Polynomial Digital-to-Analogic Converter

Abstract

This work presents a new topology for a digital-to-analog converter (DAC) based on polynomial reconstruction. The proposal is to reconstruct a signal with the information of amplitude, time and exponent. The general idea is to use a reconstruction based on signal morphology using a polynomial approximation method, thus being able to significantly reduce the sample rate of the D/A converter. The system generates four base waves and combining them it is possible to generate any exponent. In this work, 16 exponents spaced in different ways, ranging from 2 to 0.25, were tested. The system result shows an RMS error of less than 636 µV for the sharpest curve. A reconstruction of an ECG signal is performed where there is a reduction in the sampling rate of 95.08% compared to a converter with a linear sampling rate of 2 KS/s.