The Internet of Things (IoT) represents a vision of a highly interconnected mesh of smart devices exchanging data, without human intervention, about every aspect of each smart device's environment. Powering these devices remains a significant challenge, but one well suited to energy-harvesting solutions. Using available energy transducers and ICs, engineers can create zero-power smart devices able to address the power challenges of an evolving IoT.
The IoT expands the reach of the Internet to individual embedded devices designed to interact with machines, extending the familiar paradigm of a web of human users connected through smartphones, tablets, and computers. Unlike those user systems, embedded devices connected through the IoT must continue to operate with self-contained power and without expectation that a human user will be available to monitor available power, change a battery, or plug the device into a power outlet.
In many cases, IoT devices will be expected to operate for years beyond the ability of even the most advanced battery technology to deliver sufficient operating power. At the same time, many smart sensor applications for these embedded devices require relatively few components (Figure 1) to transmit sensor data wirelessly to other smart devices and upstream servers.
Figure 1: In concept, a typical wireless sensor node in the IoT is a simple device that combines an MCU with subsystems for sensors and wireless connectivity. System power remains a challenge for IoT devices expected to operate unattended for years (Courtesy of Texas Instruments).
With the availability of ultra-low-power components and specialized devices, engineers can scavenge sufficient power from ambient sources to provide system power (Figure 2). These so-called "zero-power" systems are able to acquire and store sufficient energy from the environment to eliminate the need for conventional batteries that eventually require replacement.
Figure 2: For the IoT, an energy-harvesting power subsystem can extract power from ambient sources and meet peak power requirements with long-life energy storage devices, enabling unattended operation for years (Courtesy of Texas Instruments).
Although many IoT applications in industrial and transportation segments may permit solar-powered operation, many others are likely to be literally hidden deeper within the environment being monitored. In most personal-fitness applications, smart sensors are buried in footwear, clothing, and equipment. These and similar applications in consumer, retail, medical, and other segments will depend on their ability to extract sufficient power from vibration, temperature differences, and RF signals from increasingly prevalent nearby wireless energy sources.
Regardless of the type of ambient energy, designers can face a significant challenge in building energy-harvesting power supplies capable of extracting maximum power from sources that can vary their energy output significantly from one moment to the next. Energy transducers such as piezoelectric devices used to convert vibrational energy into a voltage output, deliver maximum energy when operating at the resonant frequency of the vibrational source, and when operating into a load designed to match piezoelectric output impedance. In fact, to ensure maximum energy conversion, designers need to account for the typical operational characteristics for each transducer type, all of which produce maximum power at an ideal maximum-power point (Figure 3).
Figure 3: For maximum power, designers need to account for a transducer's power output curve, which varies significantly for variable-resistance transducers such as solar cells (a) and constant-impedance transducers such as thermoelectric generators and piezoelectric devices (b) (Courtesy of Cymbet).
Further compounding the complexity, the power subsystem needs to provide power management features so as to deliver power effectively to the load while efficiently managing energy-storage device charge and protection requirements. Li-ion-based storage devices typically used in these applications require precise control to avoid undervoltage or overvoltage conditions that can permanently damage the storage device.
In the face of these multiple requirements, an ideal energy-harvesting power supply can be a highly complex system itself when optimized for maximum energy extraction from a complex ambient source such as vibration and built with sophisticated load-responsive power management capabilities (Figure 4). Here, a transducer input stage uses maximum power point tracking (MPPT) techniques to respond to changes in resonant frequency to maintain the transducer at the ideal operating state. On the output stage, power management capabilities maintain energy storage devices while responding to varying power demands. Wireless communications impose a significant power load in these applications. By reducing communications duty cycle to the minimum required, energy-storage devices can deliver sufficient power to serve periodic spikes in power requirements and recharge in time to serve the next peak demand.
Figure 4: An idealized energy-harvesting power supply will ensure that the transducer is maintained at an operating point required to maximize energy output, and that both energy-storage devices and the downstream load are optimally served (Courtesy of the Next-Generation Energy-Harvesting Electronics, Engineering and Physical Sciences Research Council).
For IoT device designers, the availability of specialized devices mitigates the challenges associated with creating a sophisticated energy-harvesting power subsystem capable of delivering much of this capability. Specialized devices such the Maxim Integrated MAX17710 feature ultra-low-power operation required in these applications, along with integrated sophisticated power management features needed for energy-storage and load management. As with some other devices in this class, the MAX17710 integrates a boost converter to support very-low-voltage input sources. The MAX17710 is able to extract energy from ambient sources with output power levels as low as 1 μW, while providing regulated output at selectable levels including 3.3, 2.3, and 1.8 V with overdischarge protection. The device can directly charge an energy storage device, providing overcharge protection and undervoltage lockout to help prevent damage to the energy-storage device.
The Linear Technology LTC3108 offers multiple regulated voltage outputs, including a 2.2 V LDO to power an external microcontroller and a main power output programmable to one of four different fixed-voltage output levels. The device can charge an external energy-storage device, automatically switching the load to the energy-storage device when the input falls below a preset threshold.
Engineers can also find available devices offering more specialized and comprehensive energy-harvesting features. The Cymbet CBC915 combines an integrated MPPT capability capable of tracking the maximum power point for a wide range of variable- and constant-impedance transducers. Designed to work with Cymbet EnerChip™ devices such as the CBC050, the CBC915 provides comprehensive charge management and protection features as well as capabilities to communicate status information including EnerChip state-of-charge to companion processors.
Along with its built-in features for battery and load management, the Texas Instruments BQ25504 also provides a built-in MPPT capability to maintain the energy transducer at its maximum power point (Figure 5). The device samples the input source's open-circuit voltage, programmed using external resistors and an external holding capacitor. The device's boost converter supports input-voltage sources as low as 80 mV. Its energy-storage management features combine storage-device charging and protection features for undervoltage and overvoltage levels.
Figure 5: Designed specifically for energy-harvesting applications, devices such as the Texas Instruments BQ25504 combine features for efficient energy scavenging using MPPT techniques with those for managing both energy-storage devices and the downstream load (Courtesy of Texas Instruments).
Other specialized devices target energy-harvesting applications based on specific transducer types. The Linear Technology LTC3588-1 integrates a full-wave bridge rectifier and buck converter to maximize energy conversion from high-impedance AC sources such as piezoelectric devices. Nevertheless, the device is fully capable of supporting a full range of transducer types. Although not intended to support battery management requirements, the LTC3588-1 can store charge on an external capacitor or be configured to work with a battery backup.
While the IoT promises to connect billions of smart devices, the need to ensure adequate power for the expected years of operation of these devices remains a challenge. Energy-harvesting techniques offer a power solution well suited to the operating conditions expected for many of these smart devices. Using available ICs designed specifically for energy-harvesting applications, engineers can address emerging IoT applications with smart devices able to operate for years on ambient power sources including solar, temperature, vibration, and RF energy.
For more information on the parts discussed in this article, use the links provided to access product pages on the Hotenda website.