A brief check:

• Concerning the mechanical resonances, the SR suspensions have the following modes at around the frequencies (from 5 to 50Hz), according to the model prediction (ref):
  • at 5 Hz, RM Pitch,
  • at 11 Hz, RM Vertical,
  • at 16 Hz, TM Roll
  • at 50Hz, RM Vertical.
• About 17Hz, it could be the RM roll mode coupling.
• about 30 to 35Hz band, it would not be due to a pure mechanical resonance since the suspension does not have resonances at the frequency band.

Thank you Fujii-san for the quick check.

Except for the 17Hz noise, they are not so narrow band and coherence itself are small.

So, not the mechanical vibration but a other light source injecting to oplev (or very small fluctuation of SRM mirror)

In case you don't see the pictures attached to this report please click this link: 10216.

I measured the amplitude spectral densities (ASD) of the SR3 optic displacement in ALIGNED state and the coherence with the microphone in SR clean booth. See the pictures.

• In the ASD of the motion of the optic we can see lots of features from 8.4 Hz up to 470 Hz. Likely these features are not mechanical resonances of the suspension but mechanical resonances of the oplev hardware excited by acoustic noise.
• The coherence with the microphone also shows many peaks within that frequency range, including the vicinity of 17 Hz.

How to proceed

1. Check the residual motion in other Type B suspensions.
2. We can either
   • turn the clean booth fans off and check whether the peaks are gone or
   • use accelerometers on the oplev platforms and measure their coherence with oplev readout.
   • both of the above at the same time.

It's worth pointing out the following

1. Placing accelerometers on oplev platforms is a good strategy. However, it would require a lot of care because of the Al wrap around the hardware and hard access. Likely, the oplev would require diagonalization again after the change.
2. Measure the coherence of the SRCL signal with other optics.
3. In case we see the same coherence with microphones in other optics a faster strategy would be to turn the filter fans off.

About the files:

Directory:

/kagra/Dropbox/Subsystems/VIS/TypeBData/SR3/Noise/Measurements/20190830/

File name

SR3_ALIGNED_ASD_20190830.xml

We also want to check this 35-37Hz vibration.

There is no coherence with SR microphone from your figure.