Power over Ethernet · workflow

Plan PoE from the endpoint backward.

Name the power system and device requirement first. Then choose the PSE, check one-port power, check total power, and test the physical path.

1. Identify the power system

IEEE 802.3 PoE uses PSE detection and classification before normal power is applied. Proprietary passive PoE can place voltage on conductors without that standards-based detection. A label that says only “PoE” does not prove which system is present.

For an IEEE PD, record its Type or Class and maximum input requirement. For a passive endpoint, record exact voltage range, current, powered pairset, polarity, connector use, and the manufacturer’s explicit compatible source.

Never use a passive injector as an experiment. Unknown mode, voltage, pairset, or polarity is a stop condition, not a reasonable default.

2. Choose the supply-device class

Use a PoE switch when several IEEE PDs need network data and power. Use an IEEE injector when a non-PoE Ethernet source must power one compatible PD. Use a splitter only when it explicitly converts compatible PoE into the exact DC input a non-PoE endpoint requires. Extenders and PoE-powered switches consume part of the incoming budget.

Select the equipment class.

3. Check the port before the whole switch

Per-port capability and total budget are separate limits. The PoE Type constrains the standardized Class range. The exact port configuration or allocation must cover the PD, even when the chassis advertises a large total wattage.

Check the PSE/PD class pair.

4. Check all simultaneous loads

Add device groups with one consistent policy. A conservative design may use allocated class power. A documented fleet policy might use exact maximum input values plus explicit reserve. Do not mix class maximums, measured idle draw, and marketing “typical” figures without naming the assumption.

Keep reserve for startup behavior, illuminators, heaters, load changes, power-supply redundancy, and environmental derating. Then verify every port separately.

Plan the total switch budget.

5. Treat every powered intermediate as a consumer

A PoE-powered switch receives power as a PD and supplies power as a PSE only if its documentation says so. Its own consumption and reserve come out before downstream output. The declared downstream budget may change with the input Type or when the device uses DC instead.

Plan pass-through power.

6. Verify the cable channel

Power fit does not certify data performance or installation quality. Preserve pair twists, use components suitable for the application and environment, and test the completed channel. High-resistance or damaged terminations can create voltage loss and heat even when continuity exists. Long or exposed outdoor paths introduce separate surge, bonding, grounding, and code requirements.

7. Troubleshoot with one-variable controls

  1. Confirm the exact PSE and PD modes are compatible.
  2. Check switch detection/classification and port logs when available.
  3. Try one short known-good cable.
  4. Try one known-good compatible PSE port or injector.
  5. Reduce total PoE load without changing the endpoint.
  6. Compare startup, steady state, and temperature-related behavior.

Rank the next controlled test.

What these tools cannot observe

A browser cannot measure pair voltage, current, resistance, temperature, classification events, LLDP negotiation, cable certification, or live switch policy. Treat results as declared-capability screens and arithmetic, then verify the exact hardware and installed channel.

Review sources and class boundaries.