Deploying several 9253's to take advantage of the integral LPFs and remove a high frequency component of my signals.
Required passband is 0-55Hz. Plan to use the 125Hz configuration
Significant content at 994Hz to be attenuated as much as possible. Plan to use 4 Pole
Response Time/Phase Shift is not critical. Will use the Butterworth filter.
Data Rate: Plan to use the internal timebase of 12.8MHz and a decimation rate of 200 to achieve a data rate of 500Hz for each channel.
Question: In the DataSheet at the bottom of p22, there is mention of aliasing that could occur for frequencies higher then 1/2 of the sampling frequency. I assume that this pertains to the 12.8MHz timebase, so that I will not have aliasing as long as all signal frequencies are less then 6.4MHz. I just want to make sure this is correct and that the 500Hz data rate I'm striving for is irrelevant relative to aliasing.
And yes data rate will not affect the Aliasing.
In order to determine the Nyquist frequency which is the reference frequency to avoid aliasing:
You may refer the first page of the NI 9253 Datasheet: https://www.ni.com/pdf/manuals/378060a_02.pdf
which shown the sampling frequency, fs (50k per sample per second per channel).
To avoid Aliasing, simply follow the formula, fN=fs/2 to obtain the Nyquist Frequency needed, any frequency below Nyquist frequency, will not alias.
For more information about Alias Calculation, kindly refer to this link:
The NI 9253 implements multiple filter stages to provide an accurate representation of in band and out of band signals. As a part of this multi stage design we include analog anti aliasing filters to attenuate the higher frequency content that could alias into the passband. This is different than the aliasing discussed on pg. 22. The aliasing discussed in the comment on pg. 22 is in reference to signals in the transition band of the butterworth filter itself and is a more general comment about butterworth filters.