Advanced Noise-Shaping Data Conversion Techniques

Abstract

The rapidly increasing digital device market is creating a huge demand for Data Conversion. In high-resolution data conversion applications, noise-shaping techniques are essential. This thesis first goes through the fundamentals of data conversion and noise-shaping techniques, and then discusses the latest advanced noise-shaping techniques that further boost data converter performance. Three new noise-shaping techniques (architectures) are introduced. The first technique boosts the order of the noise-shaping system. A system-level approach is used to achieve high-order noise-shaping with enhanced robustness. A prototype noise-shaping SAR ADC is designed and measured. This design is the first 4th-order noise-shaping SAR ADC published. The second technique effectively increases the bandwidth of noise-shaping data converters using a time-interleaving framework. A prototype time-interleaved noise-shaping SAR ADC is designed and measured to illustrate the technique. This prototype has the highest bandwidth among noise-shaping SAR ADCs to date. The final technique aims to solve some inherent problems in continuous-time noise-shaping systems. A Continuous Time - Discrete Time (CT-DT) hybrid noise-shaping architecture is introduced, which adopts the advantages of both conventional CT and DT structures. A prototype ADC, showing benefits of tuning-free operation and high tolerance to non-idealities.