An ultra-low power bioelectrical potential acquisition analog front-end chip for wearable devices

To enable long-term continuous monitoring of bioelectrical signals such as wearable electrocardiogram (ECG), electromyography (EMG), and electroencephalography (EEG), this paper proposes and implements a bioelectrical acquisition analog front-end (AFE) chip operating at 0.8 V supply with microwatt-level power consumption. The AFE is centered around a dynamically biased chopping stabilized capacitor-coupled instrumentation amplifier (CCIA), which explicitly decouples the processing of electrode DC offset (EDO) and baseline drift into "steady-state slow-loop (DSL) baseline suppression" and "event-triggered fast-recovery path" without sacrificing input impedance or common-mode rejection. The slow loop provides ultra-low cutoff and extremely low discharge to maintain GΩ-level input impedance; when EDO steps or saturation edges occur, the fast recovery path temporarily reduces the equivalent time constant and automatically exits after recovery, thereby avoiding the second-level recovery window caused by traditional ultra-low cutoff DSL. The chip is fabricated using a 0.18 μm CMOS process, with an AFE core area of 0.35 mm² and a total system power consumption of 1.2 μW under 0.8 V supply. The measured results demonstrate that under the programmable gain range of 60–100 dB and bandwidth modes from 0.5–100 Hz (ECG/EEG) to 7 kHz (EMG), the input reference integration noise is 0.98 μV_rms (0.5–100 Hz). The common-mode rejection ratio (CMRR) exceeds 125 dB at 10 Hz, with approximately 125 dB at 50/60 Hz. The power supply rejection ratio (PSRR) surpasses 98 dB at 10 Hz, and the input impedance exceeds 1 GΩ at 10 Hz. Under step-type EDO disturbance injection, the fast recovery mechanism in this paper can bring the output back to the linear operating region within approximately 60–70 ms, whereas only the DSL mode exhibits second-level recovery. These results validate the integrated design achieved under a 0.8 V microwatt budget, which enables low noise, high input impedance, and rapid robust recovery, making it suitable for long-lasting wearable physiological monitoring systems.