With Yokozawa-san and Washimi-san.

In order to assess the origin of the features in residual motion measurement reported in entries 10209, 10210 and 10216,  on Monday 2nd of September we placed three accelerometers around the oplev. Namely, one at the oplev lower platform, another one at the upper platform and a third one close to the second one but on the large frame. We measured amplitude spectral densities with all the filters on and with some filters off (all in the SR area and half in the BS area).

The channel names and files used are the following:

Directory:

/kagra/Dropbox/Subsystems/VIS/TypeBData/SR3/Noise/Measurements/20190902/

Files:

• SR3_ALIGNED_ASD_FFU-OFF_20190902.xml
• SR3_ALIGNED_ASD_FFU-ON_20190902.xml
• acc_oplev_upper_platform_FFU_ON.txt
• acc_oplev_upper_platform_FFU_OFF.txt
• index.txt

Channel names:

• Lower platform: CH1 ACC : K1:PEM-SENSOR_RACK_SR3_BNC8_OUT
• Upper platform: CH2 ACC : K1:PEM-SEOSOR_RACK_SR3_BNC9_OUT
• On the large frame by the upper platform: K1:PEM-SEOSOR_RACK_SR3_BNC10_OUT

The following observations apply:

• Below 7 Hz the redout of the microphone is larger with the fans off. Nevertheless, at frequencies higher than 60 Hz is smaller. Clearly there's something unusual and it shoud be checked.
• In between 7 Hz and 60 Hz  the reduction in the microphone readout floor is small when turning the filters off. Some peaks are clearly smaller.
• In the readout of the upper platform accelerometer and in the oplev for TM-L, TM-P and TM-Y some peaks are clearly smaller with the filters off, but the noise floor remains at the same level.
• The coherence of the accelerometer at the upper platform and the oplev readout is shown. Above 10 Hz we have regions of high coherence.
• The coherence between the same accelerometer and the IM-L, IM-P and IM-Y is shown in order to identify oplev injected noise. The coherence is relatively high at some frequencies, especially in IM-P from 16 to 18 Hz.

A strategy would be to filter the oplev signal with a low pass filter with a cuf-off frequency of about 9 Hz. In the files listed above there's still data to analyze.