Skip to Content

Learn About Power over Ethernet (PoE)

Introduction to Power over Ethernet (PoE) Inrush, Power Up and Power Down

Last modified by Microchip on 2025/06/26 12:30
   Learn Power over Ethernet (PoE) Dual Signature Classifications  Learn About New 802.3bt Features   

Introduction

Power over Ethernet (PoE) inrush is a critical process that ensures safe and reliable delivery of power from Power Sourcing Equipment (PSE) to a Powered Device (PD) over Ethernet cabling. Inrush refers to the controlled application of power that occurs immediately after classification, which determines the power requirements of the connected device. As the final step before the PSE fully applies power to the PD, the inrush process is carefully managed to prevent electrical surges that could damage equipment or disrupt network operation. This process is typically divided into three distinct phases:

  • PSE inrush, where initial current is limited to safely charge the PD’s input capacitance,
  • Power delay, a brief pause to ensure stability, and
  • Power applied, after which full operating power has been delivered to the PD.

By managing these phases, PoE inrush protects both the network infrastructure and connected devices, ensuring efficient and safe power delivery.

Power over Ethernet Inrush Power-Up Power-Down

Back to Top

Inrush

Power over Ethernet (PoE) PD Inrush Time Phases

Power over Ethernet (PoE) PSE Total Inrush CurrentPoE inrush is a carefully managed process that controls how power is initially delivered from PSE to the PD to ensure safe and reliable startup. The inrush sequence is divided into three distinct phases. The first phase, known as PSE Inrush (1), begins when the voltage at the PD reaches between 30 and 42 volts, a point referred to as T0. During this phase, the PSE limits the inrush current (IINRUSH) according to the assigned PoE class; for Classes 1 to 4, the current is limited to 450 mA, while for Classes 5 to 8, it is limited to 600 mA. Additionally, current may be limited per pairset, depending on the PSE’s design. The PD itself has no specific requirements during this initial phase, as it is typically charging its input bulk capacitor.

The second phase, called Power Delay (2), starts 50 ms after T0 and lasts for 30 ms. During this interval, the PSE may transition to the Power Applied phase at any time. In the Power Delay phase, the PD must limit its power draw to less than Class 3 power (typically under 13W at the PD input) and must also comply with total and per-pairset current limits. The PD needs to keep its main load disabled during this phase to ensure that its input bulk capacitor is fully charged before normal operation begins, which helps prevent excessive inrush current and potential voltage dips.

The final phase, Power Applied (3), occurs 80 ms after T0. At this point, the PD is allowed to switch to its full operational power as defined by its assigned class.

Back to Top

Power Up

Following successful detection and classification, a system enters the nominal powering state, where it spends most of its time. During this time, the PSE checks for abnormal conditions on each pairset, like overloads or short-circuits, as well as checking that the PD remains connected. As a result of classification, the PD will either have been assigned to the Class it requested or have been power demoted. In either case, both the PSE and the PD are required to conform to the requirements of the assigned Class per pairset.

In the powered state, a single-signature PD is powered as a single-load device, where each pairset is monitored for overload or short circuit independently, and the PSE continually checks that the PD remains connected. A dual-signature PD is powered per pairset independently (regardless if it is a dual-signature PD with as single load or dual load), where each pairset is monitored for overload.

Back to Top

Power Down

The PoE power removal process is designed to ensure the safe and efficient disconnection of power from a PD. This process can occur when a connection is switched from a powered PD to an unpowered PD or when a PD is physically disconnected from the network. According to the IEEE® 802.3af standard, the PSE is required to continuously monitor the connected device for its unique power signature, which indicates the presence of a valid PD. If the PSE stops receiving this power signature, such as when the PD is disconnected or replaced with a non-PoE device, it interprets this as a signal to remove power from the port. To prevent potential damage to the equipment and avoid unnecessary power cycling due to brief or accidental disconnections, the PSE waits for a short period, typically 250 ms, after detecting the loss of the PD’s signature before cutting off power. After power is removed, the PSE may periodically re-initiate the detection process to determine if a new or replacement PD has been connected, ensuring continued compliance with safety and operational standards.

Back to Top

Summary

The three-phase inrush process and power disconnection, as defined by IEEE 802.3 standards, is essential for protecting both the PSE and PD, ensuring system stability, and preventing damage to network equipment and cabling during power-up. Both stages are critical aspects of PoE system design and compliance.

Back to Top

Learn More

Back to Top