A Statistics-Based Background Timing Skew Calibration Algorithm for Time-Interleaved ADCs

Abstract

A modified statistics-based background calibration algorithm is proposed, which estimates and compensates errors due to the timing skew of time-interleaved ADCs in a digital-mixing way. The mismatch-induced error signal is corrected by calculating the cross-correlation between each sub-ADC and an additional calibration ADC. The output of the supplementary channel is delayed during data processing, different from the conventional statistics-based method that is susceptible to quantization noise. Simulation results of an 8-bit 2-GS/s four-channel time-interleaved ADC behavioral model show that the proposed method is effective and efficient.