A Bias-Electrode-Free Multichannel ExG Readout IC with Time-Multiplexed PAM-4 Body Channel Communication

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Human-Machine Interfaces

Background:

• Wearable multimodal human-machine interfaces (HMIs) like extended reality (XR) and brain-computer interfaces (BCIs) require integrated physiological signal recording.
• Body Channel Communication (BCC) transmitters can introduce common-mode interference (CMI) that degrades the quality of ExG signals (EEG, EOG, EMG, ECG).

Purpose of the Study:

• To develop a multi-channel ExG recording system integrated with BCC for advanced wearable HMIs.
• To suppress CMI from the BCC transmitter while maintaining ultra-low-noise ExG acquisition.

Main Methods:

• A time-multiplexed acquisition scheme separates ExG recording and BCC transmission into distinct time slots.
• Pulse Amplitude Modulation-4 (PAM-4) signaling achieves a 10-Mbps data rate for transmitting multi-channel ExG data.
• A bias-electrode-free ExG analog front end (AFE) enhances BCC signal amplitude.
• A charge-pump-based CMI cancellation loop with a multi-channel least-mean-square (LMS) filter and a DC servo loop mitigate interference and mismatch.

Main Results:

• The system enables reliable recording of EEG, EOG, EMG, and ECG with ultra-low noise.
• Achieved a 10-Mbps data rate using PAM-4 signaling.
• The bias-electrode-free AFE increased BCC signal amplitude by approximately 2.13×.
• Demonstrated a total common-mode rejection ratio (CMRR) of 101.1 dB and suppressed electrode DC offsets up to 500 mV.
• Maintained stable operation under 18-Vpp CMI and 15% electrode mismatch.

Conclusions:

• The proposed integrated ExG recording and BCC system effectively suppresses CMI, enabling high-fidelity physiological signal acquisition for wearable HMIs.
• The novel AFE and CMI cancellation techniques provide a robust solution for noise reduction and reliable data transmission in challenging wearable environments.