At 20:59 Andrea put the automation in SQZ_INJECTED with the angle that maximized the range 2.7 rad.

After 10 min I relized that the B1 4MHz mag was only 1.4mV instead 1.8mV.

I decided to check the alignment of the system acting on SQB1_LC_X now at -3370 um and before at -3150um, then acting on the SQB1_RetroRefl H and V.

After some iterations I reached a B1_4MHz_mag of 1.92mV.

Then I decide to took 10 minutes of shot noise at 22:01 UTC.

At 22:11 I run again a CC phase scan (22:13 start of the scan)

At 22:48 I restarted to inject SQZ into the ITF. 

Note that:

in fig 2 you can see that during the phase scan the range was lower than before but the  b1 4 MHz mag had a maximum at 1.9mV

In fig 3 a comparison between hrec (no SQZ) during the on off and in the last part of the shift

We left the system in SQZ injected with the shutter open

Figure 1 the bump in the spectrum at ~35Hz started around 21:30 UTC. It is relatively broad, and is most likely scattered light from the squeezing, involving ground motion which makes a 35Hz mechanical resonance of the scattering surface modulated by the ground motion look like a broad peak instead of a narrow resonance.

To investigate the origin of the excess noise at low frequency highlighted in Michal's entry, I estimated the BLRMS of Hrec in the band 30 and 40 Hz and correlated it with all trend *_rms channels (~9k). The BLRMS is estimated with the modified algorithm described in VIR-0484A-21. This estimate automatically ignores spectral lines, which would dominate the RMS value in the band, and glitches, which would influence the Pearson correlation.

Figure 1: The most correlated channels are those in the V1:SQB1_LC_LVDT_{BL,FL,BR,FR}_H_out_raw_DS_rms family.

I then repeated the analysis to include all the ENV_* and SQZ_* channels (~2k). In the attached figures I report those that look most interesting.

Figure 2: V1:SQZ_AA_SLOW_Cam{0,1}Y_ERR

Figure 3: V1:ENV_DT_CT_FINGER_ACC_Y_min

Figure 4: V1:SQZ_MIRROR2_X_cmd_1kHz_*

I add some details. I used as figure of merit to improve the alignment between SQZ and ITF the B1 4 MHz mag, i.e. the beat note between the CC field and the ITF beam.

As you can see I increased it between 21 and 21:30 UTC acting both

• on the SQB1 Retroreflector X and Y 
• on SQB1 LC X

You can see from fig 1 that the 35 Hz bump started after the movement of SQB1 LC X. I reverted its position on Monday morning

Please note that moving the retroreflector can also move the position of the beam on SQB1 AA mirror increasing the scattered light (to be checked)

In addition on Monday I checked the mag of the 4 MHz on the EQB1 HD Detector and I found a lot of losses. I had to move back the retroreflector to decrease this amount of losses. I don't understand why the mag into the ITF increased.

Figure 1. There are angular dither lines in the squeezing beam path, one of them is at 35Hz. So the bump at 35Hz could be related to that particular angular dithering.