Today's low-power devices enable the design of circuits, which can operate from the energies harvested from the environment. The solar cell is a popular and a widespread solution, but can not be used in all circumstances. In closed spaces where mechanical energy (from vibration and shock) is readily available, vibration energy harvesters are an ideal choice. Due to its many beneficial properties, the piezoelectric PZT crystal stands out.
Vibrations are not only an energy source, but for some structures (buildings, machines) they provide important diagnostic information. Often the sensor must be installed in places where it is difficult or costly to reach, not to mention the periodic battery replacement. A fully autonomous system with piezoelectric source is still missing from the market. There are some solutions out there, but none of them offers a complex solution for energy harvesting and remote measurement.
In my thesis, I implement a system where the tasks of measuring, processing, wireless data transmission and vibration energy harvesting are integrated on a single printed circuit board, thus providing an easy-to-use and energy-efficient measuring system. I have verified the correctness of my accelerometer with spectrum measurements. I have also shown, through laboratory measurements and a real-life application (in an engine room), that it is possible to harvest energy from a piezoelectric source with my design.
My accelerometer is able to measure vibrations up to 100 Hz without aliasing, however around 200 Hz it is guaranteed to appear. The maximum power I extracted from the piezoelectric beam was 781 uW with resistive load, and 244 uW with complex load (energy management IC). The highest measured efficiency of my energy management chip (BQ25570) was 81%. In order to take a 2000 data acceleration measurement (with RF data transmission) with my system I needed 30 mW average power. If this measurement was to be solely supplied by energy harvesting, I would need around 2-3 hours to harvest that amount of energy. In my engine room application I managed to harvest a total of 2,25 mJ energy in 4 hours and 20 minutes, which resulted in an average harvested power of 144,23 nW.
My system is thus able to conveniently perform remote measurements, however, due to the high energy demand of the RF data transmission, this measurement can not be solely supplied from a piezoelectric energy harvester, therefore it is necessary to install a higher capacity energy storage element. During my work several areas have emerged where the efficiency of the energy harvester and the conditioning circuitry could be improved (vibration modelling, increased clamping stiffness, improved MPPT algorithm and circuit).