English 
Chinese As MEMS microphones evolve to meet the growing needs of next-generation audio applications, the demand for higher signal-to-noise ratio (SNR), extended bandwidth, and enhanced sensitivity is sharper than ever. Addressing these demands, this study explores the development and implementation of a two-poly differential MEMS microphone utilizing the MOSBE (Micromachining of Silicon-Based Elements) process.
Differential MEMS Microphone Architecture
At the heart of this advancement lies a novel differential MEMS design, featuring:
- Top/bottom diaphragms and top/bottom backplates, fabricated using two polysilicon layers, enabling true differential capacitance readout and enhanced noise immunity.
- A symmetric mechanical configuration, which naturally cancels common-mode noise and improves measurement linearity.
This architecture allows for parallel-plate gap-closing capacitive sensing, which is highly sensitive to minute diaphragm deflections caused by acoustic pressure—an essential requirement for applications such as far-field voice capture, wearables, and assistive hearing devices.
Key Design Innovations
- Mesa-Isolated Diaphragm with No Release Holes
Traditional MEMS microphones often suffer from low-frequency acoustic loss due to release holes that allow air leakage. This implementation eliminates such holes and instead employs a mesa-structured diaphragm, minimizing acoustic damping and boosting performance in the critical low-frequency range (50 Hz–1 kHz). - U-Shaped Spring Suspension for Compliance and Rigidity
The diaphragm is supported by flexible U-shaped silicon springs, combining structural rigidity with localized flexibility. This hybrid approach increases effective sensing area while maintaining mechanical robustness, crucial for harsh environmental conditions. - Single Sacrificial Layer Process
Manufacturing consistency is greatly improved with a single-layer sacrificial release process, minimizing fabrication-induced variation in sensing gap thickness—a critical factor for yield and capacitance sensitivity uniformity.
Performance Benchmarking
The prototype microphone, based on an 800 μm diameter diaphragm, achieved the following measured results:
- 🎧 Differential sensitivity: −40.5 dB @ 1 kHz (Ref: 1V/Pa)
- 🔊 Bandwidth: ±3 dB from 50 Hz to 20 kHz
- 📈 SNR: 57.8 dB (comparable to commercial-grade MEMS microphones)
- 🧩 Compact form factor: Ideal for integration into multi-mic arrays
Applications & Integration Potential
Thanks to its extended bandwidth and improved SNR, this microphone design is well-suited for:
- Far-field voice recognition systems
- Smart speaker arrays and conference microphones
- Automotive voice command modules
- Medical auscultation or bioacoustic monitoring
For applications requiring high sensitivity MEMS microphones with robust MEMS architecture, SiSTC offers a range of optimized products. Learn more here:
🔗 MEMS Microphone Product Line – SiSTC
Why It Matters
Differential MEMS microphone designs represent the future of acoustic sensing—combining the low power and miniaturization of MEMS with the fidelity and range of professional-grade microphones. By eliminating release holes and introducing symmetrical dual-poly architecture, this microphone is a key step forward in bridging the gap between miniaturization and performance.
