Power over Ethernet (PoE) in Industrial Automation
Contributed By Digi-Key's North American Editors
Power over Ethernet (PoE) connectivity systems are defined by the IEEE 802.3 standard and provide a convenient way to provide both data and electrical power into machine components using single all-in-one Ethernet cables. Read the 2013 Digi-Key article An Introduction to Power-over-Ethernet for a primer on PoE’s introduction and structure.
Today there are three standard subclassifications for PoE:
- PoE Alternative B uses separate wires to carry data and power. More specifically, this PoE subclassification specifies the use of Cat-5 Ethernet cable with four twisted pairs — with two pairs carrying data and two pairs carrying power. PoE Alternative B arrangements can therefore only support a data rate of up to 100 Mbps (100BASE-TX) — even though cables rated for Gigabit Ethernet are used.
- Alternative A is also limited to a data rate of 100 Mbps but uses the same two pairs to carry both data and power. This means PoE Alternative A arrangements are compatible with Ethernet cable containing two twisted pairs as well as full Cat-5 four-pair cables.
- 4PPoE uses all four twisted pairs to transmit power and can therefore provide higher currents. The twisted pairs also carry data at the highest data rates of Gigabit Ethernet and beyond.
Figure 1: Here is one wiring variation of a Power of Ethernet (PoE) installation. (Image source: Design World)
These three main standards are often simply referred to as mode A, mode B, and four-pair PoE. Different pin configurations of each mode are possible. No matter the variation though, any powered devices (PDs) must be able to accept both mode A and mode B connections.
Ultimately, it’s the PoE system’s power sourcing equipment (sometimes abbreviated PSE) that determines which mode is used and may support just one or multiple modes. A PD can indicate its compatibility with PoE by the resistance across the powered pairs. A fixed 25-kΩ resistance indicates general compliance with standards while a changing resistance can be used to request a specific supply mode.
Figure 2: MAX5969A/MAX5969B ICs install in PoE powered device (PD) assemblies. The IEEE 802.3af/at-compliant ICs provide PDs with detection and classification signatures as well as an isolation power switch with inrush-current control. (Image source: Maxim Integrated)
Differential signaling and supported power capacities
Ethernet cable transmits data on twisted pairs using differential signaling. This means that each wire in a twisted pair carries the same information so that the signal-receiving device can measure the voltage difference between the two. Such arrangements are far more reliable than designs that simply track voltage on a single wire with respect to ground, as these allow for the detection and rejection of any electromagnetic interference (EMI) affecting the cable. PoE use of twisted pairs also means that the voltage in a wire pair can be increased to transmit power without negatively affecting the data signals also carried by the cable.
Figure 3: Various PoE arrangements employ two-pair and Cat-5 four-pair Ethernet cable. A main PoE benefit is how only a single cable needs to be laid to connect PoE devices. (Image source: Getty Images)
As the PoE standards have developed, the amount of power that can be transmitted has increased. These are represented by four generations of PoE or Types:
- The original Type-1 PoE supports power delivery of up to 13 W with a voltage range of 37 to 57 V. This is typically sufficient for devices such as wireless access points and door access panels.
- Type 2 or PoE+ supports power delivery of up to 25 W with a voltage range of 42 to 57 V. With this level of power, devices such as security cameras, RFID readers, and alarm systems can also be supported.
- Type 3 supports power delivery of up to 51 W, with a voltage range of 42 to 57 V. This is sufficient to power laptops and control panels.
- Type 4 supports power delivery of up to 71 W with a voltage range of 41 to 57 V. This is particularly useful for powering LED lighting, enabling smart lighting without a mains power supply.
For Type 1 the maximum cable resistance is 20 Ω but the higher currents of later generations limit this to 12.5 Ω.
Network Components Commonly Found in PoE Installations
Devices used to construct PoE networks include:
- PoE switches, which are network switches that provide PoE on their ports. They are the basic building blocks of extended PoE networks — typically acting as the PSE in most networks.
- PoE injectors add power to an unpowered Ethernet cable. They can be integrated into a design to add power to non-PoE networks. For example, consider a system in which a non-PoE network switch connects a device to the network. If the goal is to power that device through its Ethernet cable, the installer would plug the cable from the network switch into a PoE injector and then run a second cable from the injector to the device. The injector requires its own power supply.
- PoE splitters separate the power and data from a PoE cable, allowing a non-PoE compatible device to be powered using a separate input. A splitter can be thought of like an injector in reverse.
Figure 4: This PoE splitter (see the Phihong PTM for more information) can provide up to 45 W on select IEEE802.3 systems with overcurrent as well as overvoltage protection. (Image source: Phihong USA)
- PoE hubs are essentially a stack of injectors. Multiple unpowered cables are plugged into one side and then cables plugged into the other side become powered.
- PoE extenders enable Ethernet networks to operate beyond their normal 100-m range.
PoE Example Applications
The ability to supply power as well as transmit data via a single cable is indispensable for many applications, as it simplifies and reduces the cost of automated functions. In fact, PoE is particularly useful in locations where there’s no power supply available. What’s more, because of the low voltages involved, installation of a PoE cable does not require an electrician. This can result in significant cost savings if the alternative would be installing additional electrical sockets. It may also be possible to use existing network or telephone cables already present in a facility for PoE.
Because voltages are low, the system is safer. This also means that conduits and earthing enclosures are not needed … which in turn further reduces the cost of installation. The maximum voltages stay within safe limits, and the PSE sends a test current at 10 V before supplying the full power. The full voltage is only applied if a 25 Ω resistance is detected on the PD. This prevents attached devices from being damaged.
When changes are required, it is relatively easy to change Ethernet-connected devices and swap cables. This requires that a technician simply plug the new devices’ cables into network switches. In contrast with automated machine sections running off ac power (which can in some cases require entire automated floors to be isolated while work is carried out), PoE is plug and play. That means changes can be made while the network continues to operate. It’s also easy to employ device-usage data to control power into specific devices. This ability to turn the power into a device on and off can significantly reduce power consumption.
Consider one PoE application: PoE lighting systems are an increasingly common application with wide applicability. They involve LED lights with sensors and lighting controls that connect via Ethernet cables and switches. One use is in warehouse facilities to mimic natural daylight cycles to improve the health, wellbeing, and productivity of workers. Integration of motion sensors and predictive algorithms run on the PoE controllers allow the most efficient use of light to save energy and trim operating costs.
Another relatively new application for PoE is motors. Integrated PoE motors can reduce the amount of cabling needed for discrete automation, as they eliminate the requirement for dedicated feedback cables between motor and external motion controller. In units with a drive integrated into the motor housing, the motor can simply receive control commands along with power over a single Ethernet cable. This reduces the total installation footprint while simplifying the installation process.
Such integrated PoE-ready motors can receive either motion control programs or real-time commands over the Ethernet data connection.
Power over Ethernet (PoE) is useful for devices requiring both power and data connections. It reduces installation costs, provides design convenience, and is safer and more reliable than segregated power and data connections. There are two types of devices making increased use of PoE:
- Devices such as lights that traditionally only required power … but are increasingly being sold as smart devices relying on data transmission for newer and relatively advanced capabilities.
- Components such as electric motors that have begun leveraging the increased capacities of PoE as an economical, safe, and convenient option for both power and data connectivity.
No wonder PoE has fast become a core technology in consumer goods as well as smart buildings and machine automation leveraging industrial internet of things (IIoT) capabilities.
Disclaimer: The opinions, beliefs, and viewpoints expressed by the various authors and/or forum participants on this website do not necessarily reflect the opinions, beliefs, and viewpoints of Digi-Key Electronics or official policies of Digi-Key Electronics.