Signal Distortion
An analytical phenomenon in digital signal processing describes the superposition of high-frequency components back into the primary baseband when the sampling rate falls below the Nyquist limit. In many high-speed analog-to-digital converters, alias folding occurs when signals beyond the half-sampling frequency fold across the Nyquist boundary and appear as spurious in-band tones.
Sampling Limit
The Nyquist criterion establishes the absolute boundary above which signal frequencies must not extend without causing distortion. Designers set the sampling rate to at least twice the highest frequency component of interest to prevent the occurrence of alias folding in the digitizer. When the input contains noise or harmonics above this rate, the sampling process folds these components back.
This folding behaves like a mirror reflection around the half-sampling frequency, where the distance of the high-frequency component past the boundary determines its new, lower frequency position in the digitizer output.
Filtering Process
Analog low-pass filters placed before the converter restrict the bandwidth of the incoming signal to prevent the return of unwanted frequencies. Active filter topologies or passive resistor-capacitor networks reject out-of-band signals before they reach the sampling stage, thereby limiting the risk of alias folding. The transition band of the filter must be narrow enough to attenuate signals before they reach the Nyquist threshold.
High-performance instruments utilize multi-pole filter stages to achieve steep attenuation curves.
Suppression Metric
A quantitative evaluation of anti-aliasing performance is measured by the ratio of the attenuated fold-back tone to the original input amplitude. System builders verify this suppression by applying a known test frequency above the Nyquist limit and measuring the amplitude of the resulting artifact. Effective attenuation prevents the unwanted folding of out-of-band noise, which would otherwise raise the noise floor of the instrument and degrade the signal-to-noise ratio.
The achieved suppression must exceed the dynamic range of the analog-to-digital converter to ensure clean signal acquisition.