Cable Issues: Type, Attenuation, Crosstalk, Termination, and TX/RX
Key Takeaways
- Layer 1 troubleshooting starts with media type, distance, connectors, pinout, bend radius, and link indicators before any logical check.
- Incorrect cable type or category rating can cause no link, a down-negotiated speed, errors, or intermittent connectivity.
- Attenuation is signal loss over distance; crosstalk is unwanted coupling between pairs or adjacent cables, measured as NEXT and FEXT.
- Bad terminations, split pairs, damaged connectors, and poorly seated patch cords commonly create CRC errors and flapping links.
- Transposed transmit and receive pairs or fiber strands prevent a link unless devices support Auto-MDI-X or auto-polarity correction.
Cable Issues and Layer 1 Symptoms
Physical problems are often simple, but they masquerade as complex network failures. A bad patch cord produces application timeouts. A run that exceeds the 100-meter copper channel limit (90 m horizontal plus 10 m of patch cords) appears as random packet loss. A fiber pair with transmit (TX) and receive (RX) reversed looks like a dead switch or server because the link never establishes. Domain 5 carries 24% of the N10-009 exam weight, so this is the single largest scored troubleshooting area on a test of up to 90 questions, 90 minutes, passing at 720 on a 100-900 scale.
Always start with the evidence closest to the signal: link lights, interface status, negotiated speed and duplex, error counters, cable-test results, optical levels, and the physical path. CompTIA's troubleshooting methodology runs top-down from identifying the problem, but technicians work bottom-up from Layer 1 because the cheapest fixes live there.
Common Cable Faults
| Issue | Typical symptom | Useful check |
|---|---|---|
| Incorrect cable type | No link, wrong speed, intermittent link | Verify copper category, fiber mode, connector, transceiver support |
| Attenuation | Errors or drops on longer runs | Distance, cable rating, optical RX level, excessive patching |
| Crosstalk (NEXT/FEXT) | CRC errors, retransmits, poor throughput | Test pairs, inspect bundling, separation, termination quality |
| Bad termination | Link flaps, one pair fails, autoneg problems | Cable certifier or wire-map tester |
| TX/RX transposed | No link on fiber or legacy copper | Swap strands, verify polarity or Auto-MDI-X |
| Damaged connector | Link works only when cable is moved | Inspect latch, pins, ferrule, dust cap, seating |
Incorrect Type or Rating
The media must match the job. Cat 5e supports 1GBASE-T to 100 m; Cat 6 adds headroom and supports 10GBASE-T to roughly 55 m; Cat 6A carries 10GBASE-T the full 100 m. The intermediate rates 2.5GBASE-T and 5GBASE-T (NBASE-T) were designed to run 10G-class speeds over existing Cat 5e/6 to feed Wi-Fi 6/6E access points. Fiber adds its own rules: single-mode (OS2, 9 micron core, yellow jacket) and multimode (OM3/OM4/OM5, 50 micron core, aqua/lime jacket) are not interchangeable, and the optic must match wavelength, connector, speed, and distance.
| Scenario clue | Likely direction |
|---|---|
| Negotiates 100 Mbps instead of 1 Gbps | One of the four pairs is broken; 1000BASE-T needs all 4 |
| Short patch works, long run fails | Attenuation, distance over 100 m, termination, certification |
| New fiber link stays down | Fiber mode, optic type, polarity, optical power |
| PoE device reboots under load | Cable quality, PoE budget, pair condition, power class |
Attenuation and Crosstalk
Attenuation is signal-strength loss across distance, measured in decibels (dB). Excess loss creates frame errors, low optical RX power, or a link that trains only at a reduced rate. Crosstalk is unwanted coupling between conductors. NEXT (Near-End Crosstalk) is measured at the transmitting end; FEXT (Far-End Crosstalk) at the far end. Untwisting more than the 0.5 inch limit at a jack, crushing cable under furniture, or running copper alongside fluorescent ballasts and motors all raise crosstalk and the error rate.
Bad Termination, Pinout, and TX/RX
Ethernet depends on pairs, not individual wires, so a split pair can pass a simple continuity test yet fail at gigabit speeds because two wires from different pairs were terminated together. Use a wire-map certifier when symptoms are intermittent or a run must meet a category rating. Fiber termination faults include dirty ferrules, mismatched connectors, bend radius below the typical 10x cable diameter minimum, and damaged patch cords. Most modern copper ports support Auto-MDI-X, so straight-through versus crossover rarely matters, but fiber polarity still does: two strands connected TX-to-TX and RX-to-RX will never link.
Worked Example: The Intermittent Drop
A user reports that their workstation "loses the network" a few times an hour, but ping to the gateway works when you test it. On the switch you see the port flapping (up/down repeatedly) and a slowly climbing CRC counter. You swap the user's patch cord for a known-good one and the flapping stops, but CRC errors continue at the same slow rate. That tells you the patch cord was one fault and the permanent run is a second fault. You certify the horizontal run and find a split pair introduced when the jack was punched down with two conductors from different pairs swapped. Re-terminating to the correct T568B pinout clears the CRC errors.
The lesson: a single symptom can hide two independent Layer 1 problems, and CRC errors that survive a cord swap live in the permanent cabling, not the patch cord.
Practical Troubleshooting Flow
- Confirm symptom and scope to separate one bad drop from a switch-wide outage; ask whether other users on the same switch or VLAN are affected.
- Check link status and negotiated speed to confirm Layer 1 trained at the expected rate rather than a fallback rate.
- Swap known-good patch cords and ports to isolate endpoint and patch faults in seconds before opening a certifier case.
- Inspect media, connectors, and labels for wrong type, physical damage, dirty fiber, kinked jackets, or a mislabeled path.
- Run a cable certifier or measure optical levels to capture attenuation, NEXT/FEXT, pair faults, length over 100 m, or polarity evidence.
- Document the fix to prevent repeat troubleshooting and keep the cabling map accurate for the next technician.
Common Traps
Technicians waste hours when they assume the cable is fine because the link light is green. A green light only proves a link trained at some rate, not that it trained at the right rate or that it passes clean frames. Two more traps: blaming a NIC or switch for a problem that swapping a $3 patch cord would have proven, and clearing error counters in the middle of an investigation so you lose the trend evidence. Always observe whether counters climb under live load before clearing them, and only clear when local change practice allows.
Exam Focus
For N10-009 Domain 5, match the symptom to the layer. No link, link flapping, CRC errors, unexpected speed, low optical RX power, and polarity faults all point to Layer 1 before you touch routing, DHCP, or application config. When a question offers both a physical and a logical cause for a symptom that includes errors or a dead link, the physical cause is almost always the intended answer.
A new single-mode fiber connection between two switches does not come up. The optics match the required speed and fiber type. What should be checked first?
A 1000BASE-T copper run links up but negotiates only 100 Mbps after a wall jack was reterminated. What is the most likely cabling problem?
Which issues are physical cabling concerns? Select three.
Select all that apply