Search

This paper proposes a full-band calibration scheme of timing mismatch for Time-Interleaved Analog-to-Digital Converters (TI-ADC) based on Automatic Identification (AI) detection scheme. Besides estimating the value of timing mismatch, AI detection scheme also judges the odd–even property of the Nyquist zone (NZ) which the input signal belongs to and thus adaptively adjusts the calibration polarity for full-band application. On the other hand, Successive-Approximation-Register (SAR) correction technique is employed to speed up the convergence process of calibration with low cost. The efficiency of the proposed calibration scheme is verified by MATLAB simulation and implementation on PCB. Both results show that with an input signal whose bandwidth is within any NZ, the proposed calibration methodology is effective. Compared with the traditional calibration schemes, the proposed calibration method achieves fast convergence speed with 6 × 1 0 3 samples and costs less hardware with 2.1 k gate counts. [ABSTRACT FROM AUTHOR]