Parallel Adc Sampling Channel Mismatch Error A Real-time Estimates And Correction Method

One of the most effective methods to improve the system's sample-rate is the time-interleaved ADC. However, channel mismatch errors of this structure, including time error, gain error and offset error will seriously degrade the performance of the whole system. Real time and accurate estimation and correction of these errors are the key to obtain accurate signal reconstruction.Firstly, the three typical errors caused by channel mismatch is analyzed in this paper, then the sampling signal's spectrum formula under the condition of channel mismatch errors has been derived, and further estimation method of the three kinds of errors has been established. The estimation algorithm which including Fourier transform algorithm, judgement of spurious frequency position, Fourier inverse transform, mode and phase calculation etc can estimate these errors fastly and real timely. The program of error correction proposes corresponding correction methods for each error characteristics. Compared with linear gain and offset error characteristics, time error correction is more difficult. A time error correction method of fractional delay filter based on Farrow structure has been proposed, which uses time error as the filter input. It need not update the error filter coefficients even if the time error changed. So the method can ensure that time error correction in real time.In order to realize the hardware design of real-time estimation and correction of channel mismatch, the error estimation algorithm and the Farrow structure filter correction which has advantages of simple topology and design process and low resource usage rate is applied. With DSP blocks and other logic resources in FPGA, the hardware design and verification of error estimation and correction has been completed.Using verification test, further we analyze the accuracy of error estimation algorithm, the impact of the sampling rate to signal frequency ratio and Farrow structure parameters on correction results and the correction method application based on Farrow structure filter.