We present a one-dimensional (1D) theoretical model for the design analysis of a micro thermal convective accelerometer (MTCA). Systematical design analysis was conducted on the sensor performance covering the sensor output, sensitivity, and power consumption. The sensor output was further normalized as a function of normalized input acceleration in terms of Rayleigh number Ra (the product of Grashof number Gr and Prandtl number Pr) for different fluids. A critical Rayleigh number (Rac = 3,000) is founded, for the first time, to determine the boundary between the linear and nonlinear response regime of MTCA. Based on the proposed 1D model, key parameters, including the location of the detectors, sensor length, thin film thickness, cavity height, heater temperature, and fluid types, were optimized to improve sensor performance. Accordingly, a CMOS compatible MTCA was designed and fabricated based on the theoretical analysis, which showed a high sensitivity of 1,289 mV/g. Therefore, this efficient 1D model, one million times faster than CFD simulation, can be a promising tool for the system-level CMOS MEMS design.
29 December 2020 in Journal of Microelectromechanical Systems
A two-dimensional (2D) model was developed, for the first time, to describe the characteristics of the micromachined thermal expansion-based angular motion (TEAM) sensor, which has been validated by the experimental results. Scaling analysis on the performance characteristics through the 2D model was conducted to optimize the TEAM sensor’s design in terms of the normalized distances between the microheaters and temperature detectors in $x$ and $y$ directions, the thickness of the thin film, the heater width, the cavity depth, and the heater temperature. Furthermore, the proposed 2D model was normalized by two dimensionless numbers, namely Rayleigh number Ra and Peclet number Pe, with a critical Rayleigh number (Ra $^{\ast }_{\mathrm {c}} =18$ ,000) identified to differentiate linear and nonlinear operation regimes of the TEAM sensor. In particular, our 2D model is much faster than the conventional CFD model by three orders of magnitude (18.91s versus 5.5h), enabling rapid system-level optimization of the critical design parameters. Accordingly, the TEAM sensors with three pairs of platinum thermoresistive temperature sensors were designed and fabricated. The fabricated device demonstrated a normalized sensitivity of $11.8~\mu \text{V}$ /°/s/mW based on the working fluid of air, which was more than three times better than previous thermal angular motion sensors. Thus, with the experimental validation, the proposed 2D model should be a reliable tool to realize the systematical design optimization of TEAM sensors integrated with on-chip microelectronics for future industrial IoT applications. [2020-0355]
We report, for the first time, a novel two-dimensional (2D) model for the micro thermal expansion-based gyroscope (µTEG) to predict the sensor’s performance, which has been validated by the experimental results. Scaling analysis on the sensor’s performance characteristics by this model enables the optimization of µTEG design, including the normalized distances between the heater and temperature detectors in two directions, the thin film thickness, the heater width, the cavity depth and the heater temperature, to achieve extremely high sensitivity (11.78 mV/°/s) and low power consumption (12.8mW). According to the analysis by 2D model, the sensitivity of the optimized µTEGs by using the working gases (SF6 and C4F8) with larger density, better than the best published µTEG (1.287 mV/°/s) by one order of magnitude, can reach the level of the commercial product (>6 mV/°/s). In particular, our new 2D model can significantly save the CPU time in comparison with the conventional CFD model (1.92s versus 5h) to realize the efficient systematical optimization of the key design parameters. Thus, the proposed 2D model can be a useful tool for µTEGs’ system-level designs for industrial IoT applications.
12 November 2024 in IEEE Transactions on Industrial Electronics
In this article, we propose a micro flow sensor with enhanced sensitivity, resolution, and low temperature drift, utilizing a new strategy of thermal feedback (TF) and digital power distribution (PD) across two microheaters. The optimization of the TF-based flow sensor and its temperature compensation capability were validated using computational fluid dynamics model. The flow sensor was then fabricated using a platinum-sensitive material through an in-house developed MEMS process. Two sensors with different distances (Dsh) between the microheater and microsensor were tested in a thermostatic chamber with nitrogen gas flow ranging from −9 to 9 m/s. Sensor 2, with a Dsh of 34 μm, achieved the highest sensitivity of 27.71%/(m/s), which is 2.6× higher than our prior work. Additionally, the TF-based flow sensor exhibited an inherent temperature drift of less than 4% F.S. over an ambient temperature range of 0 to 50 °C, even without compensation. Furthermore, our sensor system demonstrated superior long-term stability, with a zero offset of less than 0.04% F.S. within 10 min, indicating an intrinsic resolution better than 1.44 mm/s. Therefore, this new strategy that combines digitized signals with thermal feedback proves beneficial for designing highly robust and low-temperature drift flow sensors.
This work introduces a novel tunable piezoelectric MEMS microphone, featuring a fully clamped membrane and dual-sensing electrodes. The device achieves a baseline sensitivity of -39.31 dB, with tunability enabled by a DC tuning mechanism. The key innovations of this subject include: 1) a discrete electrode design enabling simultaneous tuning of sensitivity and resonant frequency while preserving acoustic sensing functionality; 2)utilizing the reverse piezoelectric effect to achieve large tunable ranges with minimal tuning voltages; and 3) a cost-effective performance tuning methodology that eliminates the need for structural modifications; 4) ascertain the tuning mechanism. Furthermore, a refined equivalent circuit model provides insights into the electromechanical behavior of the device, enabling the optimization of tunable microphones and acoustic transducers. Experimental results demonstrate the tuning ability successfully. The acoustic experiment shows that the output amplitude can be changed by up to 182.18% under ±0.5 V DC tuning under 1 kHz acoustic input. The electrical experiment reveals a maximum resonant frequency change of 6.85% with ±10 V DC. The proposed microphone is a promising candidate to be employed in many next-generation audio applications, such as adaptive voice systems, AI-driven speech recognition, and noise cancellation.[2025-0005]