RM-26999 Signalconditioner; impact of noise on total accuracy

[LVFan]

I want to use the device to measure 130Vrms by using sampling the downscaled output with a PCIe-6346.

The specification states that in +-200V range the system noise is 21mV.

Does this mean this noise is present on the output of the conditioner?

130V supplied to input would result in an output of 0,65V. And 21mV noise on top of that output would mean an possible deviation of 3.3% which I find quite big.  

Is this correct?

So, I have to filter that noise in some way via software or lowpass?

[santo_13]

It is not clear from the specifications as well.

Based on (4) I would assume the 21mV RMS noise is added to the input signal and may not be at the output. If you have the hardware, it would be best to measure the noise and offset yourself.

Here, noise refers to a broad spectrum random noise and averaging would be only way to reduce random noise.

[LVFan]

Yes this annotation 4) confused me as I would have expected that noise is specified for the outputsignal.

So the visible noise at +-200V on output should be 0,105mV?

I dont have the device and want to make sure it meets requirements before I buy it.

[santo_13]

You probably need to talk with Tech Support at NI, so they can check with R&D on how to interpret the specs.