Mismatch Compensation Method Of Time-interleaved Sampling System Based On The Calibration Signal

Limited by the current fabrication technique of semiconductor,utilizing a single ADC is unable to meet the requirements of high resolution and high sampling rate at the same time.Thus,it is an effective way to achieve high-speed and high-precision sampling by employing parallel ADCs.Nonetheless,due to the impact of the mismatch errors among the sub-channels,the performance of TIADC has been greatly reduced.For the TIADC mismatch errors presented in the system,the existing estimation methods generally require interrupting the normal signal acquisition process for error estimation.This severely restricted the application of the TIADC system in real-time acquisition area.This thesis focuses on analyzing the sources of various channel mismatches as well as the effects of these mismatches.Through the analysis of the spectrum of the mismatch errors,this paper further clarifies the correspondence between the mismatch parameter and spectral distortion.Based on the feature of the foreground estimation methods,this paper improves the divisional sampling structure to breaking the frequency limitations in the conventional divisional sampling structure.Through the adaptive feedback control of the calibration signal frequency,we get the calibration signal which fit the target signal in the certain frequency band.By analyzing the spectrum of the I/O based on calibration signal,the corresponding transfer function of the mismatch in the specific frequency band can be achieved.Using this function,the output spectrum sampled with the target signal can be compensated.Subsequently,this paper improves the divisional sampling structure,which can estimate the mismatch error without interrupting the target signal acquisition.In this division structure the test signal is switched to sampling channels in certain sampling instant,ensuring the normal collection of the target signal.Later,this paper studies on the functional realization of this TIADC structure as well as the technical difficulties need to be addressed,which includes: 1.The prime relationship between the sampling frequency of the two input signals.2.How to reconstruct the input signal from the sample sequence with some missing points.3.The application conditions and performance of the compensation algorithms.Finally,this paper tests and verifies the error estimation algorithm applied in the improved system,and verifies the feasibility of the algorithm within a dual signal division acquisition system.As the transfer function of mismatch obtained,we compensate the output spectrum and verify the compensation algorithms based on transfer function of mismatch.The main achievements of this thesis include: Propose an improved divisional sampling structure with an adaptive feedback of the frequency of the target signal,adjusting the frequency composition of the calibration signal,which fitting the frequency characteristics of the target signal.Then,using the adjusted calibration signal as a reference input,by comparing the input and output spectrum,we get the mismatch transfer function in the corresponding frequency band.Finally we take the advantage of this transfer function and compensates the channel mismatch errors in time-interleaved sampling system.On the other hand,this thesis has investigated the key technique of the improved divisional sampling structure experimentally,laid an illustriousness foundation for the realization of the Time-interleaved sampling system.