PoE Mode A vs Mode B: What Is the Difference?

PoE Mode A and Mode B describe which pairs in an Ethernet cable carry the power. The IEEE 802.3 standards call them Alternative A and Alternative B. In Mode A the power shares the pairs that carry data on 10/100 links, on pins 1, 2, 3 and 6. In Mode B the power uses the other two pairs, on pins 4, 5, 7 and 8, which 10/100 Ethernet leaves spare.

The data is unaffected either way. Power is applied as a common-mode voltage across both wires of a pair, through the centre taps of the Ethernet transformers, so the differential data signal on the same wires carries on as normal. This is sometimes called phantom power.

The power sourcing equipment (PSE) picks the mode. A compliant powered device (PD) must accept power on either set of pairs, in either polarity, so for most installations the mode is invisible. It matters when mixing non-standard equipment, testing cables or troubleshooting.

The table follows the standard RJ45 pin numbering. Our RJ45 FAQ covers the T568A and T568B wiring that maps these pins to wire colours.

Mode A is the usual choice for endspan sources such as switches, while midspan injectors have traditionally used Mode B. Under 802.3af, midspans were limited to Mode B; 802.3at later allowed them to use Mode A on gigabit links.

A PD cannot know in advance which kind of PSE it will meet, or whether a crossover in the cabling has swapped pairs. So the standard requires it to accept power on either pair set, in either polarity. Inside the device, two diode bridges or active bridge circuits, one on each pair set, steer the incoming power to the right polarity before the device’s own isolated DC-to-DC converter. A PD that only works in one mode is not compliant, and is one of the first things to suspect when a device powers from one switch but not another.

On 1000BASE-T and faster links, all four pairs carry data, so there are no spare pairs. Both modes still work, because power rides on the centre taps of the transformers whichever pairs are used.

The quickest check is a handheld PoE tester plugged in place of the device. It shows whether power is on pins 1, 2, 3 and 6, on 4, 5, 7 and 8 or on all four pairs, along with the voltage and class negotiated. Without a tester, a managed switch‘s PoE status page often reports the mode and the power each port is delivering. A cable tester confirms that all eight conductors in the twisted-pair cable are connected, which matters for Mode B and four-pair PoE. Never probe a live PoE port with a multimeter on bare pins without an adapter, because a short can trip or damage the port.

With standards-based equipment the mode rarely needs attention. It matters in three cases. First, passive PoE gear often puts a fixed voltage on pins 4, 5, 7 and 8, the same pins as Mode B but without any handshake, so a passive device may not power from a Mode A switch. Second, cables or couplers that only wire two pairs will carry Mode A power but not Mode B, and will stop four-pair PoE++ entirely. Third, PoE testers report the mode, which helps track a fault to the cable or the source.

Our industrial PoE switches follow the IEEE modes, and our guide to industrial Power over Ethernet covers pairs, standards and power budgets together.

We supply PoE switches, PoE computers and powered devices that follow the same IEEE modes and validate each link before the system ships as standard. Send us your device list and cabling, and we will specify the PoE set-up to match.