 09:32 UTC -> background measurement (still wait 30’ for stabilization)
• 1450000830.0 (10 UTC) -> START Background measurement
• LASER ON: 11:04 UTC (1450005318) (wait for thermalisation 30 min), until 16h05 UTC

Here is the command used: python3 photodiode_calibration.py 1450007100 16200 1450000830 3600 WE

The following table contains the old/new gains and offset for each photodiode:

WE Rx vs Tx_PD1 and WE Rx vs Tx_PD2 calibrations
Around 16h15 UTC, removed WSV and put back Rx sphere
Background measurement: 16h40 UTC (1450024800), for 2 hours
At 18h37 UTC, laser switched on
Measurement: 1450034000, for ~6h
Command launched: python3 Rx_sphere_calibration.py 1450034000 22000 1450029000 2600 WE
with the following correction coefficients (these coefficients are used in the code to take into account the new Tx_PD calibration gains, but without changing the ADC gains themselves for long term monitoring):
coeff = -0.739903/-0.740169,-0.784602/-0.783431    #new/old Tx_PD1 and Tx_PD2 WE, 2025/12/17
 

(See figures 6&7 for the ratios and background levels)

Two possible results, depending on which photodiode is used as reference (Tx_PD1 used as reference for WE, since PD2 have a sligth dependence with humidity):
 
1/ Rx sphere new gain (w.r.t Tx_PD1): 0.689194 W/V    offset -0.001159 W
correction factors:
    Rx_sphere: 0.997907 +/- 0.000053
 
2/ Rx sphere new gain (w.r.t Tx_PD2): 0.689149 W/V    offset -0.001164 W
correction factors:
    Rx_sphere: 0.997843 +/- 0.000054

 

In the frame of the ALS shift we had to block the beam on the LB in order to block the EIB.

We found that the flip mirror at the output of the PMC is not working, nor from the VPM (TTL_inj process, CH3) or directly pushing the button.

While trying to debug it, the cable of the flip mirror accidentally passed through the beam in front of the input window of the PMC. Of course it got burnt and made some plastic dust. Few seconds later, we observed a decrease of the PMC TRA power by 10% (from 0.8V to 0.72V ). All the missing power seems to be in the order 2 mode, so it seems to be pure mismatch (plot 1).

The IMC TRA power has been adjusted to 18W by the Automation with the IPC1. We compared the spectra of the different inj loops before and after the accident (plot 2), and since there are no major differences we decided to keep everything like this and monitor the transmission of the PMC in the next days. 

This morning we wanted to make a characterization of the green beam coming from the north arm (the west being unavailable).

We blocked the EIB and relocked INJ, the north arm and ALS in reflection, but the IMC started to be very unstable and unlocking (we had to manually realign the bench and the mc mirror, but it kept unlocking).

We then decided to reverse all the operations and relock the injection in a stable way.

After speaking with Paolo we found out that a possibility could be to switch back to the old bpc alignment strategy (by setting to 0 IMC_ON in the BPC DSP card) when the EIB is blocked, in which the references are only the quadrants on the EIB (while now the beam follows SIB1 in tz and ty once the IMC is locked).

ITF found in UPGRADING Mode and North Arm locked.

At around 10:30 UTC Camilla and Mathieu started to work in laser lab; activity concluded at around 13:00 UTC.

Yesterday afternoon, we have dismounted the NE_Rx sphere of the NE PCal, to bring it back to LAPP for extra-noise investigation. We also took the steering mirror to check its reflection vs aoi at LAPP.

We put a beam dump on the bench, see picture.

We also stopped the PCal laser. The laser driver is still on, but the pump diode is switched off.

ITF found in UPGRADING Mode and North Arm locked. I found SIB2_SBE, FCIM_SBE, and MC_MAR_Z_ON loops open.
At 06:59 UTC I closed SIB2_SBE.
At 07:01 UTC I closed FCIM_SBE.
From 8:21 to 9:08 UTC De Laurentis and Gargiulo went in DET Lab for wafers replacement.
At 10:46 UTC MC_MAR_Z_ON loop closed by Ruggi.
From 13:30 UTC Tringali went to WEB to disconnect microphones of newtonian array.

The power supply Hameg29 network interface was configured in USB. We have configured it in DHCP around 14h00 UTC, and it was then available from the remote control process PCal_PowerSupply.

Yesterday we did calibration activities on the NE PCal.

We first performed the installation of WSV at NE at approximately 8h UTC, switched off the laser and started the installation of WSV on the reflection bench.

ADC channel calibration NE:
- Calibration for the temperature channel 5 ADC 0 (10h10 UTC):
with a multifunction calibrator Time series connected to the input of the follower circuit for the PCAL_NE_WSV_temp channel. Here the list of actions performed:
• Changing the channel gain to 200 and offset to 0. in the SNEB_dbox_rack config
• Signal injection from 0 to -6V with 2V steps with the Time Series
• Computing the calibration coefficient (gain and offset) with a linear regression
• updating the coefficient in the config at approximately 10h30  UTC

- Calibration for the power channel ch6 ADC 0 (10h40 UTC):
with a multifunction calibrator Time series connected to the input of the follower circuit for the PCAL_NE_WSV_DC channel. Here the list of the actions performed:

• Changing the channel gain to 2. and offset to 0. in the SNEB_dbox_rack config

• Signal injection from 0 to -6V with 1V steps with the Time Series

• Computing the calibration coefficient (gain and offset) with a linear regression

• updating the coefficient in the config at approximately 12h00 UTC

You will find attached the linear regression associated to both the temperature and power calibration (see figure 1&2).
Here is the tab summaring the old&new values of the gain and offset:

NE Tx_PD1/2 calibration vs WSV measurement 

• Start of background measurement : 11h30 (UTC) for 1h

• Laser turned on at 1.3W : 13h30 UTC

We will have a first glance at the data around 16h00 UTC to see the quality of this latter. We checked the results at about 16h15 UTC and the values were consistent with what was expected (data were stable all along the measurements), we thus interrupted the measurement and run the analysis scripts from which we extracted the following gains/offsets (see figures 3,4,5)

Here is the command used: python3 photodiode_calibration.py 1449930618 5400 1449919818 3600 NE

NE Rx calibration vs Tx_PD1 photodiode

BACKGROUND START 18 UTC (for 1 hour) (see figure 6)

LASER ON 20.50 

MEASUREMENT START AT 22 UTC (for 8 hours) (see figure 7 for the ratio measurements)

Here is the command used:  python3 Rx_sphere_calibration.py 1449957618 28800 1449943218 3600 NE

(using Tx_PD1 as reference is hard-coded)

Finally, the analysis results are as follows:

• Rx sphere new gain : 0.729448, new offset : -0.002623 (using Tx_PD1)
• Rx sphere old gain :  0.730861 , old offset : -0.002632

On the ADC configuration, we updated only the ADC voltage calibration used for the two WSV channels (temperature and power), but we did not update any gains nor offsets for the Tx_PD1,PD2 or Rx photodiodes: all the values are still as they were during O4, since March 2024.

Figure 1. After 17 hours the temperature transient is still ongoing, this means that there is no point in closing the north arm cryotrap valves on Thursday evening, as the north arm will not have the time to arrive at a steady state my Friday morning. The temperature change is more than one degree Kelvin, this should allow for an estimation of the cooling power the cryotrap. For example assuming that the mirrors are perfect black bodies (which they probably are not so it is an approximation) the power needed to change the temperature by 1.5 degrees is 3.4 Watts.

ITF found in upgrading mode.

At around 15:30 UTC Romain started to work on SDB1 (68396); activity concluded at around 16:30 UTC.

At around 15:37 UTC Antonio closed the WEST valves ( 68398 ).

In parallel the CAL team worked at the Pcal power calibration.

All the activities concluded at 19:00 UTC.

SBE

SIB2 opened many times by the guardian; properly closed.

Today around 15:35 UTC the west arm cryotrap valves (both WI and WE) have been closed, so that the cryotrap stop cooling the west arm mirrors. Beforehand the WI etalon loop was open, its correction was already at zero since 11:00UTC. 

Figure 1 shows the effect, the temperature of WI has clearly started increasing. We will see tomorrow what is the timescale of the warming, once the temperature stabilizes.

The goal is to measure the effect of the cryotrap on the IM/EM radius of curvature. Simulations reported in the Advanced Virgo TDR mention it is 2m. That wouldn't be enough to explain the 7m discrepency measured at Virgo compared to LMA measurements, but maybe the cryotrap effect is larger than expected due to cooling of other parts of the vacuum system than just the cryotrap.

The goal of the measurement was to measure the angular transfer functions from the marionetta to the SDB1 bench (a first attempt was made in https://logbook.virgo-gw.eu/virgo/?r=68232 ).

SDB1 bench angular control closed at 15h34m41 utc.

Noise injection on SDB1 TX started at 15h50m23-UTC.

We adjust the amplitude of the noise at 15h53m50 utc and start collecting some data (FIG.1) for 6 min.

Then we change the shape of the noise to cover some slightly higher frequencies: new data taking starting at 16h03m57-UTC (FIG.2), for 6 min.

We inject a noise in TY starting at 16h12m58-UTC (see FIG.3) but we do not see the effect on the bench LVDT.

After adjusting the shape and amplitude of the noise, we obtain the noise spectrum shown in FIG.4. Collecting data from 16h22m16-UTC, for 6 min.

Report of the shift of December 10th.

Motivations:

Task of the shift was to evaluate the level of the DAC noise of the input marionette, in particular if its contribution to the low-frequency noise budget is somehow limiting or not. In order to derive such information, we worked in Low noise configuration, reallocating the locking actuation from the input marionettes, to the END ones.

Once in this configuration, the idea was to implement an active subtraction filter (gamma) to reduce the coupling among CARM and DARM loops, i.e. the driving amount that from the common CARM actuation, goes into the DARM actuation.
With the aim to reduce enough this noise contribution, we wanted to eventually derive an upper limit of the input actuation noise, having the comparison of two istances in which the input MAR relays were respectively closed and open, i.e. the High power configuration and the low-noise configuration, respectively.

To be noted that all the actions performed in this configuration couldn't be appreciated on Hrec, since there's a possible miscalibration of the END actuators models, which affects the good reconstruction of the sensitivity.

Hence, profiting of good injections on DARM, all the projection have been made towards a reconstructed sensitivity starting from a calibrated DARM in [m], and afterwards in strain.

Recap of the shift:

Once in the configuration reallocated to the END marionette, we performed a noise injection on CARM loop in order to measure the coupling to DARM. Given the hypothesis that the coupling, at least in the region 10-50 Hz can be assumed frequency independent, we implemented a filter which is basically a static gain with value equal to the measured coupling. 

The gamma subtraction filter has been obtained trough the relationship:

Gamma = DARM_corr / CARM_SLOW_corr *(1-OL_darm) / DCP / OL_darm.

where DARM_corr / CARM_SLOW_corr is the TF during the CARM measurement, while OL_darm and DCP are obtained from the model used to calibrate the spectra of DARM.
Those terms are necessary in order to get rid of the fact that during the injection the measured coupling is compensated by the DARM corrector, thus the right coupling is obtained by opening its loop.
As shown in fig.1 the initial coupling with DARM was of about 3%, after the implementation of the subtraction filter, the coupling was reduced of about a factor 10 at 20 Hz. This result is confiremed by looking at fig.s 2 and 3 where CARM_slow loop projections towards DARM, before and after the implementation of the subtraction filter, are shown.

At this point, we were in a sufficient clean condition that allowed us to perform the main activity of the shift, which was the derivation of the level of the DAC noise of the IN marionette, to verify that the estimated noise is in accordance to the known models used to produce the LF noise budgets.

In order to obtain a direct measure of the DAC noise level, we proceeded by opening the relais of the input marionette. After this action, a slight improvement of the reconstructed calibrated sensitivity has been observed (see fig.4). This improvement could be 'reproduced' if we made the hypothesis that the sensitivity curve with the high-power configuration is equal to the quadratic sum of the known DAC model and the sensitivity curve with the low-noise configuration. If we made this approximation, we obtain one similar response (fig.5). This reconstruction doesn't give a precise value of the DAC level of the marionette, but at least tells us that the known model can be assumed as an upper limit, confirming that we are not underestimating the DAC contribution to the LF noise budget.

list of useful GPS:

gps1=1449415808;dur1=300;%CLEAN; CARM subtracted; DAC ON
gps2=1449417098;dur2=300;%CLEAN; CARM subtracted; DAC OFF
gps3=1449406448;dur3=200;%CLEAN; CARM coupled
gps4=1449402128;dur4=400;%NOISE; CARM coupled
gps5=1449415008;dur5=400;%NOISE; CARM subtracted
gps6=1449418948;dur6=160;%injection on DARM

Figure 1. Following the decrease in RH power the arm power increased by about 2%, peaking around 14:00 UTC. Shortly after the interferometer unlocked and then relocked with the power in the arms decreasing. This indicates that a better mode matching by ~2% is possible, and it corresponds to flatter end mirrors.

Figure 2. Shows in purple the position of the order 2 mode in the arm FSR before the RH change, and in blue at 14:00 UTC close to the peak in the arm circulating power. The mode is at 10.9kHz-11.1kHz when the power in the arms is maximum,  this looks similar to the previous test in June of common RH change when the power in the arms was maximum for 11.15kHz,  https://logbook.virgo-gw.eu/virgo/?r=67078

Figure 3 shows the position of the order 4 mode, it moves from 21.6kHz to 22.3kHz. Note that the position of the 56MHz sideband (assuming 56436993Hz frequency) is at 22.36kHz (assuming an arm FSR of 49'969Hz based on the O4 OptChar paper), so the change in EM RoC makes the order 4 mode and one of the 56MHz sidebands overlap.

Figure 4 shows the position of the orde 5 mode. It is not clearly visible before the RH change, and it is clearly at 27.6kHz at the peak of arm power. For the other 56MHz sideband (upper or lower) the peak in the arm FSR is at 27.60kHz, exactly the same frequency.

In conclusion the change in EM RoC increases the power circulating in the arms, but it also corresponds to making the 56MHz USB and LSB co-resonant with the order 4 and order 5 modes. So just changing the EM RoC doesn't seem like a good solution to improve mode matching, and the input beam needs to be change to modify the beam radius at the level of the input mirror, which requires acting on two actuators of the input beam. However, it is much simpler, and it doesn't seem like it creates any real problems to have the sidebands overlapping with the carrrier HOM.