An All-Digital Background Calibration Technique for M-Channel Downsampling Time-Interleaved ADCs Based on Interpolation
Chenxi Han,Shubin Liu,Yuhao Zhang,Lichen Feng,Hongzhi Liang,Zhangming Zhu
DOI: https://doi.org/10.1109/tcsii.2022.3233413
2023-01-01
IEEE Transactions on Circuits & Systems II Express Briefs
Abstract:An all-digital background calibration technique for gain, timing, and bandwidth mismatches in the time-interleaved analog-to-digital converter (TI-ADC) is presented in this brief. The proposed calibration technique employs an interpolation filter and a Hilbert filter to generate a reference sequence for each sub-ADC. Subsequently, adaptive filters are used to calibrate the mismatch errors, the coefficients of which are updated by the least mean square (LMS) algorithm. The advantage of the technique is that it only requires the output sequence of the TI-ADC without any additional information or extra reference channels. Moreover, it is effective for the $M$ -channel TI-ADC sampling of single-tone, multi-tone, and wideband signals in any Nyquist Band. The frequency shifting is analyzed via filter bank theory, avoiding the complex signal introduction. The simulation for 4-channel TI-ADC with 5 sinusoidal input signals, shows that the proposed technique can improve the SNDR and SFDR by 37 dB and 54 dB, respectively.