Redundancy effect on the performance of digitally-assisted SAR ADCs

This paper presents a thorough assessment of the impact redundancy has on background calibration performance in SAR ADCs with static and dynamic non-idealities. A configurable model that is capable of implementing different SAR architectures is used to inject various implementation impairments and and analyze the impact redundancy has on them. The effects of a redundant capacitor array on calibration are first analyzed in the presence of static non-idealities, after which then incomplete settling effects and noise are added. The effect of redundancy in the spread of the calibrated ENOB of multiple simulations is assessed. A tradeoff between increased tolerance to settling errors and calibration performance is shown, allowing to optimize the amount of redundancy and number of SAR cycles for minimum energy consumption for a given target ENOB.