HIGH-LINEAR ANALOGUE-CODE-ANALOGUE SYSTEM WITH WEIGHT REDUNDANCY BASED ON CURRENT CONVERTERS

Abstract

At present, analog-to-digital and digital-to-analog converters are widely used for operations with signals and allow recording and storing signals, outputting a signal in analog form, performing operations with signals, and generating arbitrary waveforms. In a number of applications of analog-to-digital and digital-to-analog converters for signal processing, high resolution, low noise and high linearity of the analog-to-digital-to-analog transfer characteristic are required to provide low non-linear, frequency, phase distortion of signals and a high signal-to-noise ratio. A method for constructing a high-linear analog-code-analog system with weight redundancy based on current converters is considered. The features of the use of redundant positional number systems in analog-to-digital and digital-to-analog conversion are considered. The ability of redundant positional number systems to ensure the absence of «gaps» in the conversion characteristic caused by deviations of the real weights of digits from their calculated values, as well as the ability to significantly correct the initial errors of digital-to-analog converters, namely the linearity of the conversion characteristic, is indicated. The ability of redundant positional number systems to provide, despite the lengthening of the bit grid, a significant increase in performance is indicated. The scheme of analog-code-analogue converter based on redundant positional number systems with accumulation of buffer samples of signals in memory is presented. Calculation of errors of the analog-code-analogue signal reproduced by the converter is given. The relevance and practical expediency of using redundant number systems to obtain high bit capacity and speed of analog-code-analog systems using inaccurate components is substantiated.