On Self-Aware Mixed-Signal Systems Based on S-Δ ADC

Abstract

In this paper the authors discuss the issues related to the self-awareness of high-resolution, mixed-signal circuits and systems, based on S-? ADC, which is the most important and sensitive module and the key element for analogue to digital conversion. The basic methodology and framework for improving the self-awareness of such systems are presented. The methodology is based on efficient real-time measurements of a high-resolution, mixed-signal system using pseudo random signal source, real-time calculation of a distance between responses, the possibility to adapt measured circuit to minimize the distance, and changing the parameters of a reference system according to learning rules. The use of pseudo-random noise as a signal source leads to efficient and cost-effective measurements that run in parallel to the main signal processing. The calculation of the distance between the system and its reference are theoretically analysed and verified using Matlab model. The response of a system together with the response of high precision analogue to digital converter (ADC) is compared to the response of a bit-true model of a reference digital circuit. The differences are calculated using simple area-efficient cross-correlation algorithm. Together with adaptation strategy and tuning circuitry it forms the basis for self-awareness of mixed-signal circuits.