Here is a calibrated noise plot of the Guralp seismometer box. This is the same noise measured on Friday, measured at TP3 (just after the first gain stage), with the inputs shorted.
The Guralp calibration is:
TP3 noise
noise in m/s = -------------------
10 * 802(V/(m/s))
The 10 is from the gain of 10 between the output of the seismometer and the input of the breakout box, and the 802 V/(m/s) is from the calibration data that came with the seismometer.
From elog 881 by Rana, in the ~1-50Hz band, the calibration of the Ranger seismometer is 488*10^6 counts/(m/s). Using DataViewer, I estimated that the nighttime ground motion measured by the Ranger is ~3500 counts, and the max daytime ground motion is ~8000 counts. This is what was used for the nighttime/daytime lines in this plot.
It seems like the noise of the Guralp box is fine just as it is, and we don't need to worry about replacing the first gain stage (differential instrumentation amp) with a lower-noise op-amp, since at even the lowest freqs, we have almost a factor of 100 at night, and better than that at higher freqs.
NOTE about the plot: the legend isn't showing the correct colors for the night and day motion - obviously the nighttime motion is the lower RED line, and the day is the higher GREEN line.
Yet another note: When I was measuring the counts on the Ranger, I forgot to subtract the mean, so these numbers are overestimating the ambient ground motion. The blue curve is correct however. |