To estimate the noise floor of the Ranger, Rana and I locked the mass on the seismometer, so there will be no (aka minimal) signal from the motion of the ground in the pickup coil. We should be seeing primarily the noise of the readout electronics. We also put the Ranger on top of one of the foam lids from the Seismometer Isolation Boxes to further isolate from ground motion (this didn't change the signal noticeably).
In this plot, Green is the locked-mass-on-foam noise floor, blue is the regular spectra, with the SR560 AC coupled, and the red is the regular seismic spectra with the SR560 DC coupled. There doesn't seem to be a noticeable difference between blue and red (the spectra were taken at different times of day, with the red taken at night, when we generally expect things to be quieter). I'm leaving the SR560 DC coupled. (Rana had found it earlier this afternoon GND coupled....not sure yet why).
Also, we're not sure that the green curve is true readout noise, vs. how much of it is specifically due to the fact that the mass is locked down. Especially around a few hundred Hz, the green curve is much higher than the other 2, and at a few tens of Hz there is some weird peak action. However, this will be okay as a first-run noise estimate for the Ranger's noise floor.
The question at hand is: Do we need to redo any of the Ranger's readout electronics (i.e. replace the SR560 with a Pomona Box circuit) to lower the noise floor, or is it okay as-is?