Advanced low-power filter architecture for biomedical signals with adaptive tuning

Area of Science:

• Integrated circuit design
• Biomedical engineering
• Analog signal processing

Background:

• Source-follower filters offer low power but are unsuitable for low-frequency biomedical signals.
• Biomedical applications require filters with low cutoff frequencies and minimal power consumption.
• Existing filter designs face challenges with process variations in weak inversion operation.

Purpose of the Study:

• To present a low-power, second-order composite source-follower-based filter architecture.
• To optimize the filter for biomedical signal processing, specifically Electrocardiogram (ECG) and Electroencephalogram (EEG) applications.
• To enable filter operation in the weak inversion region for reduced area and power dissipation.

Main Methods:

• The filter is designed to operate in the weak inversion zone, reducing capacitor area and power dissipation.
• A compensation technique based on magnitude comparison is employed to address process, voltage, and temperature variations.
• The design utilizes UMC-0.18μm CMOS technology with a 1.0V supply.

Main Results:

• Achieves a tunable cutoff frequency range from 0.5 Hz to 150 Hz.
• Exhibits extremely low power dissipation of 6nW at 150 Hz.
• Demonstrates a compact silicon area of 0.065 mm², a dynamic range of 75 dB, and THD below -40 dB for a 300 mVpp signal swing.
• Validated robustness against process, voltage, and temperature variations through on-chip tuning.

Conclusions:

• The proposed filter architecture is a promising solution for low-power biomedical devices.
• The weak inversion operation and compensation technique enable efficient and stable performance for biomedical signal processing.
• The filter's characteristics meet the demanding requirements of portable and implantable medical electronics.