Physical-layer questions often hide several compatibility checks inside one sentence. A cable can fit physically and still be wrong for the distance. A transceiver can fit the slot and still use the wrong wavelength or medium. A link can come up and still accumulate errors.
The full Domain 1 and Domain 5 guides teach the media catalog and troubleshooting process. This Quick Review is for deciding which physical detail matters first.
Fast rule: For a wired-link problem, check compatibility before replacing hardware: medium, distance, connector, transceiver form factor, speed or protocol, wavelength, polarity, received signal, and power requirements.
1. Physical-layer decision map
| Question | Start here | Strong clue |
|---|---|---|
| Which medium fits? | Distance, bandwidth, interference, environment | Long campus run or electrically noisy path points toward fiber |
| Why is there no link? | Cable, connector, optic, polarity, interface state | Compatible devices but no optical link can point to TX/RX or optic mismatch |
| Why is the link unreliable? | Counters, termination, attenuation, interference | Increasing frame errors under load point toward a physical or interface problem |
| Why does data work but the AP stays off? | Power over Ethernet | Check port power state, switch budget, endpoint requirement, and cable pairs |
| Which optic belongs here? | Slot, protocol or speed, wavelength, fiber, distance | Matching form factor alone is not enough |
2. Choose among copper, fiber, and direct attach copper
The medium should fit the path, not just the port.
| Medium | Best-fit clue | Main concern |
|---|---|---|
| Twisted-pair copper | Office access link, short horizontal run, endpoint may need PoE | Category, distance, termination, interference, and pair condition |
| Single-mode fiber | Long-distance, campus, provider, or high-capacity path | Correct single-mode optics, wavelength, cleanliness, loss, and polarity |
| Multimode fiber | Shorter building or data-center fiber link | Correct multimode optic, distance, connector condition, and polarity |
| Direct Attach Copper (DAC) | Very short switch-to-server or switch-to-switch link | Limited reach and device compatibility |
| Coaxial | Cable broadband, antenna, video, radio, or legacy use | Connector and impedance must fit the application |
Copper is often the practical access-layer choice because it is inexpensive and can deliver power. Fiber is valuable when distance, bandwidth, electrical isolation, or electromagnetic interference makes copper a poor fit. DAC is attractive for short equipment links where separate optics and patch fiber would add cost without adding useful reach.
Do not treat fiber as automatically better. A 2-meter device connection does not benefit from a long-reach optic simply because the optic sounds more capable.
3. Make copper cabling decisions from the requirement
For the common copper categories emphasized in the study guide:
- Category 5e commonly supports 1 Gb Ethernet to the normal 100-meter channel limit.
- Category 6 supports 1 Gb over the full channel and can support 10 Gb on shorter runs.
- Category 6A is the common copper choice for 10 Gb over a full 100-meter channel.
The exam clue is usually the required speed and distance, not which category number is largest.
A higher cable category does not repair:
- Poor termination
- Excessive distance
- Damaged conductors
- Sharp bends
- Crosstalk or outside interference
- Bad patch cords
- An interface configured for the wrong speed or duplex
Unshielded twisted pair (UTP) depends on pair twists and good installation practices to control interference. Shielded twisted pair (STP) adds shielding, but the channel must use compatible components and appropriate grounding practices. Installing one shielded segment inside an otherwise unsuitable channel does not magically eliminate interference.
Plenum-rated cable is chosen for fire and smoke characteristics in applicable air-handling spaces. Plenum does not mean faster Ethernet.
4. Separate single-mode from multimode decisions
Single-mode and multimode fiber are different media choices, not interchangeable labels for the same cable.
| Question | Single-mode clue | Multimode clue |
|---|---|---|
| Typical deployment | Long campus, provider, backbone, long-reach link | Shorter building or data-center link |
| Core behavior | Small core supports one propagation mode | Larger core supports multiple propagation modes |
| Optic choice | Use an optic designed for the single-mode link | Use an optic designed for the multimode link |
| Failure clue | Wrong optic, wavelength, polarity, contamination, or excessive loss | Wrong optic, distance, polarity, contamination, or excessive loss |
The safe Network+ decision is to match the transceiver and fiber type. Do not assume a physically compatible connector proves the optical system is compatible.
Fiber connectors also need cleanliness. A connector can look attached and still have enough contamination to raise loss or destabilize the link.
5. Recognize connector clues without overthinking them
| Connector | Medium | Recognition clue |
|---|---|---|
| RJ45 | Twisted-pair copper | Common Ethernet endpoint and switch connection |
| LC | Fiber | Small connector commonly used with modern pluggable optics |
| SC | Fiber | Larger square push-pull connector |
| ST | Fiber | Round bayonet-style connector associated with older installations |
| MPO | Multi-fiber | Dense parallel optical links |
| F-type | Coaxial | Threaded connector common in cable broadband |
| BNC | Coaxial | Bayonet connector used in selected radio, video, and legacy applications |
Connector questions are often elimination questions. First identify the medium. Then choose a connector that belongs to that medium and application.
A connector adapter can solve a mechanical mismatch, but it does not change fiber type, wavelength, protocol, speed, or the optical power budget.
6. Treat transceiver selection as a compatibility stack
A transceiver is not correct just because it slides into the slot.
Check these in order:
- Form factor: Does the module fit the device slot?
- Protocol and speed: Does the module support the required link type and rate?
- Medium: Is it intended for copper, single-mode fiber, multimode fiber, or another supported medium?
- Wavelength: For optical links, do both ends use a compatible optical specification?
- Distance: Is the module designed for the actual link length and loss?
- Connector: Does the patching match the transceiver interface?
- Peer: Does the far end use compatible link settings and optics?
Small Form-factor Pluggable (SFP) and Quad Small Form-factor Pluggable (QSFP) are transceiver families. They do not identify one universal speed, wavelength, protocol, or reach.
Exam clue: “Both modules are SFP” proves form-factor compatibility, not link compatibility.
If a replacement optic fits but the link stays down, compare the old and new module labels for speed, medium, wavelength, and reach before replacing the cable or switch.
7. Use polarity and signal strength to explain optical failures
A duplex fiber link needs a transmit path in each direction. The transmitter on one end must reach the receiver on the other.
If the components are otherwise compatible but the link has no light or never comes up:
- Verify the transmit and receive strands.
- Inspect and clean connectors.
- Confirm the correct patch fiber.
- Confirm optic type and wavelength.
- Check received optical signal strength.
- Check for excessive loss, bends, damage, or a bad splice or connector.
A visual fault locator can help reveal selected breaks or severe faults. An optical power meter answers a different question: how much optical signal is actually being received. The Network Troubleshooting Tools Quick Reference covers tool selection in more detail.
Do not infer fiber health from link status alone. A marginal link can remain up while errors or flaps increase as the optical budget worsens.
8. Separate data connectivity from Power over Ethernet
Power over Ethernet (PoE) can fail even when the cable still carries data.
When an access point, camera, or phone will not power up, check:
- Whether the switchport provides PoE.
- Whether the endpoint's power requirement is compatible.
- Whether the switch has enough remaining power budget.
- Whether PoE is enabled on the port.
- Whether the cable and required pairs are intact.
- Whether an intermediate patch panel, injector, or other component supports the intended design.
A basic continuity or data test does not prove that the endpoint is receiving adequate power.
Scenario clue: If a separately powered device passes traffic on the same cable but a PoE endpoint stays off, move power delivery much higher on the theory list.
Do not solve a power-budget problem by changing the IP address. Keep the failing dependency in view.
9. Read interface evidence before replacing the medium
A physical problem should produce physical evidence.
| Evidence | What it suggests | Next comparison |
|---|---|---|
| Link down | No usable physical connection or interface unavailable | Cable, optic, polarity, port state, peer state |
| CRC errors increasing | Frames are arriving corrupted | Termination, cable condition, interference, optic signal, speed or duplex |
| Runts or giants | Unexpected frame sizes or physical/interface problems | Counters, duplex, MTU context, NIC or interface behavior |
| Drops | Traffic was discarded | Queue or congestion evidence before blaming the cable |
| Link flaps | Intermittent physical, power, optic, or negotiation problem | Event timing, signal level, connectors, power, peer logs |
| PoE fault | Power negotiation, budget, standard, cable, or endpoint issue | Per-port power state and switch budget |
One old error counter is weaker evidence than a counter that continues to increase while the user reproduces the problem. Establish timing and trend before replacing components.
10. Scenario comparisons
A new 10 Gb copper run must span the normal full channel
Category 6A is the common fit from the current study-guide decision set. Category 6 can support 10 Gb on shorter runs, so the full-distance requirement changes the answer.
A campus building link is much longer than an office copper run
Use fiber rather than stretching twisted-pair copper beyond its intended channel. For a long-reach fiber design, single-mode is the stronger clue. Then match optics, wavelength, connectors, and distance at both ends.
Two SFP modules fit, but the link stays down
Do not conclude that the modules match. Compare protocol or speed, wavelength, media type, reach, connector, and peer configuration. SFP describes the form-factor family.
A duplex fiber link has compatible optics but no link light
Check TX/RX polarity. Each transmitter must reach the receiver at the other end. Then inspect cleanliness, patching, and optical signal.
An AP works with a separate power adapter but not when powered only from the switch
Investigate PoE. Check the switch's available budget, port power state, endpoint requirement, and cable pair condition. The successful data path does not prove successful power delivery.
Users report intermittent failures and CRC errors rise during the incident
The increasing counter is useful physical evidence. Compare the copper or fiber path, termination, interference, transceivers, optical signal, and interface settings. Do not start with DNS or routing when corrupted frames are accumulating on the local link.
A cable is labeled Cat 6A but still performs badly
The category label does not prove the installed channel is healthy. Check distance, termination, patch cords, damage, bends, interference, and both interfaces.
11. Common exam traps
- Choosing fiber only because it sounds faster, without considering distance, environment, ports, and cost.
- Treating single-mode and multimode fiber as interchangeable because the connector fits.
- Assuming SFP or QSFP identifies the link speed.
- Matching only the local optic and forgetting the far-end optic.
- Replacing a fiber cable before checking TX/RX polarity.
- Treating a clean-looking connector as proof that the fiber end face is clean.
- Choosing the highest copper category when a lower category already meets the stated speed and distance.
- Assuming plenum rating improves network speed.
- Using shielded cable as a one-piece fix without considering the full channel and grounding.
- Treating successful data transfer as proof that PoE is healthy.
- Treating every nonzero interface error counter as a current fault instead of checking whether it is increasing.
- Treating drops as automatic proof of bad cabling when congestion or queues may be responsible.
- Using a toner to certify cable performance. A toner traces or identifies a cable.
- Using a visual fault locator when the needed measurement is received optical power.
- Changing Layer 3 settings to repair a clearly evidenced Layer 1 problem.
12. Rapid review grid
| Clue | Think first | Do not confuse it with |
|---|---|---|
| Long campus fiber run | Single-mode fiber | Multimode just because both use fiber connectors |
| Short rack-to-rack high-speed link | DAC may fit | Long-reach optical design |
| 10 Gb copper, full normal channel | Category 6A | Category 6 shorter-run capability |
| Module fits, link down | Full transceiver compatibility stack | Form factor alone |
| Compatible fiber parts, no light | TX/RX polarity and signal path | Automatic switch failure |
| Data works, powered endpoint does not | PoE budget, port, endpoint, cable pairs | IP addressing |
| CRC count rises with failures | Physical or interface corruption | Name resolution |
| Need to trace an unknown copper run | Toner | Cable certification |
| Need actual optical receive level | Optical power meter | Visual fault locator |