While looking at the Ranger seismometer's output to figure out what our max typical ground motion is, Rana and I saw that the DAQ output is at a weird level. It looks like even though the input to the DAQ channel is being saturated, the channel isn't outputing as many counts as expected to Dataviewer.
Sharon and I checked that the output of the seismometer looks reasonable - sinusoidal when I tap on the seismometer, and the the output of the SR560 (preamp) is also fine, and not clipping. If I stomp on the floor, the output of the SR560 goes above 2V (to about 3V ish), so we should be saturating the DAQ, and getting the max number of counts out. However, as you can see in the first figure, taken when I was tapping the seismometer, the number of counts at saturation is well beneath 32768counts. (16 bit machine, so the +-2V of the DAQ should have a total range of 65536. +2V should correspond to +32768counts.) The second figure shows 40 days of seismometer data. It looks like we saturate the DAQ regularly.
I did a check of the DAQ using an HP6236B power supply. I sent in 1V, 2V and 2.2V (measuring the output of the power supply with a 'scope), and measured the number of counts output on the DAQ.
|
Input Voltage [V] | Counts on Dataviewer | Expected counts from 16 bit machine
| |
1 | 18983 | 16384
| |
2 | 29331 | 32768
| |
2.2 | 29347 | 32768
|
I'm not sure why the +1V output more than the expected number of counts (unless I mis-measured the output from the power supply).
Moral of the story is...when the DAQ is saturated, it is not outputting the expected number of counts. To be explored further tomorrow... |