Energy-Efficient Sub-nV/√Hz Noise and Robust PSRR/CMRR Fully Differential CMOS Op-Amp for High-Fidelity Biomedical Signal Processing in IoT Devices
DOI:
https://doi.org/10.3103/S0735272725030057Abstract
This paper presents the design of an energy-efficient, fully differential CMOS operational amplifier (op-amp) fabricated in a 90 nm CMOS process, optimized for precision analog and biomedical applications, particularly within Internet of Things (IoT)-based health monitoring systems. The proposed architecture features a cross-coupled two-stage cascode topology with rail-to-rail input/output stages and single Miller compensation, ensuring frequency stability and wide dynamic range under low-voltage conditions. To further improve bias stability and signal integrity, the design integrates cascode current mirrors and a low-power common-mode feedback (CMFB) circuit that maintains output symmetry and stable common-mode levels. Operating at a 0.9 V rail-to-rail supply, the op-amp achieves a high open-loop gain of 86 dB, a power supply rejection ratio (PSRR) of 100 dB, and a common-mode rejection ratio (CMRR) exceeding 120 dB, thereby ensuring robust performance in noisy and variable environments common to battery-operated IoT devices. The mid-band input-referred noise is measured at 81.16 nV/√Hz, making the amplifier suitable for low-amplitude biomedical signal acquisition tasks, such as electrocardiograms (ECG), electroencephalograms (EEG), and electromyograms (EMG), which typically exhibit microvolt-level signal magnitudes. To enhance transient response, a slew-rate enhancement circuit is employed without compromising the power efficiency. The total power consumption is only 637 µW, with support for sub-µA quiescent operation, enabling compatibility with energy-constrained wearable and implantable biomedical platforms. With its combination of low noise, high gain, strong PSRR/CMRR, and low power operation, the proposed op-amp is well-suited as an analog front-end building block in next-generation biomedical sensors and mixed-signal IoT systems.