Understanding Power over Ethernet for video surveillance

The basics of PoE explained, plus a plea to the vendors to help IP video system designers

I’m not an EE, so what the heck is PoE?

PoE was, and is, supposed to make the powering of devices easy. You take your camera or other device that accepts power via the Ethernet port, you plug in the RJ45 jack to the port, and you walk away. Inside the head end, you plug the other end of the same Ethernet cable into a PoE switch or PoE injector and voila, power is magically delivered to the device along with the data connection. In theory, all of the normal worries are gone. AC power or DC power is irrelevant, and you don't even have to worry about over-powering a camera that, were you to fry it, could potentially set you back a few thousand dollars in equipment costs and man hours!

PoE was supposed to be this way, but practical reality has diverged from the perfect world concept in such a way that the actual installation is almost never that easy. So set aside the “perfect world” notions you have, and let’s start with the basics, so you can understand how PoE works.

There are four classes of PoE: Class 1, 2, 3 and 0. Each PoE classification denotes a range of power that is available to the end device as well as the power that must be available on the port of the power sourcing equipment (PSE):

PoE Classifications

  • Class 1 --  4.5 watts at PoE port; 3.84 watts at device
  • Class 2 --  7.5 watts at PoE port; 6.49 watts at device
  • Class 3 --  15.4 watts at PoE port; 12.95 watts at device
  • Class 0 --  15.4 watts at PoE port; .44 to 12.95 watts at device

In the world of PoE there are two kinds of switches that can provide PoE; the kind that operates with a “guarantee per port” and the kind that operates with a “total power budget”. Both kinds of switching are useful but there is a significant difference between them. If you happen to have a switch nearby, look at it and see if you can tell into which one of the above two categories your switch falls.

A switch that guarantees a certain wattage per port -- 15.4 watts per port, for example -- means that you can be sure that no matter how many Class 3 or Class 0 devices are plugged in, the switch will be able to power them. Of course, these switches tend to be bigger, more expensive and ill-suited for use outside of a nice climate controlled room, but they do prevent errors in power planning.

The second type of switch mentioned above -- the kind with a total power budget -- can only power as many PoE devices as it has power to spare. Imagine that you are working with a 4-port switch that carries a total power budget of 30 watts. This kind of switch could power four Class 2 cameras (4 devices x 7.5 watts = 30 watts needed). It could also easily power four Class 1 devices (4 devices x 4.5 watts = 18 watts needed). Continuing with that math, it would be able to power Class 3 or Class 0 devices, but it could only power two of those types of devices.

Power planning is where the rubber meets the road, and it brings up a challenge in our industry.

What happens if a chosen device (i.e., a PoE powered camera) does not clearly specify the PoE class and instead simply gives an operating wattage? You might think that this is OK since a camera which says “6.01 watts” is within the Class 2 specifications and therefore must be Class 2. But that’s where reality often diverges from common sense. In theory, what is supposed to happen is that a device is clearly labeled with a PoE classification so that when said device is plugged into a PSE device, the power budget has been worked out such that each device will receive its required PoE.

What I believe the security industry needs – right now, since PoE is happening today -- is clear labeling of the correct classification of PoE on each and every device that uses PoE. It is all well and good to place the operating or maximum wattage on the device, but industry manufacturers need to take the next step!

This content continues onto the next page...