What is Power over Ethernet (PoE) Technology? Ethernet, also known as Ethernet, uses passive cable as the transmission medium and transmits electromagnetic waves via the Enter command. Ethernet refers to various Local Area Network (LAN) systems that adopt the IEEE 802.3 standard. Power over Ethernet (PoE) is a power distribution technology. A more attractive concept is the ability to simultaneously transmit power and data to any device connected to an Ethernet network. For this purpose, it was adopted as the IEEE 802.3af standard in June 2003 for transmitting and receiving power signals over Ethernet. [align=center]Figure 1: Schematic diagram of providing power to Ethernet via a single CAT-5 cable that transmits Ethernet data, while simultaneously transmitting power and data to any device connected to Ethernet (IP phones, wireless access points, etc.)[/align] Table 1: IEEE 802.3af PD requirements begin with a 25K and less than 0.12μF characteristic identification. For this valid PD characteristic, all the above parameters must be detected by the endpoint PSE or intermediate PSE. Table 2: If the endpoint PSE or intermediate PSE detects the above parameters, it indicates that the Ethernet device is an invalid PD. This power distribution technology uses the IEEE 802.3af standard based on Power over Ethernet. Devices with power consumption less than 12.95W, such as IP phones and wireless access points, can be powered via a single CAT-5 cable that transmits Ethernet data, as shown in Figure 1. This means that not all network devices requiring continuous power, such as IP phones, wireless access nodes, and security cameras or network surveillance cameras, need local AC power. It also means that devices do not necessarily need to be placed near wall power outlets; in some cases, the power cord can be omitted. Power over Ethernet (PoE) not only eliminates the need for cumbersome wall transformers but also facilitates the rollout of a whole new suite of devices that combine data and power interfaces and are backward compatible with existing 10, 100, or 1,000 Mbps Ethernet equipment. IEEE 802.3af revolutionized Ethernet applications, primarily as a power delivery protocol rather than a data protocol. Power over Ethernet (PoE) Systems PoE connections are entirely controlled by the PSE, whose port voltage (VPORT) transmits link status to the PD. In a PoE system, the client device receiving power via existing Ethernet is called the Powered Device (PD). The device supplying power to the PD is called the Power Provider (PSE). Through each RJ-45 port, the power consumed by the PD is limited to 12.95W, and the PSE output is limited to 15.4W. Provided the Ethernet cable and physical layer (PHY) transformer are well balanced, each PD can expect a maximum continuous current of 350mA. Considering the voltage drop on CAT-5 Ethernet cables (maximum 100m), the IEEE 802.3af standard specifies different voltage ratings for PDs and PSEs. Longer cables result in significant voltage drops, forcing PSEs to output voltages higher than the typical 48VDC to deliver as much power as possible to the PD. Therefore, voltages up to 57VDC are common on Ethernet cables. PoE networks can be implemented using endpoint PSEs or intermediate PSEs . Endpoint PSEs (Power Supply Equipment): Endpoint PSEs integrate an Ethernet switch and power supply within the same device. That is, a PSE combines the IEEE 802.3af power supply functionality with the functionality of a Data Terminal Equipment (DTE), or the repeater functionality of current Ethernet switches and hubs. These PSEs are located at the other end of the Ethernet connection (i.e., the end of the network connection) and are called endpoints. These endpoints typically deliver power via data pairs (signal pairs), as these pairs are always connected to the PD. This type of Ethernet switch is sometimes referred to as having "online power" (see Figure 2). Endpoint PSEs are best suited for deployments of new network infrastructure. [align=center] Figure 2: In a PoE system connection between an endpoint power supply (PSE) and a powered device (PD), power is transmitted via signal pairs[/align] Intermediate PSE (Power Supply Equipment) An intermediate PSE is installed on the connection between the data switch and the PD. This type of PSE is called a midspan or middle span, and power can be injected into the network cable using an intermediate PSE. The intermediate PSE provides power through "free pairs" in the CAT-5 cable, as shown in Figure 3. The data pairs pass directly through; that is, for an intermediate-based network, the PD receives data from an existing non-802.3af switch and obtains power from the intermediate. This type of intermediate PSE is more cost-effective when only a few Ethernet devices require power. Such an example is typically a local area with 4 to 24 ports, which is part of a larger multiport network (Figure 3). [align=center]Figure 3: Simplified block diagram of a PoE power supply system design using the MAx5935 PSE controller and MAX5940 PD interface. This PoE power supply system can operate on Gigabit Ethernet. The PD must be backward compatible with the PSE application, therefore power must be received from an endpoint PSE switch.[/align] Differences between endpoint PSEs and intermediate PSEs The difference between an endpoint PSE and an intermediate PSE is that an endpoint PSE can choose to transmit power and signals simultaneously using signal pairs, or it can choose to transmit power using idle pairs. Generally speaking, a PSE must be able to provide power through either signal pairs or idle pairs, but not both simultaneously. Powered Device (PD) PD Detection When a powered device (PD) is connected to an Ethernet link, the PSE must detect whether each Ethernet device requires power. Therefore, the PD must exhibit characteristics different from traditional Ethernet devices. The IEEE 802.3af standard's PD requirements begin with a 25kΩ Φ20nF feature identification. This feature allows the PSE to detect devices requiring power by measuring their "detection characteristics"—common-mode termination—distinguishing them from other Ethernet devices that do not require power. A PD only needs to possess these detection characteristics, and its link must be in detection mode, to be detected. Table 3: Five levels of PD power rating and their rating characteristics . The PSE's specific method for PD detection: To achieve this detection, the PSE detects devices requiring power by measuring two VI (voltage-current) points and calculating the resistance from the slope between them to determine the port's common-mode termination. This utilizes a current-limiting voltage probe signal line from 2.7V to 10.1V. Table 1 lists the parameter conditions that the PSE must meet to detect a PD as valid in the detection state. The parameters in Table 1 allow a 1.9V series voltage offset because a diode bridge is typically used to control the voltage polarity. Each PD uses two such diode full-bridges (see Figure 4, a simplified design block diagram of a PoE power supply system using a MAX5935 PSE controller and a MAX5940 PD interface/controller), because the PD must be backward compatible with intermediate PSEs. The 10mA current offset is due to the inherent leakage within the PD. Additionally, Table 2 provides a series of other parameter conditions; any detection meeting these conditions will determine the Ethernet device as an invalid PD. [align=center]Figure 4: Schematic diagram of PSE power control application for simplification. The PSE power control silicon is MAX5935[/align] PD Power Classification Today, two main types of power devices are driving the growth of PoE: wireless LAN access points and VoIP phones, as well as other types of Ethernet devices (RFID readers, PDA chargers, mobile phones, and even laptops). The earliest impetus for the integration of power supply and Ethernet came from VoIP phones. The IEEE 802.3af standard also includes an optional feature called power classification. This feature allows the PSE to manage its power budget more precisely. The PSE uses a second measurement called classification to determine the peak power requirements of the PD. With this information, the PSE can supply power to devices that need it without damaging those that don't, and can effectively allocate available power. To implement this optional power grading method, the PSE applies a probe voltage between 14.5V and 20.5V. In response, the PD exhibits a certain characteristic (grading current) indicating to the PSE the maximum power that the PD will draw. A PD can consume up to 12.95W of power when receiving power (typically 48VDC). If the PD is not connected or is off, the PSE stops supplying power and continuously monitors the 25kΩ current of the active PD. The PSE manages the maximum power it delivers to connected PDs at any given time. Table 3 lists the different power levels that can be supplied to a PD and their corresponding grading characteristics. The grading current in the table is the current on the PD; the PSE current range is wider. For example, at level 2, a PSE must be able to recognize currents from 16mA to 21mA. By selecting a suitable PSE controller IC, other functions beyond the IEEE 802.3af standard can also be implemented: namely, a hard limit on the power output of the PSE to each port. Another crucial emergency function is the PSE's ability to prioritize power supply to ports, determining which ports should be powered first or disconnected when UPS or backup power is depleted. The switch can then maintain power to the most critical Ethernet ports. These ports might include E911 telephones, tag readers, certain surveillance cameras or access points, or operational data circuits. These fail-safe features are integrated into the PSE controller IC and can be configured via hardware wiring or software, aiding in power budget management during emergencies. Therefore, software-configurable PSE controller ICs should be sought. Detecting a disconnected PD : The PSE uses the "Sustain Power" feature to detect when the PSE powers on the PD, following IEEE 802.3af... The standard requires the PSE to monitor the PD's "hold power" characteristic. The PD must draw a minimum current of 10mA so that the PSE knows it is still connected. Power-sensitive applications like thermostats can reduce power consumption by pulse modulation to keep the "hold power characteristic (MPS)" current at 10mA, with the pulse interval between 75ms and 250ms. The PD must also have a common-mode impedance less than 26.25kΩ and a common-mode impedance greater than 50nF in parallel. Typically, the PD's bypass capacitor and the load will form a common-mode impedance greater than 26.25kΩ. The PSE also needs to detect whether the PD is disconnected . The IEEE 802.3af standard defines both AC and DC methods for detecting PD disconnection. For example, consider the scenario where the PD is disconnected and a legacy Ethernet device is immediately plugged into the same RJ-45 jack on the switch. If the 48VDC power supply is not immediately disconnected after the PD leaves, the legacy device will be damaged. AC impedance measurements are generally more accurate than pure DC resistance measurements. A small-amplitude common-mode AC voltage is simultaneously supplied to the Ethernet cable along with the data signal and 48VDC. The AC current is then measured and the port impedance is calculated. If the PD has not been disconnected, this value should be below 26.25kΩ and between 100MHz. For further details on DC and AC disconnection detection methods, designers should refer to the IEEE 802.3af standard. Regardless of the method chosen, measurements must be performed quickly, and power must be removed promptly after the PD is disconnected. Power Control Silicons for Power over Ethernet and Their Advanced Features Currently, some IC vendors are producing chips that meet 802.3af PSE requirements. Some of these solutions use microcontroller peripherals to provide the Power over Ethernet interface, but rely on controller software for most of the work. More powerful devices can autonomously detect and classify active PDs and manage overcurrent and disconnections with minimal software overhead. These devices may only require system software to determine whether there is sufficient power margin to meet the PD's power requirements. Current Multiport PSEs Silicon chips for PSEs are now readily available on the market, most commonly as PSE controllers that control four-port online power supplies. Devices with serial-compatible interfaces and programmable registers offer options for use with MCUs. Due to emergency considerations, advanced features across various operating modes are becoming increasingly important, and their importance is expanding. Figure 5 shows a simplified diagram of PSE power control using the MAX5935 controller chip, characterized by: four independent -48V PoE ports; and available operating modes including automatic, semi-automatic, manual, shutdown, and debug modes. Automatic mode allows the device to operate without software intervention. Semi-automatic mode (on-demand) continuously detects and hierarchically connects to devices on the ports, but does not power the ports until a software instruction is received. Manual mode allows complete software control of the device, which is very useful for system diagnostics. Shutdown mode terminates all activity and shuts off the power to the ports. Finally, debug mode allows fine-grained stepping of the device state machine for detailed system diagnostics. An application example of a PSE controller is shown in Figure 4, using the MAX5935 PSE controller and the MAX5940. This is a simplified block diagram of a PoE power supply system for a PD interface/controller. It shows the connection between the PSE and PD in Gigabit Ethernet. Since Gigabit Ethernet does not support intermediate power injection, 100/10M Ethernet can only be connected to one endpoint PSE switch. The MAX5940 PD interface controller can also add one or two as needed. The PD also requires a DC-DC converter to output +Vout -Vout. In summary, the IEEE 802.3af standard achieves power supply for Ethernet devices through Ethernet signal pairs or spare pairs, thus eliminating the hassle of using AC adapters and further expanding the application areas of Ethernet technology. Using the ubiquitous RJ-45 ports (hose), it not only provides data packets but also power. Because of these advantages, Power over Ethernet (PoE) has become a rapidly advancing new technology, fundamentally changing the way low-power devices are powered. Many more devices can be driven via PoE.