Linearity enhancement of multibit ΔΣ A/D and D/A converters using data weighted averaging

Abstract

A dynamic element matching algorithm, data weighted averaging, is introduced for use in multibit /spl Delta//spl Sigma/ data converters. Using this algorithm, distortion spectra from DAC linearity errors are shaped by first-order noise shaping, resulting in a dynamic range improvement of 9 dB/octave when DAC errors dominate. Combining this technique with random dithering nearly eliminates the aliasing of the DAC errors into the baseband. Simulations show that with only 1% element matching 110 dB signal-to-noise ratio (18 b) is achieved for a third-order 3-b modulator with an oversampling ratio of 128.