I measured the noise of the corrected LVDT, including the microseismic peak contribution. The aim of this characterization is not just to know the sensitivity of the corrected sensor, but to eventually use it to design the blending filters with the geophones using H-infinity methods if ever considered neccessary.

In order to measure such a noise using the IP free-swing ASD, I removed the geophone output from the corrected LVDT sensor. This subtraction removes the IP motion from the corrected LVDT because the geophones measure it too. I implemented the minus sign for the geophones at the BLEND filter banks, and measured the quantity (corrected LVDT - GEO) at the input of IDAMP block.

The two upper figures show the ASD of the corrected LVDT, geophones and the substraction (corrected LVDT - GEO) for IP-L, in logarithmic and linear scales. The substraction successfully removed the resonance peaks at 80 and 405 mHz. The lowest plot shows the coherence with the seismometer on the ground.

Although the microsesmic motion today was not small, it was not very high either. Still, in the quantity (corrected LVDT - GEO) there's relatively high coherence at 273 and 324 mHz.

Files are at:

• Directory: /kagra/Dropbox/Subsystems/VIS/vis_commissioning/sr2/inter_calibration/
• File name: SR2_corrected_sensor_noise_3.xml

It was pointed out to me that, according to the DGS, I didn't load the filter banks with the new seismometer inter-calibration factors after I saved them in SENSCORR last week. For these measurements I made sure I loaded them.