Simultaneously detecting multiple signals with high precision has long challenged microelectromechanical systems (MEMS) sensors due to unavoidable interference. A new study presents a solution: a constant-drive method using weakly coupled resonators. This technique preserves the high sensitivity of traditional mode-localized sensors while sharply reducing signal crosstalk. The method was rigorously validated through theoretical modeling, finite element simulations, and real-world experiments. Remarkably, it reduces cross-sensitivity by more than tenfold without compromising detection accuracy. This advancement opens the door for smarter, more reliable MEMS devices capable of decoding complex signal environments with clarity and precision.
Microelectromechanical systems (MEMS) sensors have become indispensable in a wide array of technologies—from autonomous vehicles and wearable medical monitors to industrial automation systems. These miniature devices offer advantages in size, power consumption, and multi-functionality. However, as applications grow more complex, so does the need for sensors that can simultaneously detect multiple physical signals. Current mode-localized sensors, while sensitive, struggle in this regard due to high cross-sensitivity between signals. When two signals overlap, they interfere with each other, degrading performance and making accurate interpretation nearly impossible. Due to these challenges, innovative approaches are urgently needed to enable high-fidelity, multi-signal detection.
On May 9, 2025, researchers at Northwestern Polytechnical University in Xi'an, China, published a study (DOI: 10.1038/s41378-025-00954-y) in Microsystems & Nanoengineering introducing a novel constant-drive technique for weakly coupled resonators. Designed to enable synchronous detection of dual signals with minimal interference, the approach addresses a fundamental limitation in MEMS sensing technology. By keeping the driving frequency stable, the method breaks free from the entanglement between multiple input signals, maintaining both accuracy and sensitivity. The research marks a leap forward in multi-signal sensing solutions for next-generation MEMS devices.
The study focuses on a tri-resonator MEMS structure—essentially three weakly coupled mechanical elements arranged to detect input signals via shifts in stiffness. In conventional mode-localized sensors, any input shifts the resonant frequency of the system, causing overlap and signal confusion. The constant-drive method counters this by fixing the drive frequency, which prevents one signal from altering the behavior of another. Theoretical analysis under ideal (non-damped) conditions demonstrated complete isolation between the signals. Finite element modeling under real-world damping effects showed cross-sensitivity dropping to just 0.054%. Experiments confirmed the effect, showing a dramatic improvement—cross-sensitivity reduced from over 26% to as low as 1.1%. Notably, this was achieved without any loss in signal sensitivity. Instead of relying on energy redistribution, as in mode-localized methods, this approach monitors energy dissipation—providing a more stable, scalable solution for dual-signal sensing.
"What we've demonstrated is more than a tweak—it's a fundamental shift in how MEMS sensors can process multiple signals," explained Prof. Honglong Chang, senior author of the study. "By eliminating the need to chase resonant frequency shifts, we reduce both complexity and energy consumption. It's a cleaner, more reliable way to measure real-world phenomena that often present in tandem—like vibrations and forces or electric and magnetic fields."
This breakthrough opens exciting opportunities for multi-signal MEMS devices in fields ranging from aerospace to biomedical diagnostics. The constant-drive technique is not limited to accelerometers; it can be extended to electric field sensors, force detectors, or any resonator-based measurement system. Its simplicity could also lower the energy cost and technical barriers for MEMS integration into mobile, wearable, or remote sensing applications. As sensor demands escalate in smart environments, the constant-drive method offers a robust foundation for building versatile, high-performance sensing platforms that meet tomorrow's technological needs.
