Power over ethernet: simplifying the equationFollow article
Power over Ethernet started out as a proprietary technology developed by network equipment vendors such as Cisco, to deliver power to VoIP (Voice over Internet Protocol) phones without using separate power supplies. Other manufacturers soon realised the advantages of sending DC power over the network: especially, cheaper cabling and higher data rates than alternative USb and aC powerline technologies.
These are benefits not only for IP telephones, but also for security cameras with pan/ tilt/ zoom functions, embedded computers, Ethernet switches, thin clients and even LCD displays. Such diversity also brought the need for standardisation.
The original IEEE 802.3af-2003 standard defined the implementation of sending power as well as data over Ethernet cabling. Category 5 cables contain four twisted pairs, but only two are used for data transfer in either 10BASE-T or 100BASE-TX. This leaves two pairs available to carry DC current. Power is delivered to the end device (PD) either by PoE enabled Ethernet switches (endspans) or by injection from a midspan power supply.
Updated in 2009, IEEE 802.3at uses phantom power techniques to allow powered pairs to also carry data. This extends its use to 1000BASE-T, which uses all four cable pairs for Gigabit/s data. In real-life applications, PoE Plus or PoE+ is able to deliver 25.5W at 44V and up to 350mA down a standard Category 5 Cable using two twisted pairs. [Midspan manufacturers such as Phihong have extended this capability by delivering power through unused data pairs as well as the network cables in use.]
PoE support from RS
RS stocks many components to support PoE designs including ICs from Texas Instruments, National Semiconductor, Linear Technology and STM Electronics. In addition, our range includes off-the-shelf solutions such as powered device (PD) interfaces from Murata, PoE enabled microcontrollers from Olimex and Phihong’s range of midspan injectors and splitters.
Article from eTech issue 6