Power Over Ethernet-Plus PSE And PD Chips With Real Two-Event Classification Play Nicely Together

April 10, 2008
Last month, MicroSemi and Akros Silicon demonstrated the first compatibility between ICs for IEEE 802.11at Power over Ethernet Plus/HiPOE power source equipment (PSE) and powered devices (PDs). The PSE chip was MicroSemiâ??s PD64001, and

Last month, MicroSemi and Akros Silicon demonstrated the first compatibility between ICs for IEEE 802.11at Power over Ethernet Plus/HiPOE power source equipment (PSE) and powered devices (PDs). The PSE chip was MicroSemi’s PD64001, and the PD chip was Akros Silicon’s AS1135.

Conducted by MicroSemi, the test predates the actual release of the final 802.11at standard. But the draft standard was released last fall, and an approved version is expected late in 2008. So, the compatibility demonstration is a genuine milestone.

Compatibility means that it is now possible to design an Ethernet switch or midspan power injector or an Ethernet device such as a video-over-IP telephone that will together negotiate variable power requirements up to about 30 W per channel on the fly. The ability to dynamically shift power among multiple loads promises lower costs for switches and midspans by sizing their power supplies based on statistical analyses of peak power demands, rather than on theoretical maximum demands.

A LONG ROAD Power-over-Ethernet (PoE) is a scheme for powering Ethernet-connected devices over the Ethernet cable itself, rather than via a wall adapter. It evolved from a proprietary Cisco approach developed for voice-over-IP phones.

Defined by the IEEE 802.11af standard, its first inception limited power to 15 W, delivered at the switch or midpoint. A simple hardware handshake established the presence of a PoEpowered device at one end of the cable and the presence of a PoE-enabled switch at the other. There were four possible maximum power levels, and once the handshake was accomplished, the maximum power available to the PD could not be changed. Basic 802.11af PoE proved adequate for some applications, especially stationary security cameras and Wi-Fi hot-spots. But 15 W wasn’t enough for applications such as video phones, which have power-hungry backlights, nor for WiMAX hotspots and surveillance cameras with pan/tilt/zoom motors. That need led to the formation of the IEEE 802.11at Working Group.

The Working Group soon established that CAT-5 Ethernet cable and its associated RJ-45 connectors could handle at least double the maximum 802.11af current. (The nominal PoE voltage at the PD is 48 V dc.) But the group also saw the need for dynamic power rationing. Without it, the manufacturers of big Ethernet switches would have to install power supplies capable of providing full power to every port.

While the 802.11at standard has been pending, several companies, including MicroSemi and Akros, have developed products that allow individual Ethernet ports to be powered with up to 60 W, using the basic 802.11af handshake. That’s really all that’s required for a security camera or WiMAX installation, given a dedicated run of CAT5 cable. That approach wouldn’t work, however, in a cube farm where workers and their IP phones would be frequently shuttled here and there.

TWO-EVENT CLASSIFICATION The process of dynamically negotiating maximum power for a PD involves both the basic 802.11af hardware handshake and an ongoing dialog between the PD and the PSE via the Ethernet data link itself

View the chip

Sponsored Recommendations

Comments

To join the conversation, and become an exclusive member of Electronic Design, create an account today!