Bench Talk

(Source: deepadesigns/Shutterstock.com)

Wireless sensor nodes (WSN) are gaining more attention in recent years. One of the main advantages of a WSN is its ability to monitor data of various parameters constantly. These sensors are used, for example, in water pipe systems to monitor fluid and flow parameters. One of the challenges of the WSN is its power supply. It is a battery-operated device, and once the batteries have reached their end of life, they need to be changed or recharged. Because the WSN can be placed in a remote location, replacing the batteries is not always practical. An existing approach to overcome this challenge is by adding energy harvesting to the WSN. Energy harvesters can scavenge available energy from the environment, such as mechanical, thermal, or photovoltaic energy, and convert it into electrical energy.

Typically, a wireless sensor node consists of four main components:

• A power source
• A microcontroller (MCU)
• A communication module
• A sensor

This structure is illustrated in Figure 1 and serves as the basic architecture for a sensor node.

Figure 1: Architecture of a WSN (Source: Mouser Electronics)

Design engineers have tried to regulate the power consumption within the WSN by programming its operation based on duty cycles. Here the sensor node is continuously in low-power mode (deep sleep) and is activated for a short time for data acquisition, calculations, measurements, and communication. Figure 2 demonstrates this working principle.

Figure 2: Diagram of a WSN duty cycles (Source: Mouser Electronics)

On average, a microcontrollers sleep power consumption is in the nW range, and when it's in the active mode, the power consumption is increased to the µW range. However, the active mode is not the main power consumption element here, as this usually happens only for a concise time (milliseconds). The sleep current is the primary source that draws out most of the energy from the batteries. Therefore, when designing a WSN, it is essential to choose an MCU with ultra-low power consumption in sleep mode.

For example, Texas Instruments offers an ultra-low power MCU, the MSP430FR600x, ideal for ultrasonic water-flow measurement. This MCU consumes in active mode approximately 120µA/MHz and in standby mode 450nA.

Another product example that is ideal for battery-operated devices is the Silicon Labs EFM32PG22 MCU. It features a 76.8MHz ARM Cortex^®-M33 processor and consumes 26µA/MHz in active mode. The sleep current of the EFM32PG22 is around 1.10µA and can go as low as 0.17µA.

Two approaches for energy harvesting extend the life of a WSN:

1. Extending the battery life by recharging the battery with the harvested energy.

For this, a power management circuit is needed to store the harvested energy in a battery or supercapacitor. The power management circuit consists of three main components: Rectification, DC/DC converter, and storage.

• Rectification: Rectification is only needed for energy harvesters that generate AC voltage, for example, electromagnetic and piezoelectric.
• DC/DC Converter: The DC/DC converter must regulate the generated voltage from the energy harvester and use it to recharge the battery.
• Storage: An energy storage element is used, such as a rechargeable battery or a supercapacitor.

This approach can be demonstrated on the e-peas DEMPV-BLE photovoltaic IoT demo kit. The DEMPV-BLE is a WSN that has on-board a 35mm x 50mm photovoltaic solar cell. This kit can measure luminosity, temperature, and humidity and send the data via Bluetooth^® Low Energy (BLE) to a smartphone. The onboard solar cell can supply sufficient energy for the kit to operate in an indoor environment. The harvested energy is stored in a 160mF supercapacitor and can operate for five hours by sending a beacon message every 10s with new measurements.

2. The harvested energy is directly used to power the WSN components, and so it becomes self-powered.

This approach supplies the generated energy directly to the WSN without the need for a storage element. The only challenge with that approach is ensuring the harvested energy can cover the startup power and the transmission power of the WSN. Let's explain this on the ON Semiconductor Zigbee Green Power Energy Harvesting Kit. A system block diagram is demonstrated in Figure 3. The kit has onboard an electromagnetic generator in the form of a switch that transforms the generated mechanical energy into electrical energy. Each switch press charges a 33µF capacitor, which feeds the NCP170 low-dropout (LDO). The NCP170 then provides a 3.3V supply to the NCS36510 transmitter. Then the voltage gradually decreases until the NCS36510 device powers down. This results in approximately 17ms of available run time for the device. The startup of the device occurs in less than 5.6ms, and a single message is immediately transmitted. After transmission, the module switches to receive mode to listen for an 802.15.4 ACK (acknowledge) from the receiver. Once received, the radio is powered down.

Figure 3: Zigbee Green Power energy harvesting kit system block diagram (Source: On Semiconductor)

What did we learn here? We discussed some of the highly power-efficient MCUs available, ideal for designing a WSN. Furthermore, we also showed two different types of WSNs that can be powered by energy harvesting. One approach is to recharge a battery that supplies power to the WSN, and the other one is that the WSN directly uses the generated energy by the harvester. Deciding which to use for the application depends on your application power consumption. If the energy harvester can cover your application's power consumption, then going battery-less will save time and money.