power spectrum noise vs sampling rate

[mssm22]

Hi guys, So I mainly worked with a DAQ that has anti aliasing filters already (4472B). Recently, we switched to 16 bit DAQs that does not have any signal conditioning. As a first step, I wanted to get an idea of noise floor so I attached a 50bohm to the input and measured the power spectrum at different sampling rate. What I found is that the noise floor changes as a function of the sampling rate, as the sampling rate is decreased the noise floor is increased. As this point am not sure what is causing that, aliasing! maybe, I do not know. Can you please help explaining this to me. Thanks

[Stephan_D]

Hey,

I'm not shure, but could this be the explanation for the noise?

Pink noise - Wikipedia, the free encyclopedia
http://en.wikipedia.org/wiki/Pink_noise

Stephan

[johnsold]

Aliasing may be an explanation. The noise bandwidth of the DAQ device (without filters) is typically somewhat larger than the Nyquist bandwidth at the maximum sampling rate.  If you reduce the sampling rate by a factor of two, the total noise energy within the input noise bandwidth remains the same but slightly more than half of it is now subject to aliasing.  Noise is "added" in the root-mean-square sense and not linearly, so it can be tricky to calculate exact values.  This is particularly true of the noise because the noise bandwidth is generally not known exactly.

Lynn