A frequency response mismatches estimation method for time interleaved analog-to-digital converter based on chirp-Z...

Area of Science:

• Electrical Engineering
• Signal Processing

Background:

• Time-interleaved analog-to-digital converters (TIADCs) are crucial for high-speed data acquisition.
• Accurate channel mismatch estimation is vital for TIADC performance, but traditional methods like FFT suffer from spectral leakage and fixed resolution.
• Existing methods struggle with precise calibration, especially near Nyquist boundaries.

Purpose of the Study:

• To propose a novel frequency-response mismatch estimation method for TIADCs using the Chirp-Z Transform (CZT).
• To demonstrate the CZT's advantages over FFT in terms of flexible frequency resolution and reduced spectral leakage.
• To validate the proposed method's effectiveness through simulations and hardware experiments.

Main Methods:

• Utilizing the Chirp-Z Transform (CZT) for flexible frequency mapping and high-precision spectral analysis.
• Adjusting CZT parameters (start frequency, bandwidth, resolution) to avoid spectral leakage.
• Comparing CZT-based estimation with traditional FFT methods in simulations and on a 4-channel 12.5 GS/s TIADC prototype.

Main Results:

• The CZT-based method significantly reduces spectral leakage, improving spurious-free dynamic range (SFDR) by approximately 7 dB over FFT in simulations.
• Hardware experiments on a TIADC prototype showed a 20 dB enhancement in SFDR after calibration using the CZT method.
• The proposed approach achieves high-resolution, leakage-free spectral analysis for accurate mismatch calibration.

Conclusions:

• The CZT-based frequency-response mismatch estimation is a highly accurate and efficient method for TIADCs.
• This novel approach overcomes the limitations of FFT-based methods, offering superior calibration performance.
• The CZT provides a robust solution for precise channel mismatch calibration in broadband TIADC systems.