Research Of Estimate And Synthese Calibration Method For Nonuniform Errors Of Parallel ADC Acquisition System

With the wide application of digital signal, the rapid development of digital signal processing technology is promoted. ADC is an important part of digital signal system and its speed and precision are required higher and higher, which develops along with the UWB radar technology and software radio technology. But now, most ADC can’t satisfy the requirement of high precision and speed at the same time. Parallel ADC acquisition technology is an effective method that improve the sampling rate and guarantee high precision. By the restriction of manufacturing engineering and production levels, the performance of each ADC can not totally identical. So, the mismatch errors exist between channels, such as time base error, gain error and offset error and so on, which causes the sampling signal heterogeneous and seriously degrade the performance of the system. In order to solve the problem of nonuniform errors of sampling system, an adaptive synthese calibration method is proposed in this paper.This paper is mainly divided into three aspects. Firstly, through the research of parallel ADC sampling system’s structure and basic principles, established the parallel ADC sampling system model based on Matlab. Profoundly analyzed the process of sampling and the origination of three kinds of typical nonuniform errors, according to the nonuniform signal frequency spectrum characteristic, deduced the computation of ADC system performance.Secondly, the traditional estimation and calibration methods of nonuniform errors are researched and their effect are evaluated. Then, established the adaptive estimation and comprehensive calibration model of nonuniform errors. There are two algorithms used to calibrate errors, which are traditional LMS (Least-Mean Square) adaptive algorithm and improved LMS adaptive algorithm. Considering the influence of three kinds of error, the adaptive algorithm estimate and calibrate three kinds of nonuniform error at one time.Finally, according to the proposed algorithm, simulation experiment, which is done under different sampling environment, such as different input signals, different sampling frequency and so on, is based on Matlab simulation platform. The feasibility and reliability of the algorithm can be proved by the calibration results which reflect the main performance index of system.