Electronic designers are now contemplating whether energy harvesting may be a viable power supply option for their applications. Continuing developments in this area are opening up opportunities for new applications and technical development are not only optimising the energy consumption of microcontrollers and wireless transmission techniques, but also the efficient transformation and storage of low quantities of the energy harvested.
Until recently, energy harvesting solutions could only be found in a few consumer applications, since, in the final analysis, price was a significant factor. These have generally been applications featuring solar cells, such as mobile phones, back up batteries, outdoor sports equipment and pocket calculators. However the situation is different in industrial applications, where the system and maintenance costs are decisive. When it is not easy to replace a battery or where the electronics are inaccessible, energy harvesting electronics can provide a solution. Such applications can be found in building monitoring, the structural monitoring of facilities, the chemical industry and in general factory automation, amongst others. The following types of energy harvesting are available as ambient energy: • Thermal energy: 25µW/cm² from the human body to 1mW/cm² from industrial waste • Photovoltaics: 100µW/cm² under artificial light to 100mW/cm² in sunlight • Vibration/movement: 4µW/cm² from human kinetic energy to 800µW/cm² from mechanical vibration/movement • RF waves over antennas: Less than 1µW/cm², only near field. Whether an energy harvesting solution is technically feasible depends on the regular availability of the energy source. It is frequently the case that a system is designed redundantly, so two or more energy sources are tapped. Some young companies are also developing new harvesting technologies. Micropelt (micropelt.com), for example, produces thermocouples manufactured micromechanically. Selecting the energy storage system In terms of its energy budget, a sunlight application is relatively simple to calculate in advance: the sun rises every day and thus in a way that can be planned for. For this type of application, the principal questions that arise are the following: at what intervals should the electronics be active?; and which energy storage system is appropriate? However, only energy storage systems which allow for a high number of charging cycles are really practical. Rechargeable button cells, for example, only have a useful life of 1000 charging cycles, equivalent to two or three years. The second decision making criterion for the energy storage system is the charging current. In energy harvesting applications, charging currents are only of the order of microAmps or a few milliAmps. Therefore, the battery must also be able to be charged with this low current. At the same time, developers face yet another challenge. Some applications will need 10 to 50mA quickly in short bursts when the electronics is in active operation, especially when wireless transmission is being used. However, it is possible to do away with energy storage systems completely. When using a thermoelectric element for example, it is may only be necessary for the application to be active, provided that the facility to be monitored is in operation and producing residual heat. A relatively new energy storage technology is the so called 'thin film battery' or microenergy cell. This technology features solid bodies, which enables safe and environmentally friendly operation. The materials generally used are lithium cobalt dioxide (LiCoO2) for the cathode and lithium or another metal as the anode, with lithium phosphorous oxynitride (LiPON) as the solid electrolyte. While the amount of lithium used is rather small, the EU does not yet have a separate category for this product group and the same freight and waste management rules apply will apply to thin film batteries as apply to normal Li-ion batteries. Table 1 sets out the different technologies and their characteristics. Selecting the charge controller The selection of a charge controller depends on the battery used and on the voltage needed for the remainder of the electronics. However, one important criterion should always be met: the selected dc/dc converter must have a high level of effectiveness at very low currents. This is where the wheat is separated from the chaff and some semiconductor manufacturers – including Linear, TI and Triune – have recently launched specialised products for this type of application. Discrete boost circuits are also possible, particularly when a capacitor is being used to store energy. When using a Li battery, however, charge monitoring and a cut off system must be implemented. Selecting the microcontroller In energy harvesting applications, the average energy consumption is decisive. The following criteria are therefore important when selecting the right microcontroller: • Power consumption with maximum computing power (µA/MHz). Here peak values of 110µA/MHz for a 32bit Cortex-M0 mcu are currently achieved (NXP). • Power consumption in sleep or deep sleep mode. In this state, the controller may still update data and status registers and require a few nA. • Wake up speed: the controller should be able to switch from the sleep state to active mode in a few microseconds. TI's MSP430 is often specified as the benchmark. • Flexible configuration of peripherals: As the wireless transmission system, the display and other external measuring devices are generally connected to serial interfaces such as spi or i²c (if they are not integrated into the mcu), it is important that these interfaces can, as far as possible, be configured independently of the microcontroller's clock frequency. Under certain circumstances, this saves additional energy, since the mcu's clock pulse can be optimised for the necessary computing task. Selecting the wireless transmission It is a matter of course that the selected wireless transmission technology has to save power. In energy harvesting applications, it is primarily transmission range that is important, with the data transmission rate a secondary consideration. Furthermore, wireless protocols that can organise themselves are of benefit, such as those where individual wireless nodes also have a router function in order to pass data on from other nodes (mesh network). However, designers face the challenge of having to wake up the wireless interface promptly from the low power mode on data transmission. Proprietary ISM band solutions or standard protocols like ZigBee are used. A recent development in this area is the 6LowPAN concept (Internet Protocol v6 via low power wireless personal networks), which works with IEEE802.15.4 wireless chips. Because many energy harvesting applications monitor facilities, it is attractive for operators to retrieve the data through a gateway via IP. For building automation or sports or medical equipment, Bluetooth 4.0 (also known as Bluetooth Low Energy) is an interesting option, since this enables direct mobile phone communication. Summary New battery technologies and high efficiency mcus open up a range of energy harvesting applications. When selecting components, always bear in mind the total energy budget; the solution will only function reliably when more energy is gained than consumed. Author profile: Patrick Delmer is supplier business manager with Arrow Electronics.