Camera Drops Failing After a Long PoE Run
You are commissioning eight new IP cameras on a single-story retail building. The farthest camera is roughly 320 ft of Cat6 from the PoE switch. Six cameras come up fine. Two cameras - including the farthest one - power up intermittently, drop video, and reboot when the exterior lights turn on at dusk. The switch reports PoE budget warnings. The customer's IT vendor terminated the runs with RJ45 plugs in the field rather than jacks, and one cable was run parallel to a 120 V lighting circuit for about 40 ft. The cameras are rated 802.3af (Class 3). The switch is 802.3at capable with a 240 W total budget.
1. List the most likely causes of the intermittent camera failures and explain how each produces the observed symptoms.
Read the worked answer
Likely causes, in order of probability:
- Field-terminated RJ45 plugs - plugs on solid Cat6 are unreliable; poor crimps, untrimmed conductors, or wrong plug type cause high resistance and intermittent contact. This alone explains dropouts and reboots.
- PoE voltage drop over 320 ft - long runs plus marginal terminations raise loop resistance. The camera may brown out (reboot) when current demand peaks, e.g., IR LEDs at dusk.
- Parallel run with 120 V lighting - 40 ft of proximity to a switching lighting circuit introduces EMI, and the dusk correlation strongly suggests the lighting circuit is the trigger (inrush/switch noise coupling onto the data pairs).
- PoE budget warnings - the switch is close to its 240 W total; if other loads are present, the two failing ports may be getting insufficient power allocation, or the switch is negotiating at the edge of the af/at boundary.
Dusk correlation is the key diagnostic clue: it links the failures to the lighting circuit, not random cable failure.
2. Describe a systematic troubleshooting sequence to isolate the root cause(s) before replacing hardware.
Read the worked answer
Work from the physical layer up, changing one variable at a time:
- Baseline the switch: read per-port PoE allocation, power draw, and error counters. Confirm the two failing ports' negotiated class and whether the budget is being clipped.
- Cable test: run a certified cable tester (length, wire map, NEXT/return loss) on the two failing runs. Look for high resistance, split pairs, or marginal length.
- Re-terminate: replace field plugs with proper jacks/patch panels or correct plugs for the cable type; retest.
- Isolate EMI: temporarily reroute or shield the 40 ft parallel section (or move the camera to a short test cable) and observe whether dusk failures stop.
- Substitute: swap a known-good camera and a known-good short cable onto the failing port to rule out camera vs. cable vs. switch port.
- Power test: measure voltage at the camera under load; if it sags, address voltage drop (shorter run, higher PoE class, or local injector).
Document readings at each step so the fix is evidence-based, not guesswork.
3. The customer asks you to 'just bump those two ports to 802.3at and move on.' Evaluate that request.
Read the worked answer
Bumping to 802.3at does not fix the real problems and may make things worse:
- The cameras are af/Class 3 devices; forcing at power doesn't correct a high-resistance termination or EMI. It may mask a voltage-drop symptom temporarily while leaving a latent fault.
- It consumes more of the 240 W budget, worsening the budget warnings on other ports.
- It leaves the dusk/EMI coupling and the field-plug reliability issues unaddressed - failures will recur.
Correct approach: fix the terminations, address the parallel power run (reroute or separate), verify voltage at the camera under load, and manage PoE budget deliberately. Only after the physical issues are resolved should you confirm each camera receives adequate power within the switch budget. Explain to the customer that 'bumping the port' treats a symptom, not the cause.