Alias Rejection at Low Sample Rates

At very low sample rates, the anti-aliasing filters for AI channels on DSA devices as well as the NI 9225, NI 9229, NI 9233, NI 9234, NI 9235, NI 9236, NI 9237, and NI 9239 may not completely reject all out-of-band signals. This is primarily due to the internal digital filter of the delta-sigma ADC, which cannot suppress signals with frequencies near the multiples of the oversample rate (sampling rate multiplied by oversample factor). These devices also employ fixed cutoff analog lowpass anti-aliasing filters, but at low sample rates, some multiples of the oversample rate can fall below the cutoff frequency of the analog anti-aliasing filter.

For example, for a device using ADCs with an oversample factor of 128X and sampling at a rate of 1 kS/s, the oversample rate is 128 kHz. Some multiples of that oversample rate fall below the cutoff of the analog anti-aliasing filter. If the signal to be digitized contains energy near these frequencies, aliasing can result.

One way to prevent aliasing is to raise the sample rate so that the first 128X multiple of the sample rate falls above the cutoff of the analog anti-aliasing filter. For example, a sample rate of 25.6 kS/s is not subject to aliasing because the first 128X multiple (3.2 MHz) is well above the cutoff frequency of the analog anti-aliasing filter. Some DSA devices support enhanced alias rejection, which automatically handles alias rejection at low sample rates. Refer to the device documentation for the specifics of your device.