What is the difference between PoE PD and PoE PSE devices?

What is the difference between PoE PD and PoE PSE devices?

Power over Ethernet is often described as a single elegant trick: one cable, two jobs. But behind every smooth deployment—every ceiling-mounted access point, every security camera peering into the dark, every smart lock clicking open—there’s a carefully choreographed conversation between two fundamentally different types of hardware. One initiates, one responds. One sends power, the other accepts it. In the language of PoE, these are the PSE and the PD.

Understanding the difference between a PSE and a PD isn’t just a matter of memorizing acronyms. It’s what lets you read a network diagram and instantly see which boxes are driving the bus and which ones are along for the ride. It explains why you can’t just plug any PoE cable into any device without thinking. And it sits at the core of every PoE standard from the original 802.3af to the latest high-wattage 802.3bt.

**PSE: The One Who Supplies**

PSE stands for Power Sourcing Equipment. It’s the device that puts electrical power onto the Ethernet cable. If PoE were a water system, the PSE would be the pump at the reservoir. In a typical office network, the PSE is usually a PoE-enabled Ethernet switch, sitting in a wiring closet with rows of RJ45 ports, many or all of them capable of sourcing power.

But a PSE doesn’t have to be a full switch. A much simpler device—the PoE Injector—is also a PSE. An injector has two or three ports: one for data coming in from a non-PoE Switch, one that outputs combined data and power to the powered device, and sometimes a data-only passthrough. Whether it’s a 48-port managed switch or a palm-sized single-port injector, the core job is the same. The PSE injects DC voltage onto specific pairs of the twisted-pair cable, monitors the connection, and manages the power budget.

Modern PSEs don’t just blindly apply voltage. They actively look for a compliant PD at the far end. The sequence is precise: the PSE sends a low detection voltage, measures the signature resistance of the connected device, and only if that signature matches a valid PD profile does it proceed to classification and full power delivery. This handshake is what prevents a PoE Switch from frying a legacy laptop’s non-PoE Ethernet port. The PSE is constantly watching. If a PD is disconnected, the PSE removes power. If a PD suddenly draws more current than its negotiated class allows, the PSE can shut the port down. In a managed switch, the PSE also reports power consumption per port, allows remote power cycling, and enforces per-port power limits—features that become critical when you’re running dozens of cameras or access points across a campus.

**PD: The One Who Receives**

PD stands for Powered Device. It’s the endpoint that receives power from the PSE and uses it to run its own electronics. The PD is the consumer: the wireless access point, the IP camera, the VoIP phone, the digital signage display, the IoT gateway. It presents the correct signature during detection, tells the PSE how much power it needs during classification, and then accepts the incoming voltage.

Internally, the PD contains a PoE controller chip and a DC-DC converter. The incoming 44 to 57 volts from the PSE is too high for most electronics, so the PD steps it down to the 5V, 12V, or 3.3V rails required by its processors, radios, and sensors. This conversion generates heat and must be done efficiently, especially in sealed outdoor devices where thermal design is already tight. A well-designed PD also handles protection: it guards against surge events, filters noise from the Ethernet cable, and ensures that if the PSE and PD are from different generations, the system gracefully falls back to the highest mutually supported power level.

Not every device that uses PoE is a native PD. A non-PoE device can be made PoE-compatible with a PoE Splitter, which acts as a proxy PD. The splitter takes the PoE input, separates power and data, and outputs a lower DC voltage on a barrel jack while passing clean Ethernet data to the device. From the PSE’s perspective, the splitter is the PD—it’s the one doing the handshake and accepting the power.

**The Core Differences, Side by Side**

The fundamental difference is role: the PSE is the power source, the PD is the power sink. A PSE typically sources power on the spare pairs (Mode B) or the data pairs (Mode A) of the Ethernet cable, while a PD must be able to accept power on either set—this is a mandatory compatibility requirement in the standard. A PSE can power many PDs simultaneously, while a PD consumes power from exactly one PSE connection at a time.

The power path is unidirectional. A PD doesn’t send power back up the cable to the switch. This seems obvious, but it means a PD can’t power another device in a daisy chain unless it’s a special PoE passthrough device—and those are really a combination of a PD (receiving power) and a PSE (sourcing power on a downstream port) in one enclosure.

On the hardware level, PSE and PD chipsets are distinct silicon. A PoE Switch controller is designed to manage 48 or more simultaneous PD signatures, run per-port power monitoring, and handle fault protection across a wide voltage range. A PD controller is a much simpler, lower-cost chip focused on signature presentation, classification, and input power regulation. You won’t find a PSE chip inside a camera, and you won’t find a PD chip driving a switch port.

**Where the Standards Draw the Line**

The IEEE 802.3 PoE standards—af, at, and bt—define the electrical behavior of both PSE and PD for each power class. For example, 802.3af defines a PSE output of up to 15.4W and a PD input of up to 12.95W (accounting for cable loss). PoE+ (802.3at) moves those to 30W PSE and 25.5W PD. PoE++ (802.3bt) pushes into 60W and 90W PSE levels, with corresponding PD input budgets. At every step, the PSE always provides more power than the PD is guaranteed to receive, because the standard accounts for worst-case power dissipation in a 100-meter cable run.

A PSE and a PD from different manufacturers will work together if both adhere to the same IEEE standard. That interoperability is the whole point. The certification process ensures that a PSE correctly detects a PD’s signature, and that a PD doesn’t draw more power than its negotiated class allows.

**Why the Distinction Matters in Practice**

If you’re a network installer, knowing the PSE/PD split is what lets you size your infrastructure correctly. A PoE Switch (PSE) has a total power budget that must accommodate all connected PDs with margin. Overload that budget, and the switch will deny power to newly connected devices, even if individual ports are PoE-capable. You need to calculate the worst-case PD draw—not just the typical consumption, but startup inrush and peak loads under full IR illumination or maximum radio transmit power—and ensure the PSE budget covers it.

In troubleshooting, the PSE/PD distinction also provides a logical fault-isolation boundary. If a device won’t power up, you test whether the PSE port is delivering power with a known-good PD. If it is, the cable run or the original PD is suspect. If a whole bank of ports fails, the PSE’s internal power supply may be faulty. This clear separation of roles makes PoE deployments vastly easier to debug than proprietary power schemes.

**The Blurred Line: PoE Passthrough**

There is one device that deliberately blurs the PSE/PD boundary: the PoE passthrough switch or extender. This unit contains a PD that draws power from an upstream PSE, and an internal PSE that sources power on several downstream ports. It’s both a PD and a PSE at the same time, with the downstream power budget limited to whatever the upstream PSE allocates minus the device’s own consumption. In a network diagram, these are annotated as both PD and PSE, and they’re increasingly common in smart-building architectures where a single Ethernet drop needs to power a small cluster of sensors, a lock, and a controller.

**The Bottom Line**

The difference between a PoE PSE and a PoE PD is the difference between providing and consuming. The PSE—whether it’s a switch or an injector—is the intelligent power source that detects, classifies, powers, and protects. The PD is the endpoint that identifies itself, accepts the negotiated power, and converts it into the low-voltage rails that run its internal electronics. One initiates the conversation; the other responds. One manages a pool of watts; the other consumes exactly what it needs.

In a world where building infrastructure increasingly runs on Ethernet cables, knowing these two roles is foundational. When you see a spec sheet that says “PoE PD,” you know that device expects to be powered by the network. When you see “PoE PSE,” you know that device will send power out into the world. They’re two halves of a single standard, designed to work together seamlessly, and understanding their difference is what keeps your network powered, safe, and ready for whatever you plug in next.