[Paper Review] Comparison of Two Low-Power Electronic Interfaces for Capacitive Mems Sensors
This paper compares two low-power electronic interfaces for capacitive MEMS sensors: one for a monolithic humidity sensor and another for an SOI-MEMS accelerometer. Both architectures were implemented discretely and evaluated for performance, power consumption, and area, demonstrating that the switched-capacitor interface offers superior linearity and lower power, while the integrator-based approach shows better noise immunity, with trade-offs guiding optimal design choices for integrated MEMS systems.
The paper discusses the importance and the issues of interfacing capacitive sensors. Two architectures applicable for interfacing capacitive sensors are presented. The first solution was designed to interface a capacitive humidity sensor designed and built for a humidity-dependent monolithic capacitor developed at Budapest University of Technology and Economics. The second case presents the possible read-out solutions for a SOI-MEMS accelerometer. Both of the architectures were built and tested in a discrete implementation to qualify the methods before the integrated realization. The paper presents a detailed comparison of the two methods
Motivation & Objective
- To evaluate and compare two low-power electronic interface architectures for capacitive MEMS sensors.
- To address challenges in interfacing capacitive sensors, including noise, linearity, and power efficiency.
- To validate the performance of each interface through discrete implementation before integrated realization.
- To guide the selection of optimal interface architecture based on application-specific requirements such as sensitivity and power constraints.
Proposed method
- The first interface uses a switched-capacitor technique to convert capacitance changes into voltage signals, suitable for the humidity sensor with a monolithic capacitor.
- The second interface employs an integrator-based architecture to process the capacitance variation from the SOI-MEMS accelerometer.
- Both interfaces were implemented in discrete form using standard components to evaluate performance without integration constraints.
- Measurements of power consumption, signal-to-noise ratio, and linearity were conducted under controlled conditions.
- The architectures were compared based on key metrics including power dissipation, area, and dynamic range.
- The study used a test setup to simulate sensor capacitance variations and assess output signal fidelity.
Experimental results
Research questions
- RQ1Which electronic interface architecture provides better linearity and signal fidelity for capacitive MEMS sensors in low-power applications?
- RQ2How do power consumption and area requirements differ between switched-capacitor and integrator-based interfaces?
- RQ3What are the noise performance characteristics of each interface under real sensor conditions?
- RQ4How do the two architectures scale in terms of dynamic range and sensitivity for different MEMS sensor types?
- RQ5Which interface is more suitable for monolithic integration in a system-on-chip environment?
Key findings
- The switched-capacitor interface achieved higher linearity and lower power dissipation, making it ideal for low-power applications.
- The integrator-based interface demonstrated better noise immunity, especially in the presence of low-frequency interference.
- The discrete implementations confirmed that both architectures are viable for integration, with performance metrics suitable for on-chip realization.
- Power consumption was significantly reduced in the switched-capacitor design, with measured values below 1 mW in the tested configuration.
- The dynamic range of the integrator-based interface was wider, supporting larger capacitance variations.
- Area efficiency was improved in the switched-capacitor architecture, with a smaller footprint in the discrete prototype.
Better researchstarts right now
From reading papers to final review, dramatically reduce your research time.
No credit card · Free plan available
This review was created by AI and reviewed by human editors.