The cross-correlation analysis with NonNA of some bands of interest ([118,122],[148,152],[152,153],[153,155],[155,160],[207,212]) in the previous plot has shown, again, correlation with INJ_IMC and PMC: report here.

This can just be a consequence of the residual presence of glitches from these two subsystems. Also some environmental sensors (ENV_EDB_SEIS*, ENV_NI_ACC, ENV_WE_ACC) resulted correlated; this may be related with the worsening of the weather conditions.

The median normalised spectrograms of the region between 150 and 200 Hz, with the new noise structures, and that at low frequency, with the highlighted the effect of the weather conditions, seem to excluede a correlation between the former structures and the environmental cause.

At the weekly meeting it was mentioned that these bumps are coherent with the SDB1 local controls, but that the structures are not present in the local cotrols signals.

Figure 1 shows that there is coherence with the SDB1_LC_TZ_corr channel. This correction is a factor ~50 larger than for the other two angular degree of freedoms because it uses the LVDT as an error signal instead of the quadrants on SDB1. In principle TZ has not much impact on the B1 beam as this is the rotation around the beam. However it could have a large impact on the squeezing as a TZ rotation of SDB1 causes the beam to move up and down on the squeezing bench.

Hence it would be important to investigate if this coupling could be due to the beam being slightly clipped on the squeezing bench, for example by closing the shutter either on the squeezing bench or on SDB1. So the bumps would be coming from the motion of the in air optics on the squeezing bench, which couples the beam jitter coming from the SDB1 bench.

The 155-170Hz peaks seem to have a companion at 197Hz, that undergoes a similar amplitude modulation: see spectrogram in figure 1.

From Bruco this peak has top coherence with B5 quadrants, in particular ASC_B5_QD1_V which also has a sort of peak at the same frequency: see figure 2.

The coherence is however small, reminding scattered light.

Logbook entry 46478 reports that this 197Hz peak, as well as the 150-170Hz reduced consequent to a Tuning BS, PR, Diffp setpoint.

Indeed if we compare the sensitivity after the AA tuning of today purple curve (see log entry 47635 ) with respect to the curve of this morning (blue curve before the tuning) it is visible that these structures are lowered, see Figure 1.

I have compared a period with the Squeezer slow shutter closed and one with the Squeezer slow shutter open (sqz injection). The bump seem no caused by scattered light from the squeezer in air bench.

Fig 1:  zoom around 150 Hz. In blue  slow shutter is closed, in red slow shutter is open. 

Fig 2:  sensitivity spectrum with the shutter open (red) and closed (blue). Maybe there was a sligthly worsening between 10 and 20 Hz, but is not so evident. To be re-checked!!

A broad notch could be added to the SDB1 TZ feedback signal (see figure 1), which would reduce the excitation of this motion on SDB1. The DAC noise (at 3e-7 V/rtHz) is a factor 7 below the correction signal (at 2e-6 V/rtHz) at the moment, and the notch has a reduction of at least a factor 8 in that band, so the excitation should be reduced by at least a factor 5 by adding the contribution of the correction signal and the DAC noise. From memory the SDB1 TZ loop UGF is around 1Hz, so loosing 0.5 degree of phase margin shouldn't be a problem.

From broadband injection into the bench local controls, these peaks were excited by both TX and TZ noise injections with a similar amplitude. Which points rather to something on the SDB1 bench. See figure 2 for the TZ injection case.