R. van Schaijk

VIBRATION ENERGY HARVESTING WITH ALUMINUM NITRIDE-BASED PIEZOELECTRIC DEVICES

This paper describes the measurement results of piezoelectric energy harvesters with aluminum nitride (AlN) as a piezoelectric material. AlN was chosen for its material properties and for its well-known sputter deposition process. For AlN devices a high optimum load resistance is required, which is favorable due to the high resulting voltage level. The output power harvested from mechanical vibrations has been measured on micromachined harvesters with different geometries. The resonance frequencies ranged from 200 up to 1200 Hz. The packaged devices had limited output powers and quality factors due to air damping caused by the package. A maximum output power of 60 µW has been measured on an unpackaged device at an acceleration of 2.0 g and at a resonance frequency of 572 Hz. The package of the harvester requires special attention, since air damping can significantly decrease the maximum power output.

Vacuum-packaged piezoelectric vibration energy harvesters: Damping contributions and autonomy for a wireless sensor system

This paper describes the characterization of thin-film MEMS vibration energy harvesters based on aluminum nitride as piezoelectric material. A record output power of 85 μW is measured. The parasitic-damping and the energy-harvesting performances of unpackaged and packaged devices are investigated. Vacuum and atmospheric pressure levels are considered for the packaged devices. When dealing with packaged devices, it is found that vacuum packaging is essential for maximizing the output power. Therefore, a wafer-scale vacuum package process is developed. The energy harvesters are used to power a small prototype (1 cm\3 volume) of a wireless autonomous sensor system. The average power consumption of the whole system is less than 10 μW, and it is continuously provided by the vibration energy harvester.

Modeling and characterization of MEMS-based piezoelectric harvesting devices

Vibrational piezoelectric harvesting devices (PHD) provide an autonomous power source for various types of sensors, actuators and MEMS devices. There have been several examples of vibrational energy harvesters published in the literature over the years. However, for many applications the generated power is not yet sufficient. In this paper, a physical model for predicting the generated electric power from piezoelectric harvesting devices is introduced. The model is based on estimating the total charge generated on a piezoelectric material when it is subjected to mechanical strain as a result of bending at the fundamental resonance frequency. Based on Euler–Bernoulli beam theory, the strain can be determined in terms of the beam deflection at purely mechanical excitation. The proposed model extends the current state of the art by consideration of the strain distribution due to the presence of an extended mass volume at the end of the beam. The constitutive equations of piezoelectricity in the sensing mode correlate the strain and the induced charge in the piezoelectric element. Using the device design parameters and the beam deflection as inputs, the power output can be calculated. The results of the model were experimentally verified for MEMS-based PHDs. The model was found to give an accurate prediction of the electrical parameters under various damping conditions. After model validation, a subsequent device optimization has been made to improve the power generation.

Shock induced energy harvesting with a MEMS harvester for automotive applications

In this work we report shock induced measurement and simulations on AlN based piezoelectric vibration energy harvesters. We compare the result with sinusoidal input vibrations, where we obtain a record power of 489 µW. The vacuum packaged harvesters have high quality factors and high sensitivity. We validate the potential of piezoelectric vibration harvesters for car tire applications by measurements and simulations.

Optimum power and efficiency of piezoelectric vibration energy harvesters with sinusoidal and random vibrations

Assuming a sinusoidal vibration as input, an inertial piezoelectric harvester designed for maximum efficiency of the electromechanical energy conversion does not always lead to maximum power generation. In this case, what can be gained by optimizing the efficiency of the device? Detailing an answer to this question is the backbone of this paper. It is shown that, while the maximum efficiency operating condition does not always lead to maximum power generation, it corresponds always to maximum power per square unit deflection of the piezoelectric harvester. This understanding allows better optimization of the generated power when the deflection of the device is limited by hard stops. This is illustrated by experimental measurements on vacuum-packaged MEMS harvesters based on AlN as piezoelectric material. The results obtained for a sinusoidal vibration are extended to random vibrations. In this case, we demonstrate that the optimum generated power is directly proportional to the efficiency of the harvester, thus answering the initial question. For both types of studied vibrations, simple closed-form formulas describing the generated power and efficiency in optimum operating conditions are elaborated. These formulas are based on parameters that are easily measured or modeled. Therefore, they are useful performance metrics for existing piezoelectric harvesters.

First autonomous wireless sensor node powered by a vacuum-packaged piezoelectric MEMS energy harvester

This paper describes the experimental characterization of piezoelectric harvesters with different dimensions. We present a record level of generated power of 85 µW obtained from an unpacked device. We have developed a wafer-scale vacuum package which shows a 100–200 fold increase in power output compared with packaged devices under atmospheric pressure. A wireless sensor node was fully powered by a piezoelectric harvester. The average power consumption was less than 10 µW while it was operating at 15 seconds duty cycle.

Harvesting energy from airflow with a michromachined piezoelectric harvester inside a Helmholtz resonator

In this paper we report an airflow energy harvester that combines a piezoelectric energy harvester with a Helmholtz resonator. The resonator converts airflow energy to air oscillations which in turn are converted into electrical energy by a piezoelectric harvester. Two Helmholtz resonators with adjustable resonance frequencies have been designed - one with a solid bottom and one with membrane on the bottom. The resonance frequencies of the resonators were matched to the complementing piezoelectric harvesters during harvesting. The aim of the presented work is a feasibility study on using packaged piezoelectric energy harvesters with Helmholtz resonators for airflow energy harvesting. The maximum energy we were able to obtain was 42.2 νW at 20 m s\-1.

Piezoelectric AlN energy harvesters for wireless autonomous transducer solutions

Piezoelectric energy harvesters with aluminum nitride (AlN) are fabricated by micromachining. Devices with different size and mass were fabricated. The devices were packaged by a top and bottom glass wafer with cavities of 200mum to prevent excessive displacements. The devices are characterized on a shaker by applying an oscillation as mechanical input and measuring the power dissipated in a resistive load. A power output was generated of ~10muW for an acceleration of 8g at the resonance frequency of 1155Hz. The power generation for AlN is comparable with lead zirconate-titanate (PZT), despite the much lower piezoelectric constants. This makes AlN a very suitable material for energy harvesters due to its easier processing.

A self-biased 5-to-60V input voltage and 25-to-1600µW integrated DC-DC buck converter with fully analog MPPT algorithm reaching up to 88% end-to-end efficiency

Energy harvesting is seen as an enabling technology for autonomous wireless sensing in automotive applications. This technology may rely on piezoelectric or electrostatic energy conversion using the energy available during the tire impact with the road. The power management system has to transfer harvested power to the load battery (few μW to mW) with the highest possible efficiency. An electrostatic harvester can generate voltages up to 40V even at low accelerations [1]. Hence, a high voltage (HV) DC-DC converter is needed to maximize the energy transfer from the harvester to the load battery. However, HV buck converters including maximum power point tracking (MPPT) algorithms have been shown only for much higher ranges of power, in the order of W [2]. On the other hand, all reported converters working in the sub-mW power range are not able to sustain high input voltages and do not include an integrated MPPT algorithm [3, 4]. Until this work, no solution in literature is able to interface with these electrostatic energy harvesters. Figure 4.6.1 shows the block diagram of the system, designed and fabricated in TSMC 0.25μm BCD CMOS (60V option). The IC is able to interface a vibrational harvester by means of few external components: a rectifier, its load capacitor CIN, and an inductor L. The converter is consisting of a power train and its control circuits. The power train is implemented with the external inductor L, and the integrated power switches MP and MN. The integrated control circuits have to provide the correct gate voltages VGN and VGP for allowing the converter to maximize its output power. All circuits are biased by means of an integrated current reference and are supplied by the load battery voltage VBAT and the DC input VIN.

Patterning of inorganic electrets

A yearlong experiment has been conducted to prove that the field emission and atmospheric ions are responsible for discharging of an electret at its edge. Based on the obtained results, a technology for patterning electret layers has been developed for technical applications. It allows fabrication of electret film structures with a feature size of 20 μm and probably less. Patterned inorganic electret lines fabricated using photolithography show no dependence of charge stability on line width based on monitoring the charge retention during one year. A general approach to electret patterning is discussed.