An oversampling analog-to-digital converter topology for high-resolution signal acquisition systems

High-precision analog-to-digital converters (ADC's) which periodically sample continuous time waveforms are required in many "signal acquisition" applications such as digital audio and instrumentation. This paper describes a specific topology for an oversampling differential pulse code modulation (DPCM)-type ADC that requires neither an analog anti-aliasing filter nor a sample-and-hold. Dithering techniques are presented which cause the quantization error to resemble an additive Gaussian white-noise source whose mean and variance are independent of the input. A smoothing technique is presented which improves the linearity of the ADC's transfer function. Results presented are based on both simulations and measurements of a test system constructed from discrete components. The potential for integration in a standard digital CMOS process is discussed.