Today's activy was focussed on the suspension in two steps, according to an old procedure, chosen also in this case.

1) installation of the marionette

2) installation of a block (hosting the upper clamps of mirror wires) to be bolted unederneat 

The crew Basti, Ceccani, Chiummo, Passaquieti and Pilo get rid of some difficulties and finaly the whole payload is suspended. 

An appreciation word should be spent in favour of the vacuum cabling crew and clean room operation for their collaborative efforts. 

In the late afternoon some mass was added to the marionette, we are so lucky as the difference of loads (Oldpayload - Newpayload) was positive (~3.7 kg).

A first balance setting was done in the evening. As usual, it has to be tuned the day after (tomorrow), as usual but with some delay.

Data are there (Alain), DSP OK (Valerio). 

Before lunch I prealign the payload using the OpLev (Inshallah), then I go ahead in the alternoon, hoping not to be too late.

As part of the activities for the Auxiliary laser interlock installation, I have put in operation this afternoon the version v2r0 of the PyPlc software able to manage a configurable number of PLC registers and the corresponding channels names.
This version is able to support the 8 registers/channels coming from the NI/WI and NI/WI Auxiliary Lasers interlocks.
The new channels names are the following:

• NI_CO2_OverTempTrig
• WI_CO2_OverTempTrig
• NI_TE_CO2Laser
• WI_TE_CO2Laser
• NI_AUX_OverTempTrig
• WI_AUX_OverTempTrig
• NI_TE_AUXLaser
• WI_TE_AUXLaser

For the occasion also the process name has been changed from TCSPlc to LaserGuardianTCS.
The process name has been updated into FbsMoni and those channels are currently available on the DAQ (See Fig. 1). The DMS need to be correspondingly updated to take into account the new channels and the new naming convention.
 

In the frame of the mitigation of stray light coming from INJ benches, today we removed the magnets from the coils of SIB1 (which have been disconnetted from years).

There was a couple of magnets and a bolt inserted into a coil at each corner of the bench (see 1st and 2nd photo), one in the horizontal and one in the vertical direction. They were replaced with a screw and a bolt of similar weight.

After our work Paolo rebalanced the bench (#50042).

Below the list of the activities communicated in control room:

- Electrical work in the Central building ( External Firm managed by M. D'Andrea );

- Network System Upgrade in the Central Building and DAQ works ( Computing and Alain );

- Continuation of Work on the new IMC payload;

- SIB1 magnets removal;

- works in DET LAB for the squeezer.

After the removal of SIB1 magnets, the balance of the bench had changed a bit. Before the operation, TX_CORR and TZ_CORR were both about 1 V; after the work, 3 V were needed for TZ_CORR (fig 1).

The motorized movable mass on the marionette has been used, in order to null the corrections: 42500 CW steps in TX and 117000 CCW steps in TZ. The whole operation took more than 1 hour (fig 2). This means that the setting of the motors is wrong: 30 mV/1000steps is a resolution which makes no sense, at least 10 times more precise and slow than really needed. The operation should not require more than 5 minutes.

The other defect of the system is that currently we don't know how far we are from the end of the run. Since the installation, the motors have been used several times, because several times there were interventions on the bench which changed the balancing. A record of the steps performed till now is missing, even because the person who uses to make the operation (me) is not used to fill a notebook. An automatic storage system would be appreciable.

During the alignment works of the homodyne detector we noticed a large background noise in the audio channels, see first attachment. After investigation of the circuit board, we found that the background noise was back to reasonable values (see second attachmenbt) by replacing the two high-quantum-efficiency photodiodes with spare ones. As compared to measurments at RM1 labs (see third attachment) the current dark noise in audio band is sensibly lower. The measurements at RM1 were done before the latest 6-dB amplification stage.

The failure of old photodiodes may be due to forward polarisation during accidental inversion of the PD mounting sockets. To prevent similar events in thge future, and to improve noise couplings, soldering of the photodiodes is planned. For the moment we reduced the current limit of the HD power supply from 500 mA to 120 mA.

Today we continued to improve the coupling with the fiber:

• we try to move the three screws of the collimator, but the process seemed to be no repeatable
• we did some iteration by moving the f=500mm lens, but we did not change the matching too much (we moved the lens by about 5cm)

At the end of the process we read on the multimeter with 1kOhm 0.870V. Then we measured the coupled power at the output of all the fibers:

• SC PD1 = 1.24mW
• SC PD2 = 1.18 mW
• Main PD1  = 1.24mW
• Main PD2 = 1.18 mW

Then we remeasured the imput power: 9.3mW outside the PLL photodiode box and 9.0 mW Just before the collimator

Thus we coupled: (1.24+1.18+1.24+1.18)/9.0= 53.8% 

The coupling can be for sure improved but we think that the power is enough to start the first test in the beat note detection. At this point we realized that there was no light at the output of the fiber from INJ. 

We profit to measure the DC level with 1.2mW of coupled power also with the two photodiodes of the main PLL. We measured

• Main PLL PD1: 1.230V with 1.24mW of power, 1.126V with no light
• Main PLL PD2: 1.229V with 1.18mW of power, 1.106V with no light

The offset of 1.1V without light is not expected. The electronic is the same used for O3 scientific run with two modification:

• an additional op amp used as buffer for the DC
• the sensor is no more the FD500 by fermionics but a Thorlabs pig tailed photodiode.

We have the impression that the problem can be caused by the buffer and if this is true the RF part of the photodiode should be not affected by this problem. Thus we want to wait the possibility to detect a beat note before unmount the photodiode from the box

Next steps:

• detect a beatnote with the two thorlabs photodiode(5GUz of bandiwdth)
• move the frequency of the SQZ Main laser in order to have the beat note around 80MHz (main PLL locking frequency)
• Try to detect the 80MHz beat note also with the Main PLL photodiode. If it is detected with the expected level, try to close the Main PLL loop
• At this point we can decide if open the photodiode or if we need to couple more power

It seems that starting yesterday night either the fiber amplifier suddenly switched off or we have an artificial null signal from the related sensors. We will have a look in the laser lab.

Yesterday afternoon and this morning we completed the work, already started by Massimo, to connect the current monitor signals to the DAQ. The probes are mounted at the output of IPS and UPS of NEB and WEB. Two kinds of probes were used: LEM HTA-400-S for IPS and LEM HTR-100-SB for the IPS . The conditioning electronics is the same (photo 1 NEB UPS, photo 2 WEB IPS).

The signal names are: ENV_NEB_IPS_CURR_(R,S,T), ENV_NEB_UPS_CURR_(R,S,T), ENV_WEB_IPS_CURR_(R,S,T), ENV_WEB_UPS_CURR_(R,S,T).

The phases were identified by comparing the current signals with the voltage monitors already available. Some coherence was measured (just few tens of seconds) in a larger frequency band between voltages and currents monitors of IPS in both NEB and WEB, while it was observed only at 50 Hz and harmonics between UPS voltages and currents (fig 1). The measurements perform better with on-line data (sampled at 20 kHz).

Concerning point 1) of the entry, yesterday we checked the actual counterweights  of the motors hosted in the old marionette and they turned out roughly 1/3 lighter than those designed for the new marionette upon nominal drawings.

Then the new IMC crew (Basti et al.) installed these lighter counterweights and balanced at the best the marionette in lab 1500 West before suspending the marionette to SA of IMC. The installation is being performedd in two steps, as done in the past in this particular case. Today the bottom part of the payload. 

State of Virgo: Unlocked.
ITF Mode: Upgrading.
Quick Summary: ISYS in Safety mode; all Suspension loops open; MC tower in air, DET Tower closed with air flux On;  IB venting in progress, all SBE loops open; TCS CO2 lasers off; INJ_MAIN in pause, all the other automation nodes OFF.
Activities ongoing since this morning: Central building electrical work.

The NNC_CEB server was stopped as it introduced large latency fluctuation at the FbmSt level and triggered some missing data in the data streams .

This could be due to the fact that it was not able to reach its devices from IP netwrok and to missing timing information

We restarted from scratch with the fiber coupling

• We checked the outuput of the fiber from PSL and we realized that the power was much lower than the last time. This is caused by the Fiber amplifier misalignment
• We connected the Fiber from PSL to the input of the fibered 50:50 splitter. We connected one of the two output of the splitter to the collimator connector, then we measured the beam size exiting from the collimator. Wev realized that it was convergent.
• We acted on the three collimator lenses in order to have a collimated beam at the output of the collimator. We performed the following measurement:

• Starting from the PLL beam parameters at the output of the squeezer we designed a telescope in order to have a beam waist of 500um at the level of the collimator: i.e. 75cm after the output of the squeezer. We placed two lenses
  • f=1000mm @ 7.2 cm after the output of the squeezer box
  • f=500mm  @23.2 cm afrer the output of the squeezer box
• With the beam profiler we cheched the beam size just before the collimator input and we measured about 500um
• We started to couple the fiber and at the end of the process we measured 0.821V with a load of 1kOhm. We disconnected the optical fiber from one of the 4 photodiode and we measure 1.16mW of output power. We measured the IR power just before the collimaor and it was  8.9 mW. Thus we coupled (1.16*4/8.9)*100 = 52% of the input power. This coupling was reached putting the telescope on the table and without performing any iteration with the lenses. We did only one iteration by moving the collimator lens. Moreover we realized that the last time we forgot to tune also tha angle of the half-waveplate before the collimator.

Comments:

1.  SQZ Main laser pick off beam power before the PLL box is 9.2mW. If we measure it after the half-waveplate we obtain 8.9mW
2. The last time when we obtained 27% of coupling we measured with the tester 3.6V DC with 1kOhm of Load. This time we obtained 52% of coupling and we measure 0.812V. To be investigated
3. With the frame for the plexiglass cover of the bench is impossible to close and open the PLL photodiode box once it is aligned. We need to misalign it each time we want open or to modify the cover of the box. Today we leave the box open because we need to complete the debugging of the Main PLL photodiodes
4. Before installing the new telescope we put in the right place the two pillars of the plexiglass cover frame. We took in account their position in the design of the telescope (fig 2 and 3). Moreover we had to move by more than 10 cm the PLL photodiode box along the west direction. Now the remaining space just enough to connect and disconnect the LEMO6 connectors for the Photodiode powering.

This afternoon we found the DAQ in a strange state :

• several decrease of data flow without any reasons
• strange latency fluctuations

Loocking at the monitoring of the IRIGB signal at each building(zoom) one can see that:

• the WE signal disappears during the nigth: this was due to a failure on rtpc9-lk2, moving the optical transceiver on the rtpc9-lk1 allows to recover part of the data)
• the NE signal disappears around 9h00 UTC , with the channel available in the DAQ
• the MC signal disappears at the same period as the NE one
• Everithing remains OK in the CEB

As there was cabling activities in the DAQ for the network upgrade, we went in the DAQ room and we found that  the optical patch panel used to distribute the Timing signals at NE,WE,COB, and MC and for the Tolm network for the Suspenson in the same areas was disconnected ,

As consequence :

• the {Sa,Sc} datas for  the MC, NE and WE suspension were absent all the  day
• all the MC,NE,WE,COB Tolm devices have to be resynchronized

To recover  the correct running conditions, the missing fibers were reconnected and the INJ, NEB and WEB Tolm devices resynchronized

A the WEB the VME crate hosting the ADC7674 and the MxDx_v1 was found unpowered . Plugin back the cable allows to restart the VME crate

We added the six 10 cm missing cables (see figure):

• one for each thorlabs photodiode to the input of the corresponding bias tee
• one from the RF output of each bais tee to the input of the corresponding RF amplifier
• one from the output of each RF amplifier to the corresponding connector on the external panel of the box

The Shift was dedicated to the following activities:

- Electrical work in the Central building ( External Firm managed by M. D'Andrea );
- Network System Upgrade in the Central Building ( Computing and DAQ Team );
- Continuation of Work on the new IMC payload;
- transportation of IM microtower from 1500N ( see logbook entry #50029 );
- hoisting system installation at END clean Room ( see #50027 );
- Fiber Amplifier misalignment investigation ( see #50025 ).

Venting is in progress

This morning while going in laser lab we realized that the PMC was misaligned.
While looking at the signals (figure attached), it seems that the increase of temperature of the last days (and/or the lack of data of the 22/11) induced a change in the behavior (power, pointing) of the fiber amplifier.
The temperature came back to a nominal value but the beam did not.

Moreover, when investigating the problem we spotted some dust on the last lens of the telescope (picture attached). 

As soon as we will get the signals back, we will futher investigate the problem and remove the dust. 

video link