Adaptive Resolution ADC Array for an Implantable Neural Sensor.

Area of Science:

• Biomedical Engineering
• Neurotechnology
• Integrated Circuit Design

Background:

• Implantable neural sensors require efficient analog-to-digital conversion (ADC) for signal digitization.
• Existing ADCs often face power consumption challenges in high-channel-count neural recording systems.
• Variable resolution ADCs offer a potential solution for optimizing power and data acquisition.

Purpose of the Study:

• To develop and characterize an analog-to-digital converter (ADC) array with variable resolution for implantable neural sensors.
• To implement a resolution adaptation algorithm that dynamically adjusts ADC settings based on neural signal content.
• To evaluate the power efficiency and performance of the proposed ADC array architecture.

Main Methods:

• Designed a 96-cell ADC array using a successive approximation charge redistribution architecture in 0.13 μm CMOS technology.
• Developed and implemented a resolution adaptation algorithm for periodic recalibration without additional hardware.
• Characterized the ADC array's performance, including resolution, power consumption, and energy per conversion.

Main Results:

• The ADC array features 96 variable resolution ADC base cells, each capable of 3 to 8 bits.
• Achieved a 100 kS/s sampling rate with power consumption ranging from 0.23 μW to 0.90 μW.
• Demonstrated a 2.3x reduction in power consumption for motor neuron signals through resolution adaptation, maintaining an effective 7.8-bit resolution.

Conclusions:

• The variable resolution ADC array effectively digitizes neural signals while significantly reducing power consumption.
• The resolution adaptation algorithm optimizes ADC performance based on neural data content, enhancing efficiency.
• This technology is promising for next-generation, low-power implantable neural interfaces.