The channels

• DAQ_LNFS_56MHz_raw_mag shows the amplitide of the 56MHz line at the level of the LNFS
• DAQ_EOM_56MHz_raw_mah shows the amplitude of the 56MHz line at the level of the EOM readout

The fluctuations at 112MHz of the  photodiode signals seems to be directly correlated to a variation of the 56MHz modulation at the EOM level.

ITF in Science Mode for the whole shift with BNS Range 57 Mpc.

Guard tours (UTC):
23:29 - 00:03
01:35 - 02:07
03:34 - 04:04
05:03 - 05:34

ITF in Science mode for all the shift, summary plot attached.

Guard tours (UTC):
18:55 - 19:29
20:41 - 21:11

The behavior of the sidebands during the last lock is quite strange: sometimes they drop to a very low level and stay down showing a certain instability; sometime they are back to a normal and stable level. This trend is common to all the PDs demodulated at 112 MHz (fig 1), and is seen also by B1p, which contains the sidebands for a certain amount (fig 2). There is nothing similar in the carrier, for example B4 (fig 3). At a first look, I cannot find anything in the alignment signals which could explain this behavior.

ITF locked for all the shift; Science mode was interrupted:
- at 6:27 UTC because of Squeezer pump phase reset;
- at 13:53 UTC because of Squeezer unlock;

ITF left in Science mode.

Air Conditioning
From 11:47 UTC the BACnetServer has been losing connection with the modules ACS_INJ and ACS_DET; to be monitored.

EnvMoni
Red flag for:
- ACS_INJ - ENV_EIB_ACC_Y
- ACS DET - ENV_EDB_MIC
Irene has been informed by mail.

Figure 1 compares data without the beam present on the B1 photodiode (purple), and data in science mode (blue). For 2mW of power on each photodiode, the quantum noise + electronic noise is a actor 2 above electronic noise alone. So improving the B1 RF electronic noise to remove it completely, would only reduce the noise by a factor sqrt(2^2-1)/2 = 0.87. Only a 13% improvement in the squeezing ellipse error signal SNR.

ITF found in Science Mode, it kept the lock for the whole shift.

Guard Tour
23:30 UTC - 1:00 UTC;
1:34 UTC - 2:04 UTC;
3:32 UTC - 4:02 UTC;
4:57 UTC - 5:29 UTC.

ITF in Science Mode for the whole shift with BNS Range 56.5 Mpc.

GRB
14:40 UTC - Fermi event created: E364085

Guard tours (UTC):
14:00 - 14:30
16:01 - 16:29
18:00 - 18:33
19:56 - 20:22

ISC
diffp_tx_set_ln3 = 0.010
diffp_ty_set_ln3 = -0.037
Values saved in ITF_LOCK ini file.

DAQ
(15-02-2020 14:00 - 15-02-2020 15:30) From remote
Status: Ended
Description: Rtpc14 unreachable - Expert rebooted it from remote.

ITF in Science mode for all the shift, summary plot attached.

Guard tours (UTC):
06:12 - 06:39
07:57 - 08:32
12:05 - 12:30

VPM / DataCollection - 13:30 UTC - DMS showed a red flag, I tried to restart unsuccessfully 'TproImgPAY', 'Pay_Img' and 'DET_Img_End', rtpc14 seems to be crashed, expert informed by phone.

Adjusting was to tune demod phases for SQZ AA

ITF found in Science Mode, it kept the lock for the whole shift.

Guard Tour
22:59 UTC - 23:38 UTC;
1:06 UTC - 1:42 UTC;
3:01 UTC - 3:35 UTC;
5:01 UTC - 5:41 UTC.

After the SQZ MM tuning action of Wednesday, there was some indication of extra noise from scattered light on external SQZ bench ESQB. So yesterday afternoon I tweaked the alignment upstreams the triple Faraday Isolator.

In order to identify noise from ESQB I took two reference spectra with the slow shutter (on ESQB) closed, see timing on Fig1 and a comparison of Hrec spectra in Fig2 with shutter open (magenta) and closed (blue).

To check how the noise from scattered light can be reduced with proper alignment on ESQB, I stopped the SQZ AA loop, then I sligtly mistuned the SDB1 Faraday Isolator waveplate to send more light on ESQB, and I increased the frequency and amplitude of a dither line on one SQZ AA mirror. Some noise is visible in Hrec around the dither frequency (85 Hz) and between 45 and 50 Hz, see Fig3. By slightly tilting the SQZ AA mirror closer to the ITF (Mirror1) the noise can be largely reduced, see Fig4.

Then tuned back the SDB1 FI waveplate, then I moved all SQZ AA dither lines to higher frquencies and amplitude, until noise on Hrec is clearly visible at the dither frequencies; then I modified the AA Matrix in SQZ_PLL_AA server to increase the loop bandwidth. I entered the DET lab and moved the tilt of output mirror of the MM telescope on ESQB to reduce such noise, see Fig6. Unfortunately I had to stop before reaching the optimal alignment, due to lack of internet connection.

After restoring the usual conditions for the squeezer AA, I chiecked that the noise in Hrec around 50 Hz is similar to the case with slow shutter closed, see Fig 7 where blue spectrum is with shutter open before alignment, yellow shutter closed, magenta shutter open after alignment.

14:00 UTC - ITF found In Science Mode with BNS Range 56 Mpc.
Adjusting Mode break at 15:15 UTC for an SQZ check.

At 20:08 UTC SQZ disengaged. When I tried to re-engage it, ITF unlocked.
ITF promptly relocked at first attempt in LN3_SQZ at 20:33 UTC. Science Mode set, BNS Range ~ 57 Mpc 

Guard Tours (UTC):
19:00 - 19:30
21:05 - 21:33

Here the GPS of hte injections performed yesterday to complete the IB and MC control noise budget:

• 1 to 4: the noise, shaped like the blue line in the 1st figure, has been applied to each IB coil. The TF has been computed between this noise (called Sc IB noise) and Hrec, and then projected on the board noise model (red curve), thus without need of any multiplication factor to take into account the number of coils. The 2 plot shows the TF and coherences between the noise and Hrec, while the 3rd plot show the projection of each coil (L,R,F,B) on the red curve of fig.1.
• 5: the same kind of injection as the ones reported in #48362 but with more noise applied on the MAR TY dof (see plot 4). The TF and coherence with Hrec are shown in fig. 5, its projection on fig. 6. The point at 18,58 Hz is not meaningful since we weren't injection noise in this point (plot 7).
• 6 to 8: noise applied on the MC MIR coils (Sc_MC_noise channel). To project the DAC noise the TF between Sc MC noise and Hrec have been computed (plot 8) and the projection on the board noise has been made taking into account that there are 2 coils (thus dividing the board noise by sqrt(2)). It is shown on fig. 9
• 8 and 9: noise applied on the MC marionetta angular corrections TX and TY (same kind of measurement as 5, but the number of coils is 2 instead of 4). The applied noise is shown in fig. 10 (blued curve) the marionetta corrections TFs and coherences with Hrec are shown in fig. 11 and their projection on fig. 12. In the projection we removed the points around 50 Hz and its harmonics for which no enough noise could be injected, as shown in fig. 13.

I've slightly refined the Etalon wp (NI from 19.26 to 19.3 and WI from 19.54 to 19.53 corresponding to the 61Mpc).

the behavior of the last few days is visible in the next plot:

1) Trend of the data set

2) Cleaned data set (cleaned from Range drop etc...)

3)  OG PR, OG cav, Range vs Input temperature. (good point is max R, min OG pr and max OG cav which is visible also in Figure 4)

4) Range vs OG PR and OG cav which shows the behavior: high range for low OG pr and high OG cav.

5) 3d plots to pick up the optimal point

ITF locked for all the shift, Science mode stopped only for two minutes in order to tune the Cosmic muon detector. Summary plot attached.

form 8:35 to 8:37 UTC - ADJUSTING MODE - Cosmic muon detector turned.

ISC
Tuning of DIFFp_AA performed twice:

07:35 UTC - TX = 0.035
09:10 UTC - TX = 0.015

The attached animated gif compares 5 data segments, each 10 minute-long:

• In red (common to all pictures), before turning on the cosmic muon detector (09:20 LT -> 09:30 LT)
• In blue: 4 segments from after the switch on, separated by one hour each (09:45 LT, .., 12:45 LT)

Some peaks fluctuate in amplitude and/or frequency but no systematic change is visible.

[The improvement at low frequency is due to the decreasing sea activity.]

ITF Mode set to adjusting the time to go to the "locale camion" and back. Then back to science mode. The device responds to a |ping| command on the network. Data acquisition will be started remotely from IP2I Lyon experts.

This night the ITF unlocked from science twice because of the wind, the weather condition get worse over the night and prevented to have a stable lock, I had also problem with the squeezer injection that get stuck with the SQZ metatron node in ENGAGING_SLOW_AA state, so I called Jean-Pierre (SQZ on-call) to check the situation;
finally I succeeded to relock so I decided to set the "earthquake configuration" and wait for an improving of the weather conditions and in the meanwhile let Jean-Pierre check the squeezer system.
Squeezer restored, restored the standard actuator configuration and science mode started at 3:53UTC.

Science mode stopped from 0:02UTC to 0:04UTC due to squeezer pump phase reset.

-guard tours(UTC):
23:05 --> 23:45
1:20 --> 1:50
3:20 --> 3:55
5:00 --> 5:30

Air Conditioning
The technical building cold water pressure is sligthly under it's threshold (2nd plot), under monitoring...

ISC
3:50UTC - DIFFp tuning:
TY -0.040
TX 0.050
values saved in ITF_LOCK.ini

SQZ
(14-02-2020 03:20 - 14-02-2020 03:45) From remote
Status: Ended
Description: squeezer injection not working, and automation blocked.
Actions undertaken: restore of the sqz.

ITF found in Commissioning Mode, SQZ Activity still in progress. From 15:35 UTC to 17:30 UTC F. Sorrentino worked inside the DET Lab to complete the alignment of the Squeezer, after that this activity was completed the ITF was set back in Science Mode at 17:37 UTC.
During the following 30 minutes the Squeezer disengaged 3 times ( 17:45 UTC, 17:53 UTC, 18:04 UTC ) due to the AA Loop. I set the ITF back on LOW_NOISE_3 and contacted the expert on the matter. From remote M. Vardaro solved the issue, switched to Adjusting mode from 18:28 UTC to 18:30 UTC during the fix. ITF in Science Mode from 18:31 UTC.

Guard Tour
18:45 UTC - 19:27 UTC;
20:37 UTC - 21:12 UTC;

Air Conditioning
From 15:35 UTC to 17:15 the UTA DET was set in "Portata Nominale" to allow F. Sorrentino to work in the DET Lab.

DAQ
- SQZMoni restarted at 18:47 UTC under request of M. Vardaro to update SQZ_PumpPhase_HV_Monitor_50Hz thresholds.
- InjMoni restarted at 18:53 UTC under request of A. Chiummo to update mean_PSL_SL__XTAL1_W_60 and mean_PSL_SL__XTAL2_W_60 thresholds.

ISC
From 21:22 UTC to 21:45 UTC, due to the red flags of DIFFp_AA, I tuned ASC_DIFFp_TX_SET. Value changed from -0.005 to 0.030 and saved in the config.

The attached file lists several glitches involving the excitation of one drum mode. The ranking is:

NE: 34

WI: 15

NI: 6

The identification of the TM responsible of the glitch is not so easy when the TM is one ETM, because the modes are almost overlapping, but finally I'm convinced that they are all related to NE ( a cross check is recommended).

I put also two plots showing the loudes events for each TM. Events like the NI one, with the notch in the spectrum, are very rare, but are there: I found two of them related to NE.

There is a peculiarity about the NE drum excitation: the decay time looks always quite short. it would be worth to check the associated thermal noise.

We measured the modulation depth of the three sideband frequency both from the arm free swinging and from a RFC cavity scan. Notice that the modulation depth change during the lock acquisition process (in particular in LN1). 

Hereafter the value that we measured when the ITF was unlocked. After the arrow, the values in LN1. These latter have been derived by taking into account the variation as read from the channels INJ_LNFS_AMPL1,2,3. 

- m_6MHz = 0.29 -->  0.002

- m_8MHz = 0.18 --> 0.1

- m_56MHz = 0.11 --> 0.195

06:00 UTC - ITF in Science Mode with BNS Range 56 Mpc.
1st ISC Commissioning break carried out by Ruggi (09:20 to 10:05 UTC).
2nd ISC Commissioning break carried out by Ruggi and Allocca (10:21 - 12:25 UTC). The ITF unlocked at 10:36 UTC. The relock was a bit difficult due to the earthquake occurred in Russia with Mag 7.0 (fig1)
3rd Commissioning break (ISC from 12:49 to 13:51 UTC, SQZ carried out by Sorrentino from 13:52 - still in progress).

GRB
09:00 UTC - Comment: [gracedb] Fermi event created: E363867

DAQ
06:00 UTC - FmTrend, FbTrend crashed / restarted

ISC
Tuning of DIFFp_AA. New value (TX = -0.005, TY = -0.037) saved in the ini file.

For O4 the OMC lock accuracy needs to be improved by a factor 10. This can be done by reducing the CARM motion by locking it on the RFC, or by increasing the gain of the OMC length lock loop. Both actions might be needed at the end, here I explore how much can be expected from working just on the OMC length loop. I will look here only at the precision, accuracy depends on whether there is an offset or not in the error signal, but that is a different topic.

Figure 1, shows in blue that resonant gain could be added below 1Hz. The OMC length lock UGF is around 2Hz. The resonant gains would reduce the phase margin by 20 degrees. This could be recovered by making a less sharp roll-off at high frequency.

Figure 2. shows the present PZT feedback, which correspond to OMC length noise injection as the PZT to length response is flat up to a few hundred Hz at least. Reducing the roll-off (red curve) would increase the feedback noise above 10Hz. But actually it would still be below the thermorefractive noise at most frequencies, only between 10Hz and 12Hz it would increase the length noise by a factor 2. Which is acceptable as the noise projection there is 3 orders of magnitude below the present sensitivity in h(t).

Figure 3 shows the impact of the resonant gains on the OMC error signal, in blue the present case and in red with the impact of resonant gains. The lock error signal RMS is reduced by a factor 7, not exactly the required factor 10, but almost there.

To get a more precise picture a measurement of the OMC length loop transfer function would be needed. It hasn't been done since spring 2018, the present UGF is not known, and the cross over between Peltier and PZT feedback at 250mHz.makes the modelling not trivial.

/users/mwas/OMC/OMCloopModel_20200213/OMCloopModel.m

The shift started withe the ITF in science mode; adjusting/troubleshooting from 22:12UTC to 00:29UTC to allow the intervention of Antonino to solve the problem of slave laser crystal temperature increasing (#48440), the ITF unlocked during these operations, it relocked without any major problem.


-guard tours(UTC):
1:20 --> 1:55
3:05 --> 3:35
5:00 --> 5:30

ISC
2:50UTC, 5:20UTC - DIFFp tuning:
TY -0.045
TX 0.025
values saved in ITF_LOCK.ini