A Pre-Processing Signal Modeling Approach for Multichannel ATI Digitizers

Abstract

Asynchronous time-interleaved (ATI) analog-to-digital converter (ADC) architecture utilizes pre-processing, low-pass filtering, and ADCs to simultaneously enhance the sampling rate and bandwidth of the acquisition system. We propose a novel approach to generate pre-processing signal in multichannel ATI architecture. The generated pre-processing signal is no longer limited to pulse wave or square wave with a fixed duty. We derive the condition for reconstructing input signal as the invertibility of the pre-processing matrix. Additionally, we analyzed the error sources in the multichannel ATI. Based on the spectral shift characteristics of the ATI, we propose an error coefficient estimation method. Using the estimated error coefficient, a digital filter can be designed to calibrate the system. A four-channel hardware prototype was developed to evaluate the performance of the proposed method, satisfactory results were obtained and the feasibility of the proposed approach was demonstrated.