Search

In this paper, the design, fabrication, and characterization of a robust MEMS thermal wind sensor utilizing ceramic film packaging are proposed. The ceramic film is fabricated by means of low temperature co-fired ceramic (LTCC) technology, achieving the excellent adhesion between silicon substrate and ceramic film without any thermal adhesive. The most important is that the adverse effects of the packaging on the performance of the thermal wind sensor can be minimized by adjusting the thickness and thermal conductivity of the ceramic film during the preparation of the ceramic slurry. The influences of thickness and thermal conductivity of ceramic film on the performance of sensor have been simulated by finite element method, and the results are in reasonable agreement with theoretical analysis. The proposed sensors were tested and calibrated in a wind tunnel. Experiments results demonstrate that the sensitivity of the proposed sensor is decreased only 4.7%, compared with the sensor without ceramic film. The maximum measurement error of the proposed sensor in wind speed and direction is 0.66 m/s and 5°, respectively. Compared with the sensor packaged with $250\mu \text{m}$ ceramic carrier, the heating power of the proposed sensor is reduced by nearly 50% and the sensitivity is increased by an order of magnitude at the same time.

This paper presents a semi-empirical model for the temperature effect of a micromachined 2-D anemometer. The anemometer consists of a self-heated double Wheatstone bridge. Eight thermistors are symmetrically arranged in two orthogonal directions, and four of them in each direction make up a Wheatstone bridge. The eight thermistors simultaneously act as both heating and sensing elements. The terminal voltage across the bridge and the bridge output voltage, under the constant current supply, are utilized to measure the wind speed and direction, respectively. The semi-empirical model has taken the thermal physical properties of the air, sensor substrate, and thermistors into account. It shows that the measurement of wind direction has a weak temperature dependence due to the symmetric config uration of the self-heated double Wheatstone bridge, which suppresses the temperature drift. The temperature compensation has been performed by means of the semi-empirical model. The experimental measurements demonstrate that, when the air temperature changes from 270 to 310 K, the wind speed varies from 0 to 30 m/s with an error less than ±1.5 m/s while the wind direction spans from 0° to 360° with the weak temperature dependence.

Modeling of the packaged wind sensor operating on constant power (CP) mode is proposed in this paper. The proposed model is based on the theory of heat conduction and forced convection. The presented model indicates that the total output dc voltage is proportional to the square root of the wind speed, and ${V}_{x}$ , as well as ${V}_{y}$ , varies with the wind direction based on cosine function. The wind sensor is fabricated by microelectromechanical systems (MEMS) technology and packaged with the glue. The test is performed in the wind tunnel and the measured results indicate an agreement with the presented model. The relative error of voltage for the wind speed is less than 14% up to 30 m/s, while that for the wind direction is no more than 10.7% from 0° to 360°.

A self-heated double Wheatstone bridge configuration for a 2-D wind sensor is presented. Without a central heating element, eight thermistors are symmetrically arranged in two orthogonal directions, and four of them in each direction make up a Wheatstone bridge. The eight thermistors simultaneously act as the heating sources and temperature sensing elements. By applying the constant current supply to the bridge, the terminal voltage across the bridge and the bridge output voltage are utilized to measure the wind speed and direction, respectively. The 2-D wind sensor has been simulated numerically and fabricated directly on the ceramic substrate. Experiments demonstrate that the measured speed varies from 0 to 40 m/s with an error less than ±2 m/s while the measured direction spans from 0° to 360° with an error less than ±3° after calibration. The double bridge configuration greatly simplifies the measurement and control circuits for the 2-D wind sensor. [2018-0171]

This paper introduces a novel sensing element scheme to improve the accuracy of 2-D micromachined thermal wind sensor. The sensing element consists of eight heaters and eight thermistors placed on ceramic substrate. The octagon-arranged sixteen resistors are divided into two wind sensing groups, cross-type group and saltire-type group. These two groups monitor wind independently, and the final measured wind speed and direction are extracted from the measurement results of the two groups. Based on this octagon-sensing design, the sensor achieves higher accuracy for wind measurement, which is verified by the experiment. The results show that for the wind speed up to 33 m/s, the sensor has a measurement error less than ±1%, and the airflow direction over the full range of 360° is determined with a maximum error of ±1.5°. [2018-0045]

In this paper, the effect of insulation trenches on micromachined silicon thermal wind sensors is explored. The sensor is composed of a central heater and four thermistors. It measures wind speed and direction through detecting the airflow-induced temperature difference on the chip surface. Three different forms of insulation trenches are designed and fabricated to explore their effects on the performance of wind sensors. The experimental results show that when the sensors work in constant voltage mode, the center and arch trenches enhance the performance, while the arch trenches have the most significant effect on the sensitivity improvement. Moreover, all the three kinds of insulation trenches reduce the power consumption for the sensors operated in constant temperature difference mode. The insulation trench design in this paper can provide a valuable reference for the design and fabrication of micromachined silicon thermal wind sensors.

This paper presents the design, fabrication, and performance of a micromachined silicon thermal wind sensor with improved sensitivity. Deep reactive ion etching (DRIE) trenches are fabricated between the heater and the thermistors to suppress the lateral heat conduction in the chip. In addition, eight thermistors symmetrically arranged in four directions around the heater form two Wheatstone full-bridges, resulting in about 50% increase of the sensitivity with respect to four thermistors. Based on these two methods, the sensitivity of the micromachined silicon thermal wind sensor is improved remarkably, which is verified by the experiment. The results show that the measurement wind speed range is up to 33 m/s in constant voltage (CV) mode with the initial heating power of 256 mW. The sensitivity is measured to be 29.37 mV/ms−1 at the wind speed 3.3 m/s, achieving improvement of about 226%, compared with that of the traditional wind sensor. Wind direction measurement results show that airflow direction over the full range of 360° is determined with an accuracy of ±5°. [2017-0052]

In this paper, a novel MEMS thermal wind sensor adopting vertical stacking thermistors is proposed for the first time. Compared with the traditional wind sensors, the novelty of this work is that eight thermistors, fabricated on two layers, can achieve improved sensitivity of wind detection. The sensor is tested in wind tunnel in calorimetric principle and the heater is supplied with constant voltage (initial power 210 mW). Experiments demonstrate that the sensitivity of the proposed senor is double that of the traditional one without increasing the chip size. Measuring range of wind speed from 0 m/s to 33 m/s and full range (360°) of wind direction detection have been achieved.

In this paper, an annular-encapsulation packaging method is proposed to develop high-performance micromachined thermal wind sensor. The sensor used for demonstrating the feasibility of this design is composed of a central heater and eight thermistors. It measures the wind speed and direction through detecting the airflow-induced temperature distribution on the chip surface. Unlike the conventional self-packaging, epoxy resin is only utilized to encapsulate the bonding wires between the sensor pads and the PCB (print circuit board), while the front surface of the sensor is in contact with the air in this annular-encapsulation packaging. Based on this approach, the unwanted heat loss from the sensor to the epoxy resin is minimized, and more efficient heat exchange between the sensor and the airflow is realized via the backside of the chip. The experimental results show that the sensitivity of the sensor with this annular packaging at the wind speed 3.3m/s is 44.29 mv/ms−1, achieving improvement of 50.8%, compared with the sensitivity 29.37 mV/ms−1 of the conventional self-packaged sensor.

This paper presents the sensitivity analysis of the micromachined thermal wind sensor based on back surface sensing mode. The effects of heating power and distance $\pmb{L}$ on the sensitivity are analyzed via simulation and experiment. $\pmb{L}$ is the length between the heater and the thermistor. Experiment has been implemented by applying three different heating power values. On the ceramic substrate, eight thermistors are divided into two groups with different distance related to the heater. The results of simulation and experiment indicate that both the increase of heating power and the distance $\pmb{L}$ can improve the sensitivity of the sensor significantly. In the constant power mode, measuring range of wind speed from 0m/s to 33m/s and full range (360°) of wind direction detection have been achieved.

In this paper, a micromachined silicon thermal wind sensor with improved sensitivity is demonstrated. Deep reactive ion etching (DRIE) trenches are fabricated to suppress the lateral heat conduction between the heater and the thermistors. Moreover, two Wheatstone full-bridges consisting of eight thermistors are designed to increase the output voltage by 50%. Based on these two methods, the sensitivity of the micromachined wind sensor can be improved remarkably, which is verified by the experiments. The results show that the measurement wind speed range can be up to 33m/s in constant voltage (CV) mode with the initial heating power of 256mW. The sensitivity is measured to be 29.37mV/ms−1 at the wind speed of 3.3m/s, achieving an improvement of about 226%, compared with the traditional wind sensor. The results also show that wind direction in a full range of 360° can be determined with an accuracy of ±5°.