Power over Ethernet Technology

    In this article, we consider the aspects of the application and purpose of PoE technology, the principles of supplying power to the line, and the impact of implementation on the power supply network.

    Adopted in July 2003, the IEEE 802.3af standard regulates the method of power supply to devices connected to an Ethernet network, which involves the use of a cable designed for data transmission.
    Wikipedia defines PoE as follows:
    Power over Ethernet, or PoE, is a technology that allows electrical energy to be transmitted to a remote device with data through a standard twisted-pair cable in an Ethernet network. This technology is intended for IP-telephony, access points of wireless networks, Web-cameras, network hubs and other devices to which it is undesirable or impossible to conduct a separate electric cable.


    From where it immediately becomes clear why this technology was invented - who wants to drag the power cable to a webcam or VoIP device, thereby increasing the cable economy of the office or home.
    It’s true that you need to rush to implement PoE everywhere and everywhere, because this implementation requires a revision of the issues of power supply and cooling (naturally, we are talking about industrial implementation)

    If you do not take into account the increased demand for power supply and cooling, you should be prepared for unforeseen shutdowns and malfunctions of the equipment. Since PoE is commonly used in communications and access control systems, the negative effect of such outages is especially noticeable.

    How is power supplied to the line?


    The following methods are provided for powering an Ethernet network:
    • switch with voltage supply in the line;
    • "Power injector" installed between the switch and the cross;
    • cross with voltage supply to the line;
    • “Power injector” for one line (the latter provides power to PoE-compatible devices in networks lacking such support)



    PoE technology does not affect the quality of data transfer. For its implementation, the properties of the Ethernet physical layer are used:

    A) The use of high-frequency transformers at both ends of the line with a central tap from the windings. A constant supply voltage is supplied to the central taps of the secondary windings of these transformers, and is also removed from the central taps on the receiving side. The use of central taps of signal transformers allows, without mutual influence, to transmit along one pair of wires and high-frequency data and a constant supply voltage.

    C) Using free pairs to connect power. Modern Ethernet cable networks that comply with the 100Base-TX standard consist of four pairs, two of which are not involved.

    Accordingly, there are two options for powering the PSE device:
    • A - on twisted pairs for which data goes
    • B - for unused pairs in the cable


    PSE devices (injectors) differ, depending on options A or B, while the splitters, i.e. PD devices, are universal. The PD device must be able to receive power in any version, including when changing polarity (for example, when a cross-over cable is used, such as computer-to-computer). It is important that the PSE only supplies power to the cable if the device to be connected is a PD device. Thus, equipment that does not support the PoE standard and is accidentally connected to the PSE device will not be disabled. The procedure for supplying and disconnecting power to the cable consists of several stages.


    The maximum permissible power level is about 15-16 W per port at a current of 350 mA and a voltage of 48 V, based on these limitations, a list of devices is drawn that can be powered by this technology.
    A typical IP phone consumes up to 5 watts of energy, an IP phone with video support, like an RFID tag reader or a fixed surveillance camera, 10-12 watts, a wireless access point, a card reader or an electromechanical lock, from 8 to 12 watts.

    The development version of the PoE + standard (IEEE 802.3at), which will allow connecting devices with power consumption up to 30 W when transferring data over two pairs (10BaseT and 100BaseTX) and equipment up to 60 W in case of transferring over all four pairs (for gigabit Ethernet 1000BaseT) . Thus, Wireless Mesh Network access points with a power of up to 50 W, managed video surveillance cameras with an energy consumption of up to 20 W and other fairly powerful devices can be powered by a single wire.

    What should I look for when implementing PoE in a corporate network?



    Let's say our office has 50 cars. There are two 48-port switches and an “injector” for supplying voltage to PoE-enabled devices. And, suppose that our devices require no more than 7W of power.

    Without the use of PoE technology, the power consumption of the switching equipment of a small office will not exceed 90 watts. A 500 W / 750 VA UPS is sufficient to support this load.
    However, if we are supplying all telecommunication equipment, webcams and other additional equipment from the UPS, then we will have to take into account the increased load (for a small office with 50 active devices it will increase by about 6 times) on the UPS, which may entail replacing the UPS with more powerful analogue.
    You may also need to perform a heat dissipation calculation because a new UPS in battery mode will generate more heat.

    via wikipedia , lan mag


    Also popular now: