Network tools: what a tester, a toner and an analyser each actually prove

Why this matters

Every tool in this lesson answers exactly one question, and the skill being examined is choosing the tool whose question is the one you have. Reaching for a cable tester when the fault is interference wastes an hour; reaching for a Wi-Fi analyser when the fault is a broken conductor wastes the same hour in the other direction.

There is a second, subtler point that recurs throughout domain 5: knowing what a passing result does not prove. A cable that passes a continuity test can still be unusable at gigabit speeds. A tool that reports success has answered its own narrow question and nothing more, and treating that as a clean bill of health is how faults survive investigation.

This lesson closes domain 2 and is where its hands-on work is applied.

The lesson

The crimper, the punchdown tool, and what a bad termination looks like

A crimper attaches a modular plug to a cable. It presses the plug's contacts through the conductor insulation and simultaneously clamps the strain relief onto the jacket. A good crimp is mechanically solid and electrically consistent; a bad one is neither, and usually only one of those is obvious.

A punchdown tool seats a conductor into an insulation-displacement contact in a jack or patch panel and trims the excess in the same stroke. It has a spring mechanism that releases at a set force — which is why pushing conductors in with a screwdriver produces terminations that work today and fail later.

The recognisable symptoms of a bad termination:

• One pair not making contact — no link at all, or a link that negotiates at 100 Mbps instead of 1 Gbps, because gigabit requires all four pairs and 100 Mbps requires two. A link that came up at exactly a tenth of the expected speed is a termination fault until proven otherwise.
• Excessive untwist — link comes up, errors accumulate, throughput is poor and worsens with cable length.
• Jacket not clamped — works on the bench, fails when the cable is moved, which is the classic intermittent.
• Split pairs — the conductors are in the right pin positions but taken from the wrong pairs. Continuity tests pass perfectly and the cable performs badly, which is the single best illustration of the "what a pass does not prove" point.

Cable testers: continuity, wire map, and what a passing test does NOT tell you

A basic continuity tester has two units. It energises each conductor in turn at one end and reports what arrives at the other. From that it can report:

• Opens — a conductor that does not reach the far end.
• Shorts — two conductors connected to each other.
• Miswires — conductors arriving at the wrong pin.
• Reversals — a pair connected the wrong way round.

That is a wire map, and it is genuinely useful: it catches the majority of hand-termination errors in seconds.

What it cannot see is everything electrical. It does not measure attenuation, crosstalk, noise, or the effect of untwisting. A split pair, as above, passes a wire-map test and fails in service. So does an over-length run, and so does a cable run alongside a fluorescent ballast.

Measuring those requires a cable certifier, which is an expensive instrument that tests against the category's specification and produces a pass or fail with figures. Installers use them because a certified installation is a contractual deliverable. A technician diagnosing one desk does not need one, and knowing that distinction is exactly what the exam is asking about.

A time-domain reflectometer sits between the two: it sends a pulse and measures the reflection, which gives the distance to a fault. That is the tool for "the cable is broken somewhere inside a 60-metre wall run".

Toner and probe for finding the other end of a cable in a crowded rack

A tone generator clips onto a cable and injects an audible signal. An inductive probe detects that signal without any electrical contact, so running it along a bundle of cables makes the right one audibly louder.

This solves a specific and very common problem: a patch panel with 48 unlabelled ports and one cable at a desk. It is the only practical method short of tracing the cable physically.

Practical points that the exam and the job both care about:

• Disconnect the cable from equipment first. Injecting tone into a live switch port is a good way to damage something.
• Tone bleeds onto neighbouring cables. The right one is loudest, not the only one making noise, so work by comparison rather than by presence.
• Some testers place a tone on a specific pair, which helps when several cables are toned at once.
• Label as you go. The tool exists because someone did not, and the job is not finished until the next person will not need it again.

Wi-Fi analysers: reading signal, noise and channel occupancy rather than bars

A Wi-Fi analyser shows what is actually in the air: which networks exist, on which channels, at what signal strength, and how busy each channel is. It turns "the wireless is bad here" into numbers.

The measurements that matter:

• Signal strength, in dBm, which is a negative number where closer to zero is stronger. Around -50 dBm is excellent, -70 dBm is workable, and -80 dBm is marginal. These are worth memorising because "bars" are not comparable between devices.
• Noise floor, and therefore the signal-to-noise ratio, which is what actually determines usable throughput. A strong signal in a noisy environment performs worse than a weaker signal in a quiet one.
• Channel occupancy, which shows how many networks share each channel and how much airtime is in use.

That last one produces the most common actionable finding in small-office wireless: every neighbouring network is on channel 6 because every router defaults there, and moving to 1 or 11 improves matters immediately.

An analyser also reveals what a client cannot: an access point transmitting at a channel width that overlaps three neighbours, or a network broadcasting on a band the client does not support. Some tools additionally capture frames, which crosses into network analysis proper — beyond Core 1's scope, but worth knowing exists.

Choosing the cheapest tool that answers the question in front of you

The habit worth building is to state the question before picking up anything.

• "Which cable is this?" → toner and probe.
• "Is this cable terminated correctly?" → wire-map tester.
• "Where is the break in this wall run?" → time-domain reflectometer.
• "Does this installation meet Cat 6a?" → certifier, and usually a contractor.
• "Is this port live and is the switch seeing it?" → link lights, then the switch's own port status.
• "Why is wireless bad in this room?" → Wi-Fi analyser.
• "Does this machine have an address and a gateway?" → the operating system's own configuration output, which costs nothing.
• "Can this machine reach that service?" → ping, then a port test.

Two principles behind that list, and both return in domain 5. First, prefer the test that eliminates the most possibilities per unit of effort — checking a link light before fetching a tester is free and rules out a great deal. Second, prefer the test that produces a recordable result: "signal is -78 dBm at the far desk" survives a handover, and "wireless seemed weak" does not.