11-04-2011 12:58 PM - edited 11-07-2011 10:54 AM
Continued from November 2011 e-newsletter ...
A hot topic in the design community lately is the use of energy harvesting to create self-powered devices that don’t need line power, and never need batteries. In particular, for wireless sensor and control applications, the promise of completely wire-free, and forever maintenance free, sensor nodes has created an intense interest in new materials that can harness otherwise wasted energy as a power source to supplant traditional primary batteries.
Energy harvesting isn’t really new – crank powered radios, motion powered watches and wind and solar have been with us for decades. But the convergence of new thin-film materials offering higher power densities, with the ever-shrinking power requirements of microcontrollers, is bringing the concept out of the lab into early adopter, real-world applications. Ambient energy sources can include vibration, heat, light, mechanical motion (meaning other than vibration) and even RF. Two particular areas of heightened interest are the use of piezoelectrics for vibration energy harvesting, and thin-film thermoelectric generators for harvesting waste heat.
The piezoelectric effect creates a voltage across certain materials when they’re subjected to mechanical strain (the effect is also reversible). Natural quartz crystals exhibit this effect, but energy harvester transducers are generally made from proprietary ceramics. For collecting vibration energy the most common configuration uses the piezoelectric material in a cantilevered beam. With one end fixed to the vibrating source and the other end free to oscillate, the transducer will output an AC voltage proportional to its mechanical deflection. A tip mass is usually added to the raw transducer to both amplify the deflection and tune the natural mechanical resonance of the beam to the frequency of the vibration being harvested. Midé Technologies, from Medford, MA offers ceramic piezoelectric beams in various geometries that are tunable to vibrations ranging from 20Hz up to 400Hz. The output voltage can be fairly high – up to 100V or more – requiring a switching buck regulator to isolate and condition the beam’s output for the load. Power output will depend on the amount and geometry of the piezo material, and the magnitude of the deflection. Continuous power output of up to several milliwatts is possible.
Thermoelectric generators (TEGs) exploit the Seebeck effect, which describes the migration of charge carriers through dissimilar materials such as a thin-film P-N couple, when subjected to a thermal gradient. With a matrix of P-N couples configured so that they’re in series electrically and parallel to the heat flux, the TEG will generate power proportional to the square of the temperature differential. In this type of arrangement it’s possible for a TEG to have power density of several watts/cm2. As an example, the tiny HV56 TEG from Nextreme Thermal Solutions (only 3.1 x 3.3 x 0.6mm) is capable of sourcing 35mW of output power with a mid-range delta-T of 50 degC. One of the design challenges for thermal energy harvesting is properly designing a “thermal circuit” to maximize the temperature delta across the TEG. To harvest waste heat for example, low thermal resistance to ambient is necessary to effectively reject heat on the cold side, while the hot side contact plate has to be large enough to maximize the heat flux through the TEG.
Photovoltaic (PV) energy harvesting is still the dominant technique of choice, for both outdoor and indoor environments. This is largely because PV converter design is well understood and PV materials are still significantly less expensive than other types of transducer materials. As a general rule of thumb, the higher the efficiency of the material, the higher its cost. Today there are dozens of commercially viable PV materials with efficiencies ranging from 5% up to the mid-20s. But for a fixed power output, lower efficiency also means more material is required, resulting in a physically larger array. So the cost vs. efficiency trade-off is an important consideration. Certain materials may have other advantages as well. In the case of amorphous thin film silicon for example, the material is flexible and can be folded around irregular shapes whereas single and multi-crystal silicon are made on rigid substrates like glass. For energy harvesting wireless sensors, a small 1.5 x 4.5 inch flexible thin-film array such as the MP3-37 from Powerfilm is capable of providing 150mW in full sun or several mW in a normally lit indoor office environment.
With several milliwatts of continuous output power from these various types of transducers that are in production today, scavenging energy from the environment is certainly a viable technique for powering low power sensor nodes. Deployments thus far have been limited to early adopters in the petrochemical, aerospace and heavy transportation industries. This might be because of the cost of the materials, or perhaps these are the most beneficial applications for the technology. Certainly some of the transducers still tend to be expensive, so they’re not always suited for every application. As the technology evolves, we may see costs driven down by new devices such as self-charging MEMS-based vibration harvesters or low cost PV materials printed on fabric with special inks. We at Avnet are intrigued by the use of ambient energy harvesting in other applications being developed today, and we’d love to hear about yours.