PoE Camera Not Powering: 2026 Fault Guide

A camera that won't power up usually comes down to one of four causes: an exhausted PoE budget on the switch, voltage lost over a long or thin cable run, a class mismatch between camera and switch (802.3af vs at vs bt), or a faulty port, injector or splitter.

These are the first four things we'd check on a "no power" call. Guessing at fixes wastes site visits, and non-standard or passive PoE, miswired cabling, or a faulty injector can damage equipment and affect warranties.

The fastest route to a working camera is to isolate which layer has failed — switch, cable, camera or power source — before you swap a single component.

What actually stops a PoE camera not powering on?

Most "no power" calls trace back to one of four faults: the switch has run out of PoE budget, the cable run is losing too much voltage to a gauge or length issue, there's a class mismatch between 802.3af, 802.3at and 802.3bt equipment, or one component — injector, splitter or port — has simply failed. This isn't a general explainer of how Power over Ethernet works; it's a triage sequence for a camera that's already dark.

The four faults to check first, in order of likelihood:

• PoE budget exhaustion — a switch has a total wattage pool shared across all ports. Add one camera too many, or one power-hungry pan/tilt model such as a PoE ColorVu camera with pan/tilt, and the last port to negotiate can be starved even though the LED shows link.
• Cable run and gauge loss — long runs, thin conductors, or damaged pairs increase resistance and drop voltage below what the camera needs to start. This is most likely near the practical limit of a Cat5e/Cat6 run.
• Class/standard mismatch — an 802.3at camera asking for more current than an 802.3af injector or switch port can supply will often fail to power up at all, rather than run in a degraded state.
• Faulty injector, splitter or port — a single dead PoE port, a failed injector, or a damaged splitter is indistinguishable from a cabling fault until it's swapped out.

Work through symptoms before swapping hardware. Isolate which layer is at fault — switch, cable, camera, or injector — by testing one variable at a time. Confirm suspected voltage or negotiation faults with a dedicated PoE tester or multimeter before condemning and replacing parts.

A handheld IP camera tester with PoE support, such as those in the RSRTENG IPC-5100 range, can show PoE voltage/class and power the camera directly for isolation testing.

Is it the switch's PoE power budget, not the camera or cable?

A camera that works on the bench through a standalone injector but stays dark on the switch has almost certainly hit the switch's PoE power budget ceiling. This is a power supply limit built into the switch itself, separate from cabling or connectors. It's one of the most overlooked causes of a "no power" fault — the Reolink support guidance on this exact symptom is worth reading if you're chasing it down.

The maths installers often skip over: an 8-port switch rated at a total PoE budget of, say, 130W cannot deliver 15.4W (802.3af) to all eight ports simultaneously. That would need 123.2W just for the cameras before switch overhead, leaving almost nothing spare. In practice, many installers find this by adding cameras one at a time until a port simply refuses to power up, even though earlier ports work fine.

The daisy-chain trap

A common pitfall is running an unmanaged PoE switch downstream of another PoE switch that's already close to its budget ceiling. Rather than failing cleanly, the shared budget can quietly starve whichever camera sits last in the chain. The fault then looks like a bad camera or cable, when it's actually total system demand exceeding what the upstream switch has left to give.

How to check it

Checking budget takes minutes and tells you whether the problem is the switch or the camera.

• Managed switch: log into the web interface and check per-port power draw and remaining PoE budget — this tells you immediately whether you're out of headroom.
• Unmanaged switch: there's no dashboard, so the only reliable test is unplugging another PoE device and seeing whether the problematic camera now powers up.

Could cable run, gauge or connections be dropping the voltage?

Yes — voltage drop along a long or poor-quality cable run is a common reason a PoE camera won't power up, and it gets noticeably worse past around 70-80m on thin-gauge cable. The camera and switch can both be fine; the copper in between simply isn't delivering enough voltage by the time it reaches the far end.

This shows up as a camera that never boots, or one that powers intermittently in cold or damp weather when resistance shifts slightly. Cable quality and cable length are usually the first two things worth ruling out.

A common UK installer pitfall is running long external cable using cheap patch-lead-grade cable. Thin stranded conductors are intended for short indoor data patching, not solid-core external-grade cable rated to carry sustained PoE current. It'll often pass a basic link-light and data test over a short bench run.

It can then fail once installed as a 60-90m external drop feeding a camera drawing real current, because stranded conductors have higher resistance per metre than solid copper. Genuine external-grade Cat5e or Cat6 cable with solid copper cores is built to carry PoE current reliably over distance, unlike thin multi-strand cable sold generically as "network cable" — worth checking against our PoE cameras guide before running a long external feed.

The quickest diagnostic is a straight swap test. Connect the same camera to the same switch port using a known-good short patch cable — under 5m — and see if it powers up cleanly.

If it does, the original run is the suspect. Measure its actual length, and inspect both RJ45 ends for damaged, loose, or oxidised crimp connections, particularly at any outdoor gland or junction point where moisture can get in.

Where the genuine run length is pushing towards or past that 70-80m mark, don't fight it with ever-thicker cable.

Is it a PoE standard/class mismatch (802.3af vs at vs bt)?

Yes — a switch or injector that only supplies 802.3af (15.4W max at the source, roughly 12.95W delivered at the camera end) can fail to power an 802.3at or 802.3bt camera. The result is a camera that won't power up, drops out, or power-cycles repeatedly. This is one of the most overlooked causes when a PoE camera not getting power symptom persists after cables and ports have been ruled out.

PoE devices don't just draw power — they negotiate for it. When a camera (the "powered device") connects, it signals its required power class to the switch or injector (the "power sourcing equipment") during a brief handshake. If the PSE only offers a lower class than the PD is asking for, the usual outcome is no power at all, not a reduced or "best effort" supply. That's why the fault often looks identical to a dead cable or faulty port.

Where this trips up UK installs

The classic mistake is treating "PoE is PoE" and mixing standards across a job. An 802.3af-only injector, rated for up to 15.4W, paired with an 802.3at or 802.3bt device — a PTZ camera or heater-equipped housing can draw well over 25W — will often fail to negotiate power reliably, even with a perfect cable run and connectors.

This is a recurring theme across troubleshooting threads on both Reolink's own support pages and independent guides such as CCTVinfo's fixes list.

Before pairing any camera with an older af-only injector, check the camera's actual rated power draw against the standard it needs. Our own [AT13F-A 4-Port PoE Extender (802.

Could the fault be a bad injector, splitter or port rather than the camera?

Yes — before condemning the camera, swap it onto a different switch port, a different injector, and a known-good short patch cable. Power delivery faults on the infrastructure side are common enough to check first, and a failed camera is only one of several possibilities.

Cross-substitution is the fastest way to isolate which link in the chain is actually faulty. It costs nothing but a few minutes.

A simple decision tree

• LED never lights, on any port, injector or cable — suspect the camera's PoE input circuit itself, or a genuinely dead unit. Test the same camera on a spare port and, if possible, a dedicated injector before writing it off.
• LED lights via a direct injector connection but not through the switch — this points to a switch-side fault: an exhausted PoE power budget, a damaged port, or a port capped below the class the camera needs. This is a pattern commonly reported in switch troubleshooting guides such as Thunder-link's coverage of PoE-from-switch failures.
• Power is intermittent — drops, reboots under load, flickering IR — check the cable run length, look for poor RJ45 crimps or damaged conductors, and check whether the injector's output sags under the camera's peak draw. This matters most with higher-draw devices such as a pan/tilt PoE camera with two-way audio and built-in white-light illumination, like our ANNKE PoE 3K ColorVu.
• NVR shows the camera as "offline" despite an active LED — treat this as a separate network-layer fault (IP conflict, PoE negotiation failure, or VLAN mismatch) rather than a power problem.