A Pattern Extraction Based Background Calibration Technique for SAR ADCs

Abstract

This brief presents a pattern extraction based digital background calibration for Successive Approximation Register (SAR) Analog-to-Digital Converters (ADCs). We demonstrate that major factors limiting the convergence rate of the conventional calibration schemes can be addressed by performing pattern extraction and using the extracted patterns to train the calibration engine. Benefiting from the proposed technique, the convergence speed of the calibration can be improved without sacrificing the stability and accuracy of the calibration loop. A lightweight pattern extraction scheme is also introduced to reduce the hardware cost and power consumption.