In a parallel time-interleaved data sampling system, timing and amplitude mismatches of this structure degrade the performance of the whole ADC system. In this paper, an adaptive blind synthesis calibration algorithm is proposed, which could estimate the timing, gain and offset errors simultaneously, and calibrate automatically. With no need of an extra calibration signal and redesign, it could efficiently and dynamically track the changes of mismatches due to aging or temperature variation. A fractional delay filter is developed to adjust the timing mismatch, which simplifies the design and decreases the cost. Computer simulations are also included to demonstrate the effectiveness of the proposed method.

Sample-time errors can greatly degrade the dynamic range of a time-interleaved sampling system. In this paper, a novel correction technique employing a cubic spline interpolation is proposed for inter-channel sample-time error compensation. The cubic spline interpolation compensation filter is developed in the form of a finite-impulse response (FIR) filter structure. The correction method of the interpolation compensation filter coefficients is deduced. A 4GS/s two-channel, time-interleaved ADC prototype system has been implemented to evaluate the performance of the technique. The experimental results showed that the correction technique is effective to attenuate the spurious spurs and improve the dynamic performance of the system.