This is an old revision of the document!
Powering of the IoT devices
IoT requires a constant and reliable power source to operate devices, sensors, and communication effectively. The choice of power source for IoT devices is from traditional batteries to cutting-edge energy harvesting technologies. The factors influencing the choice are:
- device size,
- working environment,
- intended operation,
- time of life,
- operation reliability and availability.
The majority of IoT devices use a DC power source. AC is usually converted into the DC, eventually used for powering high-power actuators.
Batteries (non-rechargeable or rechargeable) provide DC. A plain battery's voltage (1.5V, 3.7V) is usually unsuitable for direct powering of the devices. It also changes over time as the discharge process occurs. For those reasons, converters and stabilisers are used. In the case of the use of modern rechargeable batteries, a charging and discharging module (battery management) is also necessary.
Using green energy sources requires conversion and energy storage as its nature is non-consistent in the time domain. It also requires a different approach towards IoT device control algorithms as it is common that devices are required to put themselves into the low power consumption mode because of the lack of energy.
The majority of IoT devices require one or two voltage standards to power microcontrollers and sensors:
- 5V - common of older microcontrollers (such as AVR) and for most peripherals,
- 3.3V - for recent microcontrollers, also more and more peripherals are 3.3V powered.
Integration of the 3.3V and 5V components in one IoT device is not always straightforward: while controlling a 5V powered device with a 3.3V signal is usually non-problematic, the opposite requires signal conversion because some devices do not accept overvoltage on their inputs!
A special note on powering IoT devices with inductive actuators: a separate powering rail should be used. Actuators using electromagnetic components, such as relays, motors and servos, should not share a powering bus with MCU. They introduce serious inference into the power rail (both rising and falling the nominal voltage), thus frequently causing instability or rebooting of the MCU due to the over / under voltages or even leading to permanent damage. Eventually, capacitor-based filtering can be introduced to reduce power rail inference if separation is impossible.
When dealing with high current, high voltage or high inference devices, it is common to use physical separation of the signals with means of e.g. optocouplers.
[pczekalski]author subchapters
