A PoE camera that behaves all day and restarts as soon as night mode comes on often points to a load-related problem, but confirm the reboot first. Then compare the camera's maximum demand with the port limit and the switch's shared PoE budget. Work through the cable path one change at a time, using the same day-to-night transition after each change. Axis's December 2025 guidance explains why a daytime number can fool you. Typical consumption may leave out infrared illumination or a heater, while maximum power covers the more demanding operating state. Night mode gives you a repeatable trigger; it doesn't identify the failed part on its own. Axis power-consumption guidance This workflow is for CCTV installers and people responsible for small PoE networks. Arrange a maintenance window before disconnecting a camera. Work only from safe network access points, and don't open mains-powered equipment or swap hardware from an unsafe ladder position. Choose your starting point The camera's uptime resets The picture disappears but uptime continues Several cameras fail together You need to read a PoE tester result 1 Confirm a restart before testing the supply Check the camera's uptime before the failure and again as soon as the image returns. Uptime is simply the time since the device last started. Compare it with the camera's startup messages and the switch event log for the same minute. If uptime drops back to a small value, the camera really did restart. Look for a power event, scheduled reboot, firmware fault, temperature alarm, or watchdog recovery at that time. Save the logs before you change anything. If uptime keeps counting, use the IP camera online but NVR black screen workflow . A generic offline message from the recorder or a link-down entry from the switch isn't enough, by itself, to prove that power disappeared. When only the picture disappears Watch the camera's direct live view and the NVR during the failure. If direct video stays up, focus on the NVR's stream selection, credentials, decoding limit, and network path. If both views fail while uptime continues, capture the network and stream errors. Finish that branch before ordering a larger PoE switch. A video-retrieval fault and a camera restart need different evidence and usually lead to different repairs. The NF-859GT main unit and multifunction receiver in a Li Xiaofeng product demonstration. First separate a true camera restart from a video-path problem. Figure 1. Concept illustration separating a restart from video loss. Reset uptime points to a restart investigation; continuing uptime keeps attention on the video path. Correlate both observations with camera and switch logs. Hardware drawings are generic. 2 Check the port limit and the shared budget Start with the datasheet for the exact camera model. Note its required PoE type and maximum power, then check the original switch port's supported type, configured power limit, and current status. Use the switch manufacturer's allocation rules rather than comparing labels in isolation. Next, check the switch's total allocatable PoE budget and how much is already reserved. Don't mix allocated power with measured consumption: a switch may reserve more than a camera happens to draw at that moment. For example, a 120 W allocatable budget with 112 W already assigned leaves 8 W. You can't assume it will accept a camera that needs a 15 W allocation. Those figures are a teaching example, not a NOYAFA field measurement. The source rating is only part of the calculation. The camera must still receive enough power after cable losses, which Axis includes in system planning. A camera consumption figure and a switch's per-port rating are related, but they aren't automatically equivalent. Information Where to obtain it How to use it Required PoE type and maximum power Camera datasheet Check source compatibility under the stated conditions Port type and configured power limit Switch documentation and status Identify a restriction on the affected port Total budget and allocated power Switch PoE status Check whether shared allocation is exhausted Startup or power-denial event Camera and switch logs Match a restart to its recorded cause or trigger Voltage and supply indication Compatible PoE tester mode Check the source indication at that test point Cisco's Spanish-language troubleshooting guide separates budget, port, cable, and powered-device problems. Follow that sequence with the equivalent status pages on your switch; Cisco command syntax is platform-specific. Cisco troubleshooting method If several cameras restart together, look first for a shared event such as a switch reboot, common supply problem, schedule, or allocation change. If only one camera fails while its neighbors stay stable, continue down that camera's individual connection. 3 Identify the cable and read the PoE indication Before reading any tester result, prove which cable reaches which switch port. The NOYAFA NF-859GT supports port blinking as well as PoE and wire-map checks. Use port blinking to tie the physical run to the correct switch log; troubleshooting the neighboring port wastes time and can produce convincing but irrelevant evidence. Close view of the NF-859GT set. For PoE indication and wire-map checks, use only the ports and connection methods specified in the manual. For a PoE source check, use the interface and mode specified in the NF-859GT manual. Record the supply indication, energized pins, and displayed DC voltage, and note exactly how the tester was connected. A source reading taken with the camera removed isn't the same condition as a working camera under load. Figure 2. PoE screen excerpt from the NF-859GT section of the official manual, printed page 33, with added explanations. The displayed values are manual examples, not field measurements. The manual screen shows DC voltage in volts. It doesn't show the camera's night-mode demand in watts, so that number alone can't prove sustained available power or capture a brief dip during a restart. If the result suggests an unexpected supply type or polarity, check compatibility before reconnecting the camera. A non-standard supply indication is a reason to consult the equipment specifications, not permission to test an unknown injector on a camera. 4 Check the cable using the correct test arrangement For the receiver wire-map test described in the manual, disconnect the run from active equipment. Connect the main unit's line interface at one end and Remote Line A at the other, choose Line with Receiver, and run the test. That setup checks the conductor mapping between those two points. Write down an open, short, or incorrect map exactly as the screen reports it. Inspect accessible terminations or replace the suspect patch lead, then repeat the same test. A pass confirms the reported mapping; it doesn't certify cable performance or loaded PoE delivery. Fluke Networks explains why a negotiated PoE load test tells you more than a source indication: it checks delivery under a defined load. Use an appropriate load analyzer when the diagnosis needs that evidence. NF-859GT voltage and wire-map results shouldn't be presented as an equivalent load test. PoE load-testing explanation 5 Change one part of the connection at a time Save the camera configuration and note the original port, cable route, and power settings. Reproduce the same day-to-night transition through supported controls or the real operating condition. For the final check, keep infrared enabled if the installation relies on it. The NF-859GT main unit and receiver positioned separately for a repeatable comparison. Keep the trigger and observation period constant while changing one component. Figure 3. Original teaching plan for controlled component comparisons. Replacements must be compatible and independently known to work. Start with an accessible patch lead Leave the camera and switch port alone and replace one accessible patch lead with a compatible known-good lead. Watch the same night-mode trigger. If the restart stops, inspect the removed lead and its connectors, then repeat the observation before closing the case. If the camera still restarts, try a short known-good direct path on the same adequate port where access allows. Improvement points to the bypassed path as a group. It doesn't tell you which buried connector or section is responsible. Compare the switch port Keep the camera and cable path unchanged, then move to a known-good port with comparable settings and enough PoE capacity. If the symptom follows the original port, inspect that port's configuration and logs. A replacement port with different settings can't prove a hardware failure. Compare the camera Put a compatible known-good camera on the proven connection, ideally the same model. If the original camera still restarts there, keep its logs and investigate firmware, temperature limits, and service options. Save recovery information before considering a factory reset. 6 Verify the repair at the original trigger Restore the installed cable path and the intended recording profile. Observe the day-to-night transition with the relevant infrared, heater, and motion features active. During the same period, check uptime, switch events, live video, and recorded playback. Add the observation period to the result. Seeing no restart through an observed transition is useful evidence; a quick daytime preview isn't a verification of the repair. Record the comparison Field Original condition After one change Camera model and firmware Enter actual details Enter actual details Port and configured power limit Enter port and setting Enter port and setting Cable path or patch lead Identify the tested path Identify the changed part Trigger time and uptime Record time and uptime Record time and uptime Supply reading and test method Record units and connection Repeat comparable conditions Result and log event Describe the observed failure Describe the observed result If uptime never reset, move to the video-stream and recorder workflow instead. For the underlying terminology, see what Power over Ethernet means .