Power over Ethernet (PoE) is a method whereby power is transmitted to Ethernet-connected equipment (VoIP telephones, WLAN transmitters, security cameras) from the central switch. By using the existing CAT-5 cabling, the need for AC power (and wiring costs) can be eliminated. The switch is also able to control power distribution to the powered devices allowing sophisticated uninterruptible power management for vital systems.
Operation
Fundamentally, a PoE load or Powered Device (PD) must fulfill three functions in order to act in conjunction with the sending end Power Sourcing Equipment (PSE). The functions are Detection, Classification, and Under-voltage Lockout.
Detection Phase
When a PoE-enabled Ethernet cable is plugged into a PD, the PSE interrogates the device to determine if it is PoE-enabled. This period is termed the detection phase. During the detection phase, the PSE applies a voltage ramp to the PD and looks for a characteristic impedance from the load (25 kO). If the correct impedance is not detected, the PSE assumes that the load is not PoE-enabled and shuts down the PoE sending end. The system then operates as a standard Ethernet connection. If the signature impedance is detected, the PSE moves on to the classification phase. The signature identification voltage is a ramp voltage between 2.5 V and 10 V. A 24.9 kO resistor provides the correct signature impedance for detection (see Figure 1).
 |  | |
| Figure 1. Detection Impedance. | Figure 2. Classification Current (Class 0). |
Classification Phase
The PSE continues to ramp the voltage to the PD. Between 15 V and 20 V, the classification phase occurs. During this voltage transition, the PD must draw a specified current to identify the device class (see Figure 2). The simplest class (Class 0) is also implemented by the use of the 24.9 kOhm signature resistor. The classification current describes the amount of power the PD will require during normal operation. It is this information that is fed to the controller by the PSE, which allows the system to determine power budget requirements. A table of classification current and operating PD power requirements is shown in Table 1.
| Class | PMIN | PMAX | ICLASS (MIN) | ICLASS (MAX) | RCLASS |
|---|---|---|---|---|---|
| 0 | 0.44 W | 12.95 W | 0 mA | 4 mA | Open |
| 1 | 0.44 W | 3.84 W | 9 mA | 12 mA | 150 Ohm |
| 2 | 3.84 W | 6.49 W | 17 mA | 20 mA | 82.5 Ohm |
| 3 | 6.49 W | 12.95 W | 26 mA | 30 mA | 53.6 Ohm |
| 4 | Reserved | Reserved | 36 mA | 44 mA | 38.3 Ohm |
Turn on phase
After the classification phase, the PSE continues to ramp the input voltage up to 30 V, when the under-voltage lockout (UVLO) circuit is released and the PD is allowed to power up. Soft-start circuitry is required to limit the current drawn from the PSE. A typical under-voltage lockout circuit is shown in Figure 3.
Figure 3. PoE Class 0 Interface Circuit Using a MOSFET Pass-Switch.
By this process, the PSE and PD work together to determine the nature of the load and apply power only to PoE enabled equipment. The system controller at the central location can determine load requirements and allocate power according to an operational needs hierarchy during power failure from its available UPS budget.
For additional information about driving PoE compatible load equipment and circuits for implementing Class 1 through Class 3 classification, see Design Ideas DI-70 and DI-88.