Degree Type

Dissertation

Date of Award

2013

Degree Name

Doctor of Philosophy

Department

Electrical and Computer Engineering

First Advisor

Randall L. Geiger

Second Advisor

Degang J. Chen

Abstract

Spectral testing is widely used to test the dynamic linearity performance of Analog-to-Digital Converters (ADC) and waveform generators. Dynamic specifications for ADCs are very important in high speed applications such as digital communications, ultrasound imaging and instrumentation. With improvements in the performance of ADCs, it is becoming an expensive and challenging task to perform spectral testing using standard methods due to the requirement that the test instrumentation environment must satisfy several stringent conditions. In order to address these challenges and to decrease the test cost, in this dissertation, three new algorithms are proposed to perform accurate spectral testing of ADCs by relaxing three necessary conditions required for standard spectral testing methods. The testing is done using uniformly sampled points.

The first method introduces a new fundamental identification and replacement (FIRE) method, which eliminates the requirement of coherent sampling when using the DFT for testing the spectral response of an ADC. The robustness and accuracy of the proposed FIRE method is verified using simulation and measurement results obtained with non-coherently sampled data.

The second method, namely, the Fundamental Estimation, Removal and Residue Interpolation (FERARI) method, is proposed to eliminate the requirement of precise control over amplitude and frequency of the input signal to the ADC. This method can be used when the ADC output is both non-coherently sampled and clipped. Simulation and measurement results using the FERARI method with non-coherently sampled and clipped outputs of the ADC are used to validate this approach.

A third spectral test method is proposed that simultaneously relaxes the conditions of using a spectrally pure input source and coherent sampling. Using this method, the spectral characteristics of a high resolution ADC can be accurately tested using a non-coherently sampled output obtained with a sinusoidal input signal that has significant and unknown levels of nonlinear distortion. Simulation results are presented that show the accuracy and robustness of the proposed method.

Finally, the issue of metastability in comparators and Successive Approximation Register (SAR) ADCs is analyzed. The analysis of probability of metastability in SAR ADCs with and without using metastable detection circuits is provided. Using this analysis, it is shown that as the frequency of sampling clock increases, using a metastable detection circuit decreases the probability of metastability in SAR ADCs

Copyright Owner

Siva Kumar Sudani

Language

en

File Format

application/pdf

File Size

169 pages

Share

COinS