Decoupling capacitive sensing from circuit knowledge via time-domain representation
摘要
Over nearly a century, capacitive sensor readout has remained largely tied to absolute capacitance extraction and circuit-specific parameters. Here, we report a unified time-domain sensing framework that ingeniously leverages the intrinsic threshold voltage of microcontroller I/O pins, enabling a direct mapping between sensor charging dynamics and the desired physical quantities, without requiring any circuit knowledge. Circuit variations manifest primarily as time-axis scaling, and can be efficiently corrected using simple two-point anchoring. We experimentally validate the framework using non-contact liquid-level sensor, commercial capacitive humidity sensor, and linear displacement sensor. Despite substantial hardware differences, consistent sensor responses are recovered through two-point temporal anchoring, without re-identifying circuit parameters or recalibrating sensing models. By shifting the sensing representation from absolute capacitance to time-domain observables, this work challenges the long-standing capacitance-centric readout paradigm and points to a system-level paradigm shift toward robust, scalable, and platform-independent sensing. The proposed framework significantly reduces the complexity and maintenance overhead associated with capacitive sensors across diverse application domains, making it broadly applicable across nearly all industrial sectors.