How MEMS sensors work: a guide for engineers

How does a MEMS sensor work?

Micro-electromechanical systems (MEMS) are fundamental components of modern engineering, enabling advanced functionality in devices across industrial, automotive, and healthcare sectors. These microscopic systems integrate mechanical elements with electronics on a single silicon chip, allowing them to sense and interact with their environment with remarkable precision. For electronic engineers, understanding the operational principles of these devices is crucial for optimizing system performance and unlocking new design possibilities.

This article details how a MEMS sensor works, from its core components to the transduction mechanisms that convert physical stimuli into electrical signals. We will explore the working principle of MEMS and explain how these sensors operate to provide real-time, high-fidelity data that drives countless applications.

The core components of a MEMS sensor

At its foundation, MEMS functionality depends on the integration of mechanical and electrical components fabricated at the microscale. While designs vary based on the specific application, most MEMS sensors consist of three primary functional units.

1. Micro-mechanical structure: this is the primary sensing element. It is a precisely engineered structure that physically responds to an external stimulus. Examples include cantilevers, diaphragms, proof masses, or comb drives. These structures are typically manufactured from silicon or silicon-based compounds due to their mechanical properties and compatibility with semiconductor fabrication processes. The movement or deflection of the structure is the initial step in the sensing process.
2. Transduction mechanism: the transducer converts the physical movement or change in the mechanical structure into a measurable electrical signal. This is the critical link between the mechanical and electrical domains. The method of transduction is a defining characteristic of a MEMS sensor and directly influences performance parameters, such as sensitivity, power consumption, linearity, and thermal stability.
3. Application-specific integrated circuit (ASIC): the ASIC is the electronic brain of the sensor. It receives the raw electrical signal from the transducer and performs signal conditioning. This includes amplification to boost the signal strength, filtering to remove unwanted noise, and analog-to-digital conversion (ADC) to create a digital output. The ASIC processes this data and communicates it to a host processor, delivering a clean, usable signal for the application. In advanced systems, the ASIC can also embed features like a machine learning core (MLC) for local data processing.

The working principle of MEMS: from physics to data

The operational flow of a MEMS sensor follows a logical sequence. This process describes how MEMS sensors operate to provide real-time data that supports many applications, from industrial condition monitoring to automotive safety systems.

Step 1: sensing the physical world

A MEMS sensor’s job begins when it interacts with a physical phenomenon. This could be acceleration from motion, a change in atmospheric pressure, an acoustic wave, or a shift in temperature. The sensor's micro-mechanical structure is specifically designed to react to one of these stimuli. For instance, in a MEMS accelerometer, a proof mass suspended by tiny springs deflects in response to movement. In a pressure sensor, a thin silicon diaphragm bulges or flexes when subjected to external pressure. The design of this mechanical element determines the sensor's measurement range and sensitivity. STMicroelectronics builds upon these fundamental principles to deliver high-precision, low-power solutions within its MEMS technology portfolio, integrating advanced mechanical structures with intelligent electronic processing.

Step 2: transduction to convert motion to signal

Once the mechanical structure moves, the transduction mechanism translates this physical change into an electrical signal. This conversion is at the core of the MEMS sensor working principle. Several key transduction methods are used, each with distinct advantages for different applications.

Capacitive sensing

Capacitive sensing is one of the most common transduction methods due to its low power consumption, high sensitivity, and excellent stability over temperature. It relies on measuring the change in capacitance between two or more conductive plates.

• How it works: the sensor contains at least one fixed plate and one movable plate, which is part of the micro-mechanical structure. As the movable plate deflects in response to a physical stimulus, the distance between it and the fixed plate changes.
• Formula: capacitance (C) is inversely proportional to the distance (d) between the plates: C ∝ 1/d.
• Application: a motion MEMS uses this principle. When the device accelerates, the proof mass (movable plate) shifts, changing the capacitance. The application-specific integrated circuit (ASIC) detects this minute change and calculates the corresponding acceleration value. This method is used in our high-performance accelerometers and inertial modules.

Piezoresistive sensing

This method uses materials whose electrical resistance changes when subjected to mechanical stress. Silicon exhibits a strong piezoresistive effect, making this an efficient and highly integrated transduction method for MEMS.

• How it works: piezoresistors are embedded in the micro-mechanical structure at points of high stress, such as the base of a cantilever or the edge of a diaphragm. When the structure deforms, it creates mechanical stress, which in turn alters the resistance of the piezoresistors.
• Measurement: the ASIC measures the change in resistance, typically using a Wheatstone bridge configuration for high accuracy, and correlates it to the applied physical force.
• Application: piezoresistive sensing is prevalent in pressure sensors. As pressure flexes the diaphragm, the embedded piezoresistors are stressed, providing a direct and linear measurement of the pressure change.

Piezoelectric sensing

This principle relies on materials that generate an electrical charge when mechanical stress is applied. Conversely, these materials deform when an electric field is applied, making them suitable for both sensing and actuation.

• How it works: a piezoelectric material, such as lead zirconate titanate (PZT) or aluminum nitride (AlN), is integrated into the sensor mechanical structure. When an external force (e.g., from vibration or an acoustic wave) deforms the material, it produces a proportional electrical charge.
• Measurement: this charge creates a voltage that can be measured by the ASIC.
• Application: MEMS microphones and ultrasonic transducers often use this method. In a microphone, sound waves vibrate a diaphragm coated with a piezoelectric film, generating an electrical signal that represents the audio.

Step 3: signal conditioning and processing

The raw analog signal from the transducer is typically very small and susceptible to noise. The final stage of the MEMS sensor working process involves the application-specific integrated circuit (ASIC), which refines this signal into a robust digital output suitable for a host microcontroller or processor.

The ASIC performs several critical functions:

• Amplification: it increases the amplitude of the weak electrical signal from the transducer to a level that can be accurately processed.
• Filtering: it removes unwanted noise and interference from the signal, ensuring that only the relevant data is processed.
• Analog-to-digital conversion (ADC): it converts the continuous analog signal into a discrete digital format (e.g., a 16-bit or 24-bit value).
• Calibration: the ASIC often stores factory-set calibration coefficients to compensate for minor manufacturing variations and ensure high accuracy across different operating temperatures and conditions.
• Intelligent processing: advanced MEMS and sensors from ST integrate intelligent sensor processing units (ISPUs) and machine learning cores (MLCs). These features enable on-device processing of algorithms, such as activity recognition or anomaly detection, reducing system power consumption and data bandwidth.

This entire sequence, from physical stimulus to digital output, occurs in real-time and enables the instant response required for applications like airbag deployment, industrial predictive maintenance, and smartphone screen rotation.

MEMS sensors enabling advanced technology

By combining sensitive mechanical structures with powerful signal processing on a single chip, MEMS devices offer a compact, low-power, and cost-effective solution for measuring a vast range of physical parameters. The conversion of physical forces into digital data enables engineers to design smarter, more responsive, and more efficient systems. ST’s MEMS technology integrates mechanical precision with intelligent processing to provide the foundation for the next generation of smart sensors.