Blog: The value of Power-over-Ethernet (PoE)

Feature articles |
By Nick Flaherty

Power-over-Ethernet (PoE) comes into its own with IIoT and Industry 4.0 applications as the PoE standard allows unused wires or data lines within an existing Ethernet cabling system to power sensors and actuators.

In the previous two blogs, we discussed the macro considerations for power within industrial settings. We highlighted how motivation for energy and process efficiency is changing the landscape of the factory floor. The adoption of Industry 4.0 and IIoT is pushing data storage away from local servers toward the cloud at one extreme and data processing toward the sensors and actuators at the other, utilising ‘edge’ computing for fast, localised decision making.

Mixing power and data

Powering the ‘edge,’ where sensors may feature a microcontroller, memory and a wireless or wired data link, requires access to a supply. For wireless actuators and sensors, the options are battery, AC-DC adapter, or perhaps energy harvesting. However, wired Ethernet data connections are often employed to supply power as they are ubiquitous. The reasons are not just historical; wireless is inherently less secure, and most devices work in the license-exempt bands. There can be congestion, packet delays, and stability issues, especially with multiples crowded in an electrically noisy industrial environment.

The development of Power over Ethernet (PoE)

The Ethernet standard has existed since 1973, though initially a coaxial cable-based system, differential signals on twisted pair wires are now used extensively. Gigabit/s communications often use four twisted pairs. Lower speeds use only two twisted pairs – typical for industrial use to process data from sensors. PoE can use existing Ethernet installations in areas lacking AC-DC power outlets. Additionally, it can be installed without reliance on qualified electricians.

IEEE 802.3af standard

Under the 802.3af standard (IEEE 802.3 – 2003), unused pairs could deliver power; known as mode B or Midspan. The PoE midspan method uses PoE Injectors to add PoE to your networks without adding or replacing your existing non-PoE switches. These are intelligent devices; they detect whether the Powered Device ( PD) requires power and if so, enables it. PoE injectors are useful when you have a small number of devices that require power.

The IEEE 802.3af standard also allows power to transmit via the data lines as a common-mode voltage, applied to each pair. Because Ethernet uses differential signalling, the data transmission remains unaffected, known as mode A or Endspan.

Figure 1 demonstrates the use of centre taps on the pulse transformers at either end of a cable run to apply and then extract the power from the data signal.

Figure 1: PoE using two twisted pairs in an Ethernet cable

This technology, often integrated into RJ45 connectors from suppliers such as Molex and TE Connectivity, simplifies designs and saves board space.

The PoE endspan method uses a switch with an integrated PoE. You connect your end device to the PoE switch, and it will detect whether the end device is PoE compatible and, if so, enable power automatically.

PoE power levels have increased dramatically. The IEEE 802.3af PoE standard details a technology designed to provide up to 15.4 W of DC power (minimum 44VDC and 350 mA) to each device. Due to losses in the cable, the standard guarantees only 12.95 W to be available at the powered device (PD), enough for small sensors, security cameras, and similar loads.

IEEE 802.3bt Type 4

Over time, the standard has evolved to allow higher power levels under specific conditions – In 2018 IEEE introduced a new high power standard 802.3bt type 4. Under IEEE 802.3bt Type 4, a maximum of 100W from the power sourcing equipment (PSE ) is permitted, from a voltage range of 52-57V. A minimum voltage of 41.1V at the powered device (PD) is left if the cable resistance is a maximum of 12.5Ω and cable temperature fixed within certain limits. The use of all four pairs to achieve this maximum power is essential.

The original simple applications considered for PoE have now expanded, along with the increased power rating. these can now potentially supply power to:

  • wireless access points
  • routers
  • security panels
  • A suitably rated laptop or tablet  

PoE Lighting

PoE lighting is an area generating much interest; with the rapid introduction of LED fixtures, PoE power is certainly enough for emergency lighting and, in some cases, even ‘normal’ use. When lighting fixtures include both light and motion sensors, a whole world of control and monitoring becomes possible. For example, during an emergency inside a building, a battery supplied PoE safety lighting system could sense where a person is in the building, and guide them to safety by illuminating or changing the colour of selected LED lights – providing a safe path for them to follow.


Power over Ethernet (PoE) is an ingenious way of distributing power using existing cabling networks. It allows devices in remote locations access to a source of local DC power. Protocols and standards facilitate the use of these existing structures, without interrupting data flow. Industry 4.0 and the IIoT rely on such innovation for data acquisition and will employ PoE for more and more use-cases as the development of this technology progresses.

In the next blog in the series, we will look at the increasing viability of utilising energy harvesting as an alternative solution for powering IIoT sensors.

Other power blogs

Other articles on eeNews Power 


Linked Articles
eeNews Power