Continuous-Time Acquisition of Biosignals Using a Charge-Based ADC Topology.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Hardware Implementation

Background:

• Conventional fixed-rate digitization of biosignals can be inefficient in terms of bandwidth and power consumption.
• Activity-dependent and nonuniform sampling present an alternative to improve signal acquisition efficiency.

Purpose of the Study:

• To investigate continuous-time (CT) signal acquisition as an activity-dependent, nonuniform sampling method.
• To demonstrate bandwidth savings and resource-efficient hardware implementation for biosignal acquisition.
• To propose and validate a novel charge-based CT analog-to-digital converter (ADC) for neural signals.

Main Methods:

• Quantification of bandwidth saving using nonuniform quantization on electrocardiogram (ECG) and extracellular action potential signals.
• Description of CT sampling properties, including bandwidth reduction, quantization error mitigation, and aliasing impact.
• Design and implementation of a novel charge-based CT ADC in 0.35 μm CMOS technology.

Main Results:

• Achieved compression ratios of 5 for ECG and 26 for extracellular action potentials.
• Demonstrated an 8-bit resolution, 4 kHz bandwidth, and low static power consumption (3.75 μW).
• Silicon-verified measurements confirmed activity-dependent dynamic power dissipation (1.39 pJ/conversion).

Conclusions:

• CT signal acquisition is a viable, efficient alternative to conventional digitization for biosignals.
• The proposed CT ADC offers a resource-efficient hardware solution for neural signal acquisition.
• This technology enables significant bandwidth reduction and power savings, particularly for activity-dependent signals.