Energy harvesting devices are making steady improvements in performance and their manufacturers are seeking greater market acceptance of their devices. They’re being encouraged by major semiconductor IC manufacturers who are joining forces with them by offering supporting products like sensors, microcontrollers, power supplies, DSPs and solid state batteries, as well as development kits and application notes. These IC manufacturers realize there’s a large market potential for these efforts as ICs downscale in size and are called upon to operate at lower and lower power levels.
The market for energy harvesters in 2011 reached $700 million, with the majority market value going into consumer electronics according to IDTechEx, a market research and consulting services firm. It forecasts this market to grow to over $1.4 billion by 2017, with two of the leading growth segments being the wireless sensor network and military/aerospace segments (Fig. 1).
Shrinking semiconductor IC chip line geometries and lower power consumption levels come at a time when energy harvesting devices are becoming more effective and practical (Fig. 2). For example, thermoelectric energy harvesting is now possible with only a few degrees of temperature difference.
This was demonstrated by Schneider Electric for wireless communications applications using the ZigBee communications protocol. The company used a thin-film thermoelectric generator where only 3ºC of temperature difference generated 126 µW (at 210 mV), and a standard dc-dc converter with about 70% efficiency (100 µW at 2.4 V). In a star architecture, measurement and data transmission could be achieved every 5 susing ZigBee.
We don’t have to look very far when it comes to lower-power future IC chips. Intel Corp., for example, recently demonstrated an experimental IA microprocessor core capable of unprecedented low-power operation, the product of many years of research. Codenamed Claremont, it can operate at a near threshold voltage (NTV) so low that it can be powered by a small solar cell (Fig. 3). Intel’s Chief Technology Officer Justin Rattner showed this last September at the Intel Developer Forum as an example of how to develop NTV computing and to demonstrate the energy benefits of NTV designs, which promise better energy efficiency.
The Claremont, designed for high-performance computing, is a heat-sink-free processor core that can be placed in the NTV mode that dissipates less than 10 mW with minimum energy and provides five times better energy efficiency than what is possible today. It also provides a wide dynamic operational range and can run at higher frequencies (about 10 times higher) when performance is needed. “It might not be a commercial product, but the research could be integrated into future processors and other circuitry,” says Rattner.
A common form of energy harvesting comes from the piezoelectric effect. Materials like certain ceramics and crystals can generate electricity in response to an applied strain, a behavior that is reversible wherein an applied electric field can produce mechanical motion of the material.
One notable player in piezoelectric-based products is MicroStrain Inc., a leading provider of inertial measurement systems, displacement transducers, and sensing networks used in health monitoring of civil structures and military/aerospace aircraft. The company’s patented temperature-compensated differential variable-reluctance transducers provide extremely small size and high-accuracy attributes and the ability to withstand high-temperature gradients, saline solutions, and pressurized environments.
France-based Arveni s.a.s. has already demonstrated the first infra-red (IR) Philips TV remote control that requires no batteries (Fig. 4). The remote control, developed for Franceís IP TV provider SFR, contains a standard pulse microgenerator consisting of a sensor, microprocessor and an RF output stage. It accepts electrical inputs from the click of the remote’s buttons that produce a piezoelectric output. Arveni’s AR01 microgenerator produces 2.1 mJ of energy (31.6 V) in response to 3.4 N of force.
AdaptivEnergy is another company that capitalizes on the piezoelectric effect. Their Joule-Thief dc power device harnesses energy from vibrations and impact forces, and converts it to usable electrical power.
The technology of piezoelectric energy harvesting is improving. At last November’s Energy Harvesting 2011 meeting in Boston, MA, researchers from the National Institute of Aeronautics (NIA) showed how piezoelectric devices could be made to produce high-energy efficiencies and output levels. They demonstrated that piezoelectric crystals in the <33> longitudinal mode can be made to produce three times higher energy harvesting efficiency than those in the <31> transversal mode, using a hybrid piezoelectric energy harvesting transducer, and can harvest four times more energy.
And researchers at Belgium’s IMEC have produced a MEMS chip that can harvest energy from vibrations inside a car to power tire-pressure monitoring systems (TPMSs) without the need for battery power. The chip’s maximum output was just under 500 µW at its natural frequency of 1 kHz. Energy generation dropped to a little over 40 µW when a car’s tire was traveling at 40 miles/hour, which is still enough to qualify the part for TPMS and wireless communications circuitry.
The chip consists of cantilever beam with an aluminum-nitride piezoelectric layer sandwiched between metallic electrodes that form a capacitor. A mass attached to one end of the cantilever beam enables to act as a transducer, converting vibrations into electricity as the piezoelectric layer flexes. The voltage across the capacitor is harvested to drive wireless transmissions.
One of the largest suppliers of energy harvesting product from vibration is United Kingdom’s Perpetuum Ltd. It makes use of electromagnetic vibration harvested energy to power wireless sensor nodes in the monitoring of railroad wheel bearings, wheel conditions, derailments, hazardous cargo, braking systems and GPS stationary locations.
Perpetuum’s PMG FSH tranducers work on Faraday’s law principle. A magnet vibrates up and down relative to a coil, both of which are housed in a case with springs (Fig. 5). The transducer produces an ac current that is rectified to dc. A PMG FSH delivers at least 0.5 mA at 3.6 V, indefinitely. As such, it replaces a typical and more costly battery (in the long term) and provides “fit and forget it” life-cycle independence.
Thermal energy is another good source for harvesting electrical power. Germany’s Micropelt GmbH’s TGP651/751 thermoelectric generators (TEGs) produce electricity from temperature differences of 5ºC or more and output power ranging from 100 µW to over 10 mW. This is sufficient to offset most batteries in wireless sensor networks.
The TEG modules use an aluminum oxide substrate on which multiple thermocouples are laid thermally in parallel and electrically in series. Up to 128 thermocouples can be laid on a 1600-mm2 substrate to produce about 50 mV/K.
Another German manufacturer, EnOcean GmbH, offers the Dolphin platform of energy harvesting modules that can create electricity from thermal as well as solar and motion sources. Modules like the STM300 868-MHz wireless sensor transceiver feature the industry’s lowest sleep-mode current of just 200 nA and consume one-tenth the power of conventional modules.
Kits, Development Tools
Evergen thermoelectric kits from Marlowe offer a range of thermal energy harvesting for evaluation and testing purposes (see Power Electronics Technology, Feb. 2012, p. 24). The kits allow solid-to-air, liquid-to-liquid and liquid-to-air thermal energy harvesting evaluation and testing for powering sensors, actuators, valve solenoids and other small devices. They operate from temperature-difference outputs of 0.3 and 0.5 mW that can be converted to outputs of 2.3 and 5.1 W.
Energy harvesting kits are available with all sorts of devices that include not the only the energy harvesting source, but also solid-state storage devices (to store the energy charge) as well as electronic ICs like microprocessor development tools. For example, Cymbet Corp. has teamed up with Texas Instruments (TI) to offer Cymbet’s EnerChip solid-state battery with TI’s MSP430 Value Line LaunchPad development kit.
TI also offers the TPS62120 75-mA dc-dc step-down converter for energy harvesting and low-power applications. It supports inputs of 2 to 15 V and consumes a mere 11 µA of quiescent current. The part is 96% efficient and can operate from 9-V and 12-V dc or a battery.
Others also offer components for energy harvesting. For example, Advanced Linear Devices recently introduced the EH4200 family of micropower step-up low-voltage booster modules that boost the output of some TEGs, electro-magnetic coils, and single photovoltaic and infrared emitters for effective energy harvesting. And, Linear Technology Corp. offers the LTC3105 and LTC3108 step-up dc-dc converters for boosting purposes as well.
Infinite Power Solutions has teamed up with Maxim Integrated Products to offer a development kit that contains Infinite Power Solutions’ ThinEnergy MEC101 solid-state rechargeable thin-film battery and Maxim’s 17710 power-management IC to develop and evaluate self-sustaining “green” power solutions. The IPS-EVAL-EH-01 kit efficiently accepts charge less than 1µA.
Wireless Sensor Networks
A large number of energy harvesting applications involves their use in wireless sensing applications. Many of these applications serve consumer electronics, building and industrial automation, automotive uses, military/aerospace, and medical body area networks for monitoring.
According to the aforementioned IDTechEx forecast for energy harvesting, approximately 1.6 million energy harvesting devices were used in wireless sensing in 2011, resulting in $13.75 million being spent on this market segment. The study forecasts that the market for wireless sensors in industrial automation will reach $140 million, and $210 million for military/aerospace applications by 2017.
Many companies are offering starter kits and components specifically designed for wireless sensor networks. EnOcean, for example, is offering its ESK 300 starter kit, which consists of a switch module for building services, components for different switch applications, a temperature sensor module, a USB gateway, PC software for visualization, and a sample case for industrial switching applications.
Powercast Corp. is offering a licensable RF power chip-set and reference design for embedded low-power wireless charging (Fig. 6). These enable OEMs to directly embed the same functionality provide by Powercast’s P1110 or P2110 Powerharvester receivers.
TI offers the BQ25505 IC, a booster charger that can be used in wireless sensor networks. It maximizes the energy harvested from solar and TEG sources with an industry low quiescent current of just 330 nA.
Dust Networks Inc. has demonstrated a self-powered IPV6 intelligent wireless sensor network using MicroPelt’s TE-Power TEG as well as Cymbetís EnerChip solid-state battery rechargeable solid-state battery. The SmartMesh IP 6LoWPAN network runs entirely on harvested energy and does not require conventional batteries.
An important development has emerged from IMEC’s Host Centre in the form of a record low-power 2.3/2.4-GHz transmitter for wireless sensor applications compliant with four wireless standards: the IEEE802.15/15.6/4/4g and Bluetooth Low Energy. Fabricated on a 90-nm CMOS process, the transceiver consumes only 5.4 mW from a 1.2-V supply (2.7 nJ of energy) at a 0 dBm output. This is said to be three to five times more efficient than current state-of-the-art Bluetooth Low Energy solutions.
One of the impediments to wider-scale adoption of wireless sensor networks is the fact that there is no universally available source of energy available to harvest at all times. Solar energy harvesting is not possible in dark areas (near machines, in warehouses, etc.), and thermal gradients and vibrations cannot be harvested in situations where there are no such sources of energy.
These observations have been pointed out by researchers at Franceís CEA-Leti in an article entitled “Energy Harvesting, Wireless Sensor Networks & Opportunities for Industrial Applications” (http://www.eetimes.com/design/smart-energy-design/4237022/Energy-harvesting--wireless-sensor-networks---opportunities-for-industrial-applications). They emphasize that because different power densities characterize different energy sources, the source of energy to be harvested for wireless sensor networks must be carefully chosen according to the local environment (Fig. 7).
The idea that present-day renewable energy sources such as solar, wind, geothermal and hydropower cannot be easily adopted due to their intermittency and storage difficulties is shared by others like Professor Jeongmin Ahn at Syracuse University. He also points out that despite significant advances in battery technologies, their power densities cannot meet the demands for powering future wireless sensor networks. He believes that combustion devices are the answer to powering wireless networks where electrical energy is created from chemical energy available in various hydrocarbon fuels.
Ahn believes the such devices are needed due to limitations of present-day power generation systems that require moving parts and need parasitic power. Such devices would impose no parasitic electrical power requirements; use fuel, not electrical power, as the energy feedback for pumping; produce electrical power with no moving parts; and do not require high-precision fabrication. Ahn is trying to develop power solutions that meet these stringent requirements.