Give away medical masks when you place an order. learn more

Energy Harvesting for Wireless Sensor Networks

Wireless sensor networks require careful attention to power management if they are to meet their objectives in the field. Integrating energy harvesting techniques into your design can go a long way toward addressing the problem.

The increasing accessibility and performance of power miserly sensors, microcontrollers and RF transceivers is raising the potential for wireless sensor networks powered exclusively or supplemented by energy harvesting techniques. Ultra-low power wireless protocols are beginning to achieve widespread industry acceptance and standards are in active development. Sensor networks unshackled from the mains, or battery power, open the possibility for greater reliability, lower maintenance costs, improved safety, and widespread deployment.

Applications unthinkable only a few years ago are now possible with energy harvesting techniques. Newly available power management products can convert the inconvenient, intermittent and often miniscule outputs of various energy harvesting transducers (Thermo-Electric Generators, photovoltaics, piezos, electromagnetics) into usable power for modern electronics. A new way of specifying, analyzing, and designing with these power management devices is necessary to fully exploit the capabilities of the respective energy transducer elements and the sensor networks electronics that are ultimately powered by them.

Wireless sensors are not new, making them semi — or fully autonomous through the use of energy harvesting techniques requires the proper selection and design of energy transducers and power management devices. A typical wireless remote sensor node is shown in Figure 1. To date, the missing link in this system has been the power management solution. The transducers available to provide power are often very inconvenient to work with — producing a very low voltage, low impedance output or a very high voltage, high impedance output. The various elements in this system can be further broken down into power producers/regulators (transducer and power management) and power users (everything else). If the energy harvesting average output power capability exceeds the average power required by the remote sensor electronics, then you have the possibility for an autonomous system.

Figure 1: Typical Wireless Sensor System.

Before any design is initiated, it is worthwhile to run a quick feasibility analysis. This will quickly determine whether energy harvesting techniques are practical. The first step is to decide how often measurements need to be made and transmitted. We will call this measurement frequency (F). Next, we can determine how much processing power is required for the sensor, signal conditioning, data conversion, and processing to generate the desired data, plus the RF transceiver power and time required to transmit this data.

Table 1 shows the typical power requirements for a popular microcontroller and RF link system. The power requirements can vary from manufacturer to manufacturer and for the particular application. There are many choices that can be optimized based on the end application. From this, we can calculate the system duty cycle and average power. The duty cycle (D) of the system is defined as: (Measurement time (Tm)+Processing time (Tp)+Transmit time (Tt)) x Measurement Frequency (F). The average power (Pa) is simply the total power (P) x D + the standby power, which is generally small enough to be ignored.

  Processing Current/Sleep Current
Processor Power 3 mA/500 nA
RF Link 20-30 mA for 1-10 ms

Table 1: Typical Power Requirements for Microcontroller and RF Link

Energy Harvesting Source Typical Power Range K
Solar (indoor/outdoor) single cell 10 µW-40 mW/cm2 0.6-0.8
Vibration (piezo) 4 µW-100 µW/cm2 0.8-0.9
Thermal (TEGs) 25 µW-10 mW/cm2 0.3-0.5

Table 2: Typical Energy Sources and their Power Capabilities

For example, let's assume we are tasked with designing an autonomous indoor temperature sensor. This sensor will be deployed throughout a large office building and coupled with proximity sensors that can detect when a room is occupied and adjust the temperature accordingly. Deploying this type of sensor within a large building can reduce the annual heating and cooling costs significantly. The sensors require 500 µA at 3.3 V for 2 ms to measure temperature and detect an occupant. A low power microcontroller needs to operate on this data for another 5 ms. The microcontroller consumes 3 mA at 3.3 V, when processing the data. Finally, the RF link requires 30 mA at 3.3 V for 30 ms to transmit the data. The desired measurement frequency is 0.2 Hz (one measurement every five seconds).

Pa, or average power, is the key term that will tell us what types of energy harvesting transducers, if any, will be suitable for this system. Table 2 lists some typical energy transducers and the typical average power they are capable of delivering. The column labeled (K) is a power conversion constant that takes into account the type of power management block that is required to convert the transducer energy to a usable voltage, (3.3 V) in this case. A perfect power converter has a K=1. K will vary with the type of transducer employed. Generally speaking, K is proportional to the output voltage of the transducer. Since very low output voltage transducers like TEGs require an extremely high boost ratio and correspondingly high input currents, K tends to be lower than very high output voltage transducers like piezo elements. In the previous example, we can see that the average power required (Pa) is approaching the upper range of piezo transducers of a reasonable size, but is within the capabilities of TEGs and photovoltaic (PV) transducers or solar cells.

The system environment will usually dictate what type of transducer is selected. In our example, we cannot depend on an always-available light source, so PV transducers are not practical. We are at the upper end of what is feasible for piezo transducers, so we decided to use a TEG. TEGs utilize the Seebeck effect to generate a voltage across their output terminals when exposed to a temperature differential (see Figure 2).

Figure 2: Typical TEG.

Figure 3: Typical Current pulse during measure and transmit cycle.

Figure 4: VOUT ripple during measure and transmit cycles.

To further our example, let's assume that a 50 mm2 TEG is selected. One side of the TEG will be mounted to the HVAC duct in the ceiling and the other side exposed to room temperature air. Since TEGs have a very low thermal resistance, it's often challenging to develop a suitable ΔT across them, so the room temperature side will employ a heat sink. Our measurements have shown that the HVAC duct surface will average 38°C in the winter (heating) and 12°C in the summer (cooling) with an average room temperature ambient of 25°C. Through careful measurements, we've determined that the ΔT across the TEG is ~+/-10°C when mounted to the duct with a heat sink. From the manufacturer's data sheet we can see that the TEG VOUT with a 10°C dT is 180 mV. The TEG Output Resistance (ROUT) is 2.5 Ω. The maximum power available to the load occurs when the TEG ROUT = Power Converter (or load) RIN.

If we assume that our power management circuit has a RIN near 2.5 Ω, then the maximum power available to the power converter input is 180 mV2/(2.5 ω x 4) = 3.24 mW. Our power converter constant (K) is 0.4, so the total power available to the remote sensor 3.3 V output is 3.24 mW x 0.4 = 1.3 mW. Since 1.3 mW is comfortably above the previously calculated Pa of 818 µW, we can generate enough power to operate.

A power management circuit to convert the very low output voltage of the TEG to the required 3.3 V is the next challenge. A further complication is that the input voltage (TEG output) can be either positive or negative 180 mV, depending on whether the duct surface is hot or cold. While it may be possible to develop a discrete circuit to meet this challenge, it is often very difficult to achieve a solution that meets the system requirements for manufacturability, small size and reliability. Further, circuit design is extremely sensitive to stray capacitance and the entire circuit needs to be micropower to achieve the rated K factor.

Fortunately, an integrated solution exists today. An example circuit using the LTC3109 is shown in Figure 5. The LTC3109 can operate from inputs as low as +/-30 mV and will produce any of four pre-programmed output voltages (VOUT): (2.35, 3.3, 4.1 or 5 V). A switchable VOUT is provided to power the sensors only when necessary. The LTC3109 also includes a power manager that is useful for storing and utilizing excess harvested energy. Since our typical load power is less than the available energy, any excess energy can be stored for later use on CSTORE.

Figure 5: LTC3109 Power Management Circuitry.

Figures 3 and 4 show the 3.3 V output of the LTC3109 before, during and after a measurement/transmit cycle. The capacitor on VOUT is sized based on the acceptable voltage droop for one measure/transmit cycle. In our example, we've determined that a voltage droop of 300 mV is acceptable on the 3.3 V output. Using the values obtained previously, we can calculate the required COUT:

  Iload = sum of all the loads on the 3.3 V output
  Iavg = average output current of the LTC3109
  dT = duration of the load pulse
  dV = acceptable voltage droop

The actual droop in Figure 4 is much less than the 300 mV. This is due to a lower current transmit pulse duration for the simple system that was measured and the higher output capacitance.

Figure 6 shows the 3.3 V output during a temporary interruption of the energy harvesting transducer input. In this case the LTC3109 operates from the storage capacitor, CSTORE. There is no limitation on the value of CSTORE, so it can be sized for whatever system holdup time is desired.

Figure 6: Operation during input source interruption.

The basic design procedure outlined above is applicable to other types of energy harvesting transducers. Power management circuits that interface with piezo elements (high voltage AC), electromagnetic (coil/magnet) and photovoltaic (solar cells) are all readily available today. In all cases, it is necessary to first determine the average load power required to see if autonomous operation is feasible.


Average load power is the key variable to consider when contemplating the use of energy harvesting techniques to supplement or replace batteries in remote wireless sensor networks. The operating environment will always dictate what types of energy harvesting transducers are suitable and average load power will further narrow the choices. Power management solutions are now available to bridge the gap between low output power level transducers and ultra-low power microcontrollers, sensors and RF links. With all of the necessary elements in place, semi- or fully autonomous remote sensor networks have left the theoretical realm and are now poised to enter the mainstream.