Skip to content

802.3at, and the allowance that has to be granted

Type 2 is the class most ceiling-mounted access points name on their data sheet, and the one most often mismatched with the switch already on the wall. What it raises, and what it still leaves behind on the cable, are both published.

What the standard defines

IEEE 802.3at was published in 2009 and has since been folded into the base 802.3 standard. It kept the earlier behaviour intact as Type 1 and added Type 2, which is what packaging calls PoE+.

Type 2 raises the published pair. The source delivers up to 30 W at its port, and the standard guarantees the powered device 25.5 W. The gap has the same cause as before — the cable between them takes its share — and it is the reason both figures are published rather than one.

The interesting addition is not the wattage but the negotiation around it. The standard defines an additional classification step by which a Type 2 source and a Type 2 device confirm each other, either through a further exchange at the physical layer or through a data-link protocol between them. The higher allowance is granted; it is not assumed by a device that would like it. That is why two devices which both say “PoE+” on the box can behave differently on a switch that supports only one of the two methods.

Type 2 still uses two of the four pairs, and it names its cabling: a Category 5e channel or better.

What it does not promise

It does not promise thirty watts at the far end. Thirty is what leaves the port; twenty-five and a half is what the device is entitled to, and a long run is where the difference is felt.

It does not promise every port at once. A switch’s per-port class and its total power budget are separate published numbers, and an eight-port unit naming Type 2 throughout rarely has a budget of eight times the figure.

It does not promise four-pair power. Delivering current on all four pairs is a later addition, with its own Types.

It does not promise anything to a Type 1 source. A device built for Type 2 on a switch that offers Type 1 gets the lower allowance, and what happens then is the device maker’s decision rather than the standard’s.

What it works alongside

A Type 2 source powers Type 1 devices without configuration, resolving the negotiation downwards. Type 2 devices are powered by Type 3 and Type 4 sources in the same way. The power travels alongside the data on a Category 5e channel or better, so a powered link runs at whatever rate the two ports negotiate — the two things are carried on the same cable and decided separately.

The questions that come up before an order

Will a device specified for Type 2 run from a Type 1 source?

Not at full function. Some models start in a reduced mode — a radio disabled, a port unpowered — and the maker states which, on the data sheet. Others do not start at all. It is the one line to check against the switch already in place before ordering.

Is an injector labelled PoE+ the same as a Type 2 source?

Only if its own data sheet says so in the standard’s terms. The trade names are not the standard; the Type number and the published wattage are, and an injector that names neither is describing itself rather than a specification.

Last reviewed 9 September 2026