Effects of coefficient quantisation on a general class of sigma-delta based data converters-a comparative study

This paper presents results of a comparative simulation study. The aim of this study is to provide a framework for efficient architectural realization of low-pass (LP) digital modulators. To achieve this the study identifies which topologies are least sensitive to coefficient truncation and/or power-of-two quantisation. A further aim is to identify which coefficients within the topology have the greatest influence on the maximum achievable signal-to-noise ratio (SNR) and the stability of the modulator. To this end a general class of sigma-delta (/spl Sigma/-/spl Delta/) modulator whose output is single-bit and signal-transfer function (STF) is unity is investigated using Simulink/sup TM/ and Matlab/sup TM/ based behavioral simulations. The results suggest that the distributed weighted feedback type topologies may be advantageous. Modulators with power of two coefficients have also been derived empirically and can achieve an SNR of up to 101 dB with a maximum stable input equal to 72.5% of the FS feedback signal.