A Noval Calibration Based on Genetic Algorithm for Pipelined ADC

This paper introduces a multi-dimensional multiple digital-to-analog converter (MDAC) calibration for pipelined analog-to-digital converter (ADC). The proposed multi-dimensional MDAC digital domain compensation model can be compatible with pseudo-random number (PN) injection error compensation while realizing capacitance mismatch and static inter-stage gain error compensation, and can also be successfully connected with background calibration for dynamic inter-stage gain error. A multi-dimensional coefficient extraction based on genetic algorithm is also proposed, which can accurately extract the required compensation coefficient and avoid falling into local traps. The test results show that the signal-to-noise-and-distortion ratio (SNDR)/spurious-free dynamic range (SFDR) are increased by 7.24 dB/11.91 dB respectively with foreground calibration at 25°C; With the help of background calibration at 85°C, SNDR and SFDR are increased by 3.13 dB and 5.15 dB.