A Direct Digitizing Chopped Neural Recorder Using a Body-Induced Offset Based DC Servo Loop.

Area of Science:

• Biomedical Engineering
• Neuroscience
• Integrated Circuit Design

Background:

• Electrode offset (EDO) is a significant challenge in neural recording, often requiring AC coupling capacitors that limit bandwidth and increase power consumption.
• Existing neural recorders face limitations in input impedance, bandwidth, and power efficiency, hindering their application in chronic brain-computer interfaces.

Purpose of the Study:

• To present a direct digitizing neural recorder with on-chip electrode offset cancellation.
• To demonstrate a novel architecture that eliminates the need for AC coupling capacitors, enabling high input impedance and wide bandwidth operation.
• To achieve low power consumption and low input-referred noise for effective local field potential (LFP) recording.

Main Methods:

• A body-induced offset based DC servo loop was designed to counteract the electrode offset (EDO) on-chip.
• The neural recorder architecture utilizes the bulk of the input pair to generate a counteracting offset.
• The design was implemented in a 180 nm HV-CMOS process, integrating chopping without impedance boosting.

Main Results:

• The prototype achieved a high input impedance of 238 MΩ over a 10 kHz bandwidth without AC coupling capacitors.
• The device demonstrated low power consumption of 12.8 μW and a small silicon area of 0.02 mm².
• Input-referred noise in the local field potential (LFP) band was measured at 1.82 μV/√Hz, with a Neural Efficiency Factor (NEF) of 5.75.

Conclusions:

• The developed direct digitizing neural recorder effectively cancels electrode offset on-chip, overcoming a major limitation in neural recording.
• The architecture's ability to maintain high input impedance and wide bandwidth with low power consumption makes it suitable for advanced neural interfaces.
• This technology offers a promising solution for miniaturized, high-performance neural recording systems.