
01What paired LFP is designed to solve
A media converter can keep its local RJ45 link up even after another segment of the end-to-end path has failed. The connected switch or PLC may therefore continue to treat the path as available. Link Fault Pass-through associates selected copper and optical link states so that a nearby device can see a physical failure and raise an alarm or start a higher-level recovery action.
LFP does not repair fiber, test application traffic, or replace a redundancy protocol. A frozen PLC, incorrect VLAN, duplicate IP address, or blocked service can leave every physical port up. Diagnose LFP with link LEDs, switch logs, optical measurements, and one fault at a time.

02Direction 1: how a copper fault reaches the fiber side
Disconnecting Device A’s RJ45 cable first produces LINK DOWN at its copper interface. If the converter supports copper-to-fiber propagation, it may stop the associated optical transmitter, alter the optical link state, or use a supported FEF mechanism. Device B then detects RX loss or the remote fault and can set its own RJ45 port to DOWN, allowing the attached switch or PLC to see the failure.
Do not infer this direction from the words “LFP enabled” alone. Some products only pass an optical alarm to copper. FCTEL’s currently referenced product page explicitly documents optical-port alarm transmission to the copper port; that page alone does not prove reverse copper-to-fiber behavior. Confirm the exact model manual and verify it by disconnecting Device A’s copper cable.

03Direction 2: how an optical fault returns to copper
When the A–B fiber link is opened, or one converter loses receive light, the optical interface detects the fault first. A converter that supports fiber-to-copper LFP passes that state to its local RJ45 interface. If both ends lose receive light and both converters use the same function, both copper ports may go DOWN.
A single broken strand in a duplex link can be directional: one end loses RX while the other may still receive light. Whether FEF reports that one-way failure to the peer depends on the implementation. Single-fiber bidirectional products have different alarm behavior. Check TX/RX polarity, wavelength pairing, optical power, fiber mode, and the exact product logic.

04Validate both directions with a four-fault matrix
Run at least four controlled tests: disconnect Device A copper to verify copper-to-fiber behavior; open the A–B fiber to verify fiber-to-copper behavior; disconnect Device B copper to test the reverse direction; and remove power from Device B to confirm how remote power loss appears at Device A. Introduce only one fault at a time.
For every test, record Device A copper, Device A optical, Device B optical, Device B copper, upstream port state, switch logs, alarms, service interruption, and recovery time. “May go DOWN” and “implementation dependent” are test items, not conclusions. After restoring the link, check for port flapping and error-counter growth.
05Coordinate LFP with redundancy and monitoring
LFP can give an upstream switch a clearer physical event, while STP, ERPS, routing, controller redundancy, or the application still decides whether traffic moves to another path. Test the combined convergence behavior so overlapping detection mechanisms do not amplify each other and repeatedly switch the port.
Keep three alarm layers distinct: link DOWN for a physical segment, device offline for power or management loss, and application health failure for a service that stops while the link remains up. Together they help technicians distinguish broken fiber, converter power loss, loose copper, and configuration errors.
06Common mistakes and maintenance records
Common mistakes include changing only one converter, moving a DIP switch while energized against the manual, failing to record the original position, treating LFP as a ring protocol, and replacing fiber immediately after any port-down event. Read the model instructions first, change one variable at a time, and compare every result with the healthy baseline.
Acceptance record: converter models and revisions, both DIP positions, fiber and copper identifiers, the four-fault matrix, switch logs, alarm recipients, recovery times, and rollback steps. Repeat the tests after replacing either converter, changing an optical module or switch port, moving the fiber, or modifying the redundancy design.


