It's an interesting technology, but still in it's infancy, and suffering from the chicken/egg problem. I've started to deploy it in limited ways for some of my clients, and it does what it's supposed to.
Since it's still early, you have to be wareful of products claiming
802.3af when they aren't (Linksys WAPPOE devices, for instance, apply power to the spare pairs without bothering to check if there's a valid client on the other end).
Prices tend to be obscenely high, though the PhiHong stuff is on the order of $25 for the single-port midspan.
Finding client devices with PoE already in them isn't easy, life will be so much better when the client hardware is cheap enough that it's installed as a matter of course in all but the lowest-end devices.
I haven't gotten to the point where I'm buying products specifically for their PoE capabilities...
Nowadays standardized and starts to be pretty widely supported by devices.
Proprietary implementations that were used earler are going out.
I don't know. Maybe at least some small boom.
Pros: Can supply power though same wiring as Ethernet signal to power devices like IP phones, WLAN access points etc.. Less wiring needed for those devices (just net cable), possibly for easier power backup systems (PoE powered WLAN stations receiver power from one point from switch, just put UPS/generator there and you whole wireless network infrastructure does not stop if mains power goes out).
Cons: Switches that can supply power tend to be quite expansive.
More information on power over ethernet system can be found at
Yeah, there are some apps that look really neat, like that PoE clock that sets itself using SNTP, but it really should cost less than a computer that does the same thing. 8*)
There are two methods of providing power in 802.3af:
Power supplied over the spare pairs. This is the method supposed to be used for mid-span power injectors.
Phantom power over the data conductors. This method is supposed to be used for PoE switches.
Clients/nodes must be able to accept either.
Some of the high-power options they are working on use both, though there are some interoperability details to be worked out.
Obviously, Gigabit ethernet has no spare pairs, so must use the second method, or a special gigabit-compatible PoE midspan. Normal midspans will probably break Gigabit, or cause it to negotiate downwards to
There is a long tradition of supplying power on signal carrying cables, such as antenna mounted TV signal amplifiers. In that case, capacitors are used to pass the RF signal, and resistors or inductors to get the power out. It is usual to us 60Hz AC instead of DC to reduce the electrolytic effect when things get wet.
For gigabit, one could couple into a pair in a similar way, or do it between pairs. In any case, it must be done very carefully, so as not to affect the signal. One possibility is center tapped transformers to allow power to be sent between two (or more) pairs. I have not actual tried it, and don't know how hard it would be to do. It might be that the transformers aren't designed for that much current, or that the transformer core might saturate, for example.
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.