BCIwiki (bciwiki.com) — A multichannel ExG readout chip records EEG, EOG, EMG and ECG and transmits the data through the wearer's body at 10 Mbps, using separate time slots for recording and transmission, researchers reported in IEEE Transactions on Biomedical Circuits and Systems on July 29, 2026. The system targets wearable multimodal human-machine interfaces such as extended reality and brain-computer interfaces (BCIs).
Body channel communication (BCC) uses the human body as the transmission medium. Its drawback is that transmitter noise shows up as common-mode interference at the ExG analog front end, where it can swamp faint EEG and EOG signals. The team's time-multiplexed scheme keeps recording and transmission apart in time, preserving ultra-low-noise acquisition.
Because the transmitter must send multichannel data within a limited slot, it needs a high data rate. PAM-4 signaling delivers 10 Mbps. That is made possible by a bias-electrode-free front end, which raises the BCC signal amplitude by about 2.13 times and leaves enough margin to tell the 4 signal levels apart. A charge-pump-based interference cancellation loop with a multichannel least-mean-square (LMS) filter, plus a DC servo loop, reduces inter-channel mismatch and suppresses differential-mode artifacts.
Fabricated in 180-nm CMOS, the chips achieve a total common-mode rejection ratio (CMRR) of 101.1 dB, suppress electrode DC offsets up to 500 mV, and remain stable under 18-Vpp common-mode interference and 15% electrode mismatch. The abstract reports chip-level benchmarks and does not describe BCI tasks performed by wearers.