Presilicon Circuit-Aware Linear Least Squares Spectral Analysis for Time-Based Data Converters

Presilicon time-based data converter spectral analysis usually incurs long simulation times. This brief shows an approach for the estimation of the data converter's spectral components and noise floor based on a small simulation window. Additionally, it provides two useful estimates of the technique's expected accuracy. The method is based on circuit-aware linear least squares spectral analysis. It can be applied together with the accuracy estimate and knowledge of the noise spectrum's shape during the design time of time-based data converters. It provides a significant advantage over classic fast Fourier transform-based techniques for time-based data converters. Both a theoretical background and simulation results are provided for a voltage-controlled oscillator (VCO)-based analog-to-digital (A/D) converter. The technique provides a good accuracy for both the spectral components and the data converter's noise floor. The accuracy estimate expresses the technique's performance, achieving a worst case error of only 5 dB, which proves its usability.