Meaning
Signal degradation arises from the discrete nature of digital representation within analog to digital conversion architectures. This delta-sigma adc quantization noise occurs because the rounding process maps a continuous input range onto a finite set of output values. The discrepancy between the actual signal and its quantized equivalent generates an error spectrum that depends on the sampling rate and the order of the noise shaping filter.
Such artifacts are predominantly pushed into high frequency bands away from the baseband signal to increase the resolution of the converter.
Noise Shaping
High order feedback loops rearrange the error spectrum by applying a transfer function that attenuates low frequency components. The delta-sigma adc quantization noise floor drops significantly within the desired frequency range while rising in regions beyond the bandwidth of interest. Designers utilize this mechanism to achieve higher effective bit counts from hardware that otherwise lacks the necessary linearity.
Filter Characteristics
Digital decimation structures remove the out of band energy produced by the modulator stage. Sharp cutoff responses define the transition between the passband and the stopband where the delta-sigma adc quantization noise exists. Proper attenuation in the stopband prevents aliasing from folding high frequency errors back into the useful signal range.
Converter Performance
System performance relies on the interplay between oversampling ratios and the specific topology of the modulator circuit. Trade offs between speed and resolution often dictate the architecture chosen for different sensing applications. Higher order loops manage the delta-sigma adc quantization noise more effectively at the cost of increased complexity and potential instability.