Mismatch Calibration And Bandwidth Extension Methods For Time-Interleaved Analog-to-Digital Converters

Data acquisition instrument,employed as an essential constituent unit in mixedsignal processing systems,is the only bridge connecting the real analog world and the virtual digital world as well as the cornerstone of modern signal processing.Along with the increase of the frequency bandwidth and the instantaneous dynamic range for measuring,the requirement of the conversion speed as well as the conversion resolution of the sampling equipment have to be enhanced as far as possible.Due to the limitation of the manufacturing processes and materials,it is difficult even impossible to satisfy the requirement of the increasing sampling rate and resolution at the same time by utilizing a single analog-to-digital converter(ADC).Therefore how to solve the problem viz.,meeting both requirements on sampling rate and resolution with the current ADC's technology,becomes extremely urgent.Time-interleaved ADCs(TI-ADCs)utilizes multiple parallel ADCs that cooperate in a circulatory manner,which can increase the sampling rate by a factor of the number of the interleaved channel,thus can break through the bottleneck of single ADC's performance.However,this architecture is very sensitive to mismatches between the parallel channels,where a small deviation will result in a serious degradation in the overall TI-ADC's dynamic performance.Consequently,this dissertation is focused on the estimation and compensation algorithm for the channel mismatches as well as the expansion technique of the analog bandwidth.The main contributions are further explained below:Chapter 2 studies the system model of TI-ADCs and the effects of channel mismatches.We derive and demonstrate the quantitative relationship between the mismatch level and the dynamic performance.The derived formulas can be used to determine the mismatch tolerance for TI-ADCs,and provide the precision demand for system design procedure as well as the post calibration methods.Chapter 3 studies the correction methods for nonlinear mismatches in TI-ADCs,and proposes one foreground calibration method as well as an adaptive blind calibration method.The proposed foreground calibration method utilizes a training signal to estimate the mismatch coefficients,and then optimizes the system's dynamic performance by using a digital cascaded compensation structure.The results show that the proposed method can efficiently suppress the nonlinear mismatch distortions.Furthermore,compared with directly extending the linearization method for a single ADC into a time-interleaved sampling system,the proposed method can achieve the same performance by using less computational complexity.Further we propose a blind calibration method to deal with the time-variant nonlinear mismatch errors,which utilizes a certain degree of oversampling to acquire the mismatch distortions as well as the least mean square algorithm to track the mismatch variations.With the aid of discrete-time Fourier series,the M-periodic nonlinear errors are expressed with M time-independent mismatch coefficients,which substantially increases the utilization-efficiency of the analog bandwidth.We employ the complex conjugate relationship between the mismatch images to simplify the calibration structure for M-channel TI-ADCs.Simulation results demonstrate the calibration performance of proposed method with various input signals and mismatch levels.Chapter 4 presents and verifies a joint calibration method for the mixed linear and nonlinear mismatches.The proposed method also utilizes a certain degree of oversampling and the adaptive filtering structure to realize the background calibration for mixed mismatches.With mixed mismatches,the proposed method in this chapter can offer superior calibration performance than using the calibration method either for linear mismatches or nonlinear mismatches separately as well as using the calibration method for linear mismatches or nonlinear mismatches serially.Further,we consider the dependency of the mismatch parameter on the input frequencies,and extend the joint calibration method into frequency-dependent mixed mismatches applications.Simulation results illustrate the ability of proposed calibration method for mixed mismatches under different mismatch models,input signals as well as mismatch levels.Chapter 5 proposes a bandwidth expansion method by utilizing least-square-based finite-impulse-response(FIR)filters.To our best knowledge,we first derive and verify the order estimation formula for the least-square-based FIR filters.Based on the order estimates,we present the detailed configuration strategy for the filter design parameters.With the proposed configuration strategy,one can substantially reduce the design time,and the derived filter order estimation can help in providing an accurate evaluation of the computational resource from the top-level perspective.In consideration of different application requirements,we propose the order estimation and configuration strategy with priority to noise reduction and aliasing suppression,respectively.With the proposed order estimation and configuration strategy,one can accurately and quickly approximate the optimal filter design parameters,such as the minimal filter order,the weighting ratio,and the transition bandwidth,by using the given constraints of passband error and noise degradation,which can substantially reduce the design time and therefore enhance the capability for real-time systems.The performance of the filter order estimates and the configuration strategy of the filter design parameters for both application requirements is verified by comprehensive simulation examples.