PoE Frequently Asked Questions

What is PoE ?

Power over Ethernet or PoE is a cost effective solution which allow both data and power to be
provided over an ethernet cable to an endpoint.

Why use PoE ?

There are a number of reason why you would use PoE, but in most cases the overriding reason is
cost.

An example of this is: -

It can cost between $100 and $200 to run an ethernet cable to a wireless access point. But it can
costs a further $400 to run an electrical conduit to power that access point.

The alternative is to install Power Sourcing Equipment (PSE) called Endspan Equipment into the
ethernet switch, or by adding Midspan Power Insertion Equipment after the ethernet switch. This
will combine the data and power at the source to be sent through the ethernet cable. The power
from the cable can then be extracted at the endpoint with a Powered Device (PD) to operate the
access point.

This solution removes the need to run expensive electrical conduit and will cost considerably
less than $400.

Is there a PoE Standard ?

Yes, the Institute of Electrical Electronics Engineers (IEEE) 802.3af Task Force has ratified
the technology into the Ethernet standard. The IEEE 802.3af standard covers the interoperability
of using PoE products and covers the PSE - Endspan or Midspan, the Powered Device (PD) and
the cabling requirements. It is endorsed by all network players that power IP-connected devices,
such as voice-over-IP (VoIP) phones, wireless LAN (WLAN) access points (APs) and advanced
security devices, over the existing Ethernet cabling infrastructure.

How is the power put onto the ethernet cable ?

The power can be added to the ethernet cable if the switch has a built-in power source in the
form of PSE is called Endspan Equipment, as shown below -

Alternatively the power can be added to the cable after the switch by using a PSE called Midspan
Equipment, as show below -

What does the PSE actually supply ?

Under normal operation the PSE supplies a nominal d.c. voltage of 48V, this is limited to a
maximum of 15.4 Watts (at the present time).

How is the power extracted from the cable ?

To extract the power from the cable the connected device needs to be a Powered Device
(PD) such as the Ag9000.

How does the PSE know that a PD is connected ?

Before the PSE supplies power, it has to verify that the device connected is a Powered Device
(PD). It does this by checking that the PD has a conforming signature. The Ag9000 PoE module
has a built-in signature chip that does that automatically. Only upon successful completion of a
valid signature will the PSE apply power to the cable. This ensures that non Powered Devices are
not damaged by the PSE, thus making the system backwards compatible.

What is Power Class ?

This is optional within the IEEE 802.3af specification, but can be used by the PSE to manage the
overall power requirements of the system. Below is a table that shows the various power
classes detailed within IEEE 802.3af.

CLASS	Usage	Maximum Power
0	Default	15.4W
1	Optional	4.0W
2	Optional	7.0W
3	Optional	15.4W
4	Reserved for future use	Treat as CLASS 0

The Ag9000 PoE module has the option of setting the power class, if required.

What is Power over LAN?

Basically the Power over LAN or PoL is another term used for Power over Ethernet or PoE.

Do Silver Telecom's PoE modules comply with EMC
regulations?

FCC Part 15 is the EMC spec. This is not directly relevant to our PD modules, but is relevant to
the finished system. So, for example, a Wireless Access Point unit produced using our Ag9000
would have to meet FCC Part 15. In order to do this it may be necessary to add a common mode
inductor at the input to filter out noise generated by the Ag9000. The Ag9000 module already has
filters fitted internally at the input and output. These may not be sufficient for a particular
system as there are limits to how much filtering can be done on such a small module. Also the
customer's system may generate noise of its own, which can be conducted through our modules
and thus require extra filtering to remove.

It would not be helpful for us to get an FCC Part 15 certificate for our modules because this
would be specific to the one test configuration that we measured. As soon as it is put into a
customer's product then the configuration is changed and the tests would have to be repeated.

FCC Part 68 is the telecoms specification. Since an Ethernet network is not classed as a
"Telecommunications Network" by FCC Part 68 then this specification is not relevant.

click on link for a pdf version of this article