An Inductively Powered Scalable 32-Channel Wireless Neural Recording System-on-a-Chip for Neuroscience Applications

Area of Science:

• Neuroscience
• Electrical Engineering
• Biomedical Engineering

Background:

• Wireless neural recording systems often require animals to carry bulky batteries, limiting recording duration and animal behavior.
• Existing systems may suffer from power inefficiency and unstable power supply, impacting data quality and experimental feasibility.

Purpose of the Study:

• To develop an inductively powered wireless integrated neural recording system-on-a-chip (WINeR SoC) for small freely behaving animals.
• To enhance system power efficiency and enable long-duration neural recordings without external batteries.

Main Methods:

• Utilized time-division multiplexing and a novel power scheduling method to reduce power consumption in low-noise amplifiers (LNAs).
• Incorporated a high-efficiency active rectifier with optimized coils for improved power transfer and stable power supply.
• Implemented the system on a 0.5-μm standard CMOS process, measuring 4.9×3.3 mm².

Main Results:

• Achieved a total SoC power consumption of 5.85 mW at ±1.5 V with 12 out of 32 LNAs active.
• Demonstrated a measured input-referred noise of 4.95 μVrms (1 Hz-10 kHz) with inductive powering at a 7-cm coil separation.
• The closed-loop controlled active rectifier ensured stable power delivery despite coil displacements.

Conclusions:

• The developed WINeR SoC offers a battery-free, power-efficient solution for wireless neural recording in freely behaving animals.
• This technology facilitates extended recording sessions, crucial for studying long-term neural dynamics.
• The system's design addresses key limitations of current neural recording technologies, paving the way for advanced neuroscience research.