Ham Radio Coax Cable Testing — Fault Finding & Loss Measurement
A coaxial cable that looks fine externally can be hiding a corroded connector, a water-filled section, a crushed braid, or an intermittent centre conductor that wastes half your transmitter power before it reaches the antenna. This guide gives you a complete coax testing toolkit — from a £4 ohmmeter to a NanoVNA TDR — so you can quickly locate and diagnose any feedline fault.
Coaxial cable fails in predictable ways. Knowing the failure modes before starting a test helps you choose the right test method and interpret the results correctly. The five principal coax fault types in amateur radio installations are:
Water ingress
The most common outdoor failure. Water enters through cracked jacket, damaged weatherproofing at connectors, or capillary action along the braid. Water raises the dielectric constant and loss tangent dramatically, causing increased insertion loss, SWR sensitivity to rain, and eventually conductor corrosion that creates a resistive fault.
Mechanical damage
Crushing the cable (under doors, cable staples driven too hard), sharp bends beyond the minimum bend radius, or animal/insect damage cuts through the jacket and braid. This can cause intermittent contact between braid and centre conductor — showing up as an SWR that varies with cable position.
Connector failure
Cold solder joints, corroded contacts, braid strands bridging to the centre pin, and physical connector separation (from antenna weight pulling on the coax). Connector faults produce the highest insertion loss per unit length and are the most common source of sudden SWR changes.
Dielectric breakdown
High SWR creates voltage peaks along the cable. Over time at high power levels these can arc through the dielectric at weak points — damage from sharp bends, manufacturing flaws, or repeated thermal cycling. Dielectric faults are power-dependent: fine at low power, arcing at full power.
Braid corrosion
In coastal or high-humidity environments the braid oxidises, increasing its resistance. At HF this adds a few tenths of a dB; at VHF/UHF the skin depth is small enough that braid resistance matters significantly. Corroded braid shows up as higher loss across all frequencies with no obvious fault point.
The simplest and fastest coax test requires only a digital ohmmeter. It catches the most severe faults — open circuits, short circuits, and very high-resistance connections — in under two minutes. It will not detect moderate loss increases, water ingress, or dielectric damage, but it eliminates catastrophic failures before you waste time on more sophisticated testing.
Remove the coax from the radio and from the antenna (or connect a short-circuit reference at the far end). Never test a connected cable — the radio's circuits or the antenna's feed point will interfere with the measurement and may be damaged by the ohmmeter's test current.
Using a clip lead or a piece of wire, short the centre conductor to the outer braid at the far end of the cable. This reference short allows you to verify continuity of both conductors simultaneously from the near end.
With the far-end short in place, measure resistance between the centre pin and the outer barrel of the near-end connector. Should read very low — typically 0.1–2.0 Ω for RG-213 at 30 m run, scaling with cable length and conductor size. The measurement should be stable; any variation when the cable is flexed indicates an intermittent fault.
Without the far-end short, measure resistance from centre to braid at the near end. Should read open circuit — several megaohms or higher. Any reading below 1 MΩ indicates dielectric leakage — the cable insulator has been compromised, most likely by water ingress or mechanical damage. Readings below 100 kΩ indicate a serious dielectric fault.
| Measurement (far-end shorted) | Interpretation | Action |
|---|---|---|
| 0.1–3 Ω (centre to braid) | Good — both conductors continuous | Pass DC test, proceed to RF test |
| Open circuit (>1 MΩ) | Broken centre conductor or braid | Locate break — use TDR or divide-and-test method |
| Very high resistance (10–100 Ω) | Corroded contact or high-resistance joint | Check all connectors first — re-solder or replace |
| Near zero even without far-end short | Short circuit — braid touching centre | Locate short — check connectors, then TDR |
| Measurement (far-end open) | Interpretation | Action |
|---|---|---|
| >10 MΩ | Good dielectric — no leakage | Pass |
| 1–10 MΩ | Minor dielectric degradation | Monitor; may worsen with age and moisture |
| 100 kΩ–1 MΩ | Significant moisture or dielectric damage | Dry cable section; replace if persistent |
| <100 kΩ | Severe dielectric failure — water-filled section | Replace cable section |
Coax DC Loop Resistance Calculator (centre + braid, far end shorted)
DC tests cannot detect moderate water ingress, dielectric degradation, or increased conductor loss — all of which only manifest at RF. Measuring the cable's insertion loss at your operating frequency gives the most direct indication of feedline efficiency. There are three practical ways to do this with equipment commonly available to amateurs:
Method A: NanoVNA S21 measurement (most accurate)
Connect the NanoVNA's PORT 1 (transmit) to one end of the cable and PORT 2 (receive) to the other end with the far end terminated in 50 Ω (connect a 50 Ω dummy load or use the NanoVNA's built-in 50 Ω reference). The S21 display shows insertion loss directly in dB across the frequency range. A cable showing 3 dB of insertion loss at your operating frequency is losing half your transmitter power. Compare measured loss against the manufacturer's specification for the cable type and length — excess loss indicates a fault.
Method B: Signal generator and power meter
Connect a signal generator to one end of the cable and a RF power meter or spectrum analyser to the other end, both terminated in 50 Ω. Measure the output power with and without the cable — the difference in dB is the insertion loss. This method is highly accurate but requires signal generator and power meter equipment not usually found in most amateur shacks.
Method C: SWR sweep at both ends (indirect)
Connect a dummy load at the far end of the cable and measure SWR from the shack end with an antenna analyser. A good cable with a 50 Ω load shows SWR very close to 1.0:1. A faulty cable with a 50 Ω load shows SWR progressively worse from 1.0:1 as loss increases — because the loss hides the mismatch at the far end. This is why very lossy cables appear to have "good SWR" — they are lossy enough to mask any mismatch. If your cable shows suspiciously perfect SWR (below 1.1:1) across a very wide frequency range, consider whether the cable may be absorbing power due to fault-induced loss.
Test 3: TDR Fault Location with a NanoVNATime Domain Reflectometry (TDR) is the gold standard for locating the exact position of a fault in a coaxial cable run. A TDR sends a pulse down the cable and measures how long it takes for reflections to return. Since the pulse travels at a known fraction of the speed of light (determined by the cable's velocity factor), the round-trip time directly gives the distance to the fault.
Modern NanoVNA firmware — particularly NanoVNA-Saver and some direct NanoVNA-H4 firmware versions — includes a TDR function that performs this measurement by computing the inverse Fourier transform of the S11 frequency sweep. The result is displayed as distance vs. reflection amplitude, with peaks indicating impedance discontinuities (faults, connectors, and the cable end).
Launch NanoVNA-Saver (free download from GitHub) and connect the NanoVNA via USB. Set the sweep range appropriately — for TDR, use a wide frequency sweep from 1 MHz to the NanoVNA's maximum frequency (typically 300 MHz or 900 MHz depending on model). More frequency range = better distance resolution in TDR mode. Set the number of sweep points to 201 or more.
Apply the standard SOLT calibration at PORT 1 (SHORT, OPEN, LOAD). Accurate calibration is essential for TDR — errors in the calibration appear as false peaks near the measurement start. After calibration, connect the cable under test to PORT 1. Leave the far end of the cable open circuit (do not connect anything to it).
In NanoVNA-Saver's TDR panel, enter the velocity factor (VF) of your cable type. Common values: RG-58 = 0.659, RG-213 = 0.659, RG-8X = 0.820, LMR-400 = 0.850, RG-6 = 0.820. An incorrect VF gives correct peak positions relative to each other but wrong absolute distances. If you know the physical cable length, you can calibrate the VF by adjusting it until the far-end peak appears at the correct distance.
The TDR plot shows reflection amplitude vs. distance from the NanoVNA's port. You should see a clean baseline, then a peak at the cable's far end (the open circuit). Any additional peaks between 0 and the far-end peak indicate impedance discontinuities — connectors appear as small peaks, water-filled sections appear as gradual deviations or multiple peaks, and hard faults (shorts or opens) appear as large peaks. The distance readout at each peak tells you exactly where along the cable the fault is located.
TDR Fault Distance Calculator
When a TDR is not available and the fault location is unknown, the divide-and-test method is the next best approach. It works by systematically bisecting the cable run to locate the fault — similar to binary search. Each test halves the length of cable that could contain the fault, quickly narrowing it to a short section that can be replaced.
Use the DC continuity test to confirm whether the fault is an open circuit, short circuit, or high resistance. Knowing the fault type before beginning the search prevents wasted tests — an open circuit in the braid is found differently from a dielectric leak.
Locate the physical midpoint of the cable run. Either expose the cable there (temporarily remove from conduit, unclip from the wall, or gain access to a cable tray) and temporarily connect a short reference or dummy load. Perform the DC continuity test on each half separately. The half that shows the fault is where the problem lies.
Expose the midpoint of the identified faulty half and repeat the test. After three or four rounds of bisection, you have narrowed the fault to a 1–3 m section that can be replaced with a short jointed section or, more reliably, by replacing the entire faulty segment. Each round of bisection halves the search space.
Connectors are the most common source of coax faults and deserve careful, methodical inspection every time SWR unexpectedly rises or signal levels drop. The following checklist covers the most common connectors in amateur radio use.
Visual inspection checklist
- PL-259 centre pin: Should be bright silver or gold. Green or black oxidation indicates moisture ingress. The pin should not move when pushed with a finger — movement indicates a cold solder joint inside the connector body
- PL-259 body/barrel: Inspect the solder holes on the barrel — they should be filled flush. Any braid strands visible inside the barrel approaching the centre pin area indicate a potential short circuit
- SO-239 female socket: The spring contact inside should grip the male pin firmly. Any looseness, green oxidation, or visible corrosion disqualifies the connector
- N-type connector: The PTFE dielectric insert should be clean white — any browning or charring indicates previous arcing from high power with a poor contact. The coupling nut should thread smoothly to the mating connector and tighten to finger-tight plus approximately 1/4 turn
- BNC connector: The bayonet locking tabs on the male connector should engage positively and lock with a click. Worn tabs that slip or a loose mated connection cause intermittent contact
Testing connector quality with a NanoVNA
Connect a known-good terminator (50 Ω dummy load) to the suspect connector and measure S11 (return loss/SWR) across a frequency range relevant to your operating frequencies. A good connector-and-terminator combination shows return loss of 30+ dB (SWR 1.06:1 or better) up to the connector's rated frequency. Poor connectors show return loss degrading above their rated frequency — a PL-259 typically shows return loss falling below 20 dB above 150–200 MHz, confirming that PL-259 connectors should not be used at VHF.
Diagnosing Wet CoaxWater-infiltrated coaxial cable is one of the hardest faults to diagnose because the symptoms develop gradually and can mimic many other problems. The definitive sign of a wet cable is SWR readings that change with weather — rising after rain and improving when dry — combined with progressively worsening insertion loss over months. A NanoVNA TDR sweep of a wet cable characteristically shows a gradual elevation in the S11 baseline between the cable entry point and the far-end peak, rather than a sharp discrete peak that indicates a hard fault.
Drying a wet cable
Disconnect the cable at both ends. Use a heat gun on low setting applied along the cable for 15–20 minutes, starting from the lowest accessible point to allow steam to escape upward. Then connect a 12 V DC supply (no RF) across the centre conductor to braid — DC heating of 0.5–1 W warms the internal dielectric and accelerates moisture evaporation. Leave both ends open for 24–48 hours in dry conditions. Repeat the DC resistance test — if it improves significantly, the cable may be salvageable. If the dielectric resistance remains low (under 1 MΩ), the cable should be replaced.
My SWR is perfect but my signal is weak — could it be the coax?
Yes — this is the classic symptom of a high-loss coax. A very lossy cable (water-filled, damaged dielectric) absorbs power in both directions. On transmit it absorbs transmitter power; on receive it absorbs signal. The SWR appears good because the cable's own loss masks the antenna mismatch — a 6 dB loss cable hides significant antenna SWR. Measure insertion loss directly with a NanoVNA S21 test to catch this.
How often should I test my coax?
Annual visual inspection and DC continuity test is good practice for permanently installed outdoor coax. A full NanoVNA sweep (insertion loss and TDR) after any significant weather event (storms, heavy snow load), after any mechanical work on the antenna system, or whenever signal levels drop unexpectedly. Cables older than 10–15 years in outdoor installations should be inspected carefully — UV degradation of the jacket and long-term connector oxidation are inevitable.
Can I join a broken coaxial cable rather than replacing the whole run?
Yes — a properly made coaxial splice using a barrel connector and correctly fitted connectors at each end is acceptable at HF. The splice introduces approximately 0.2–0.5 dB of additional insertion loss and is a point of potential future failure. At VHF and UHF, a splice also introduces a small impedance discontinuity that slightly degrades return loss. A well-made splice is better than leaving a broken cable; a completely replaced cable section is better than a splice.
What does a TDR peak at the very start of the trace mean?
A large peak immediately at distance 0 (at the test port) indicates a poor connector at the NanoVNA connection point — the impedance mismatch of a bad connector appears as a large near-field reflection before the cable itself. Check and clean the connector at the NanoVNA port first, or re-calibrate with a better reference. This is a very common source of TDR false readings when first using the technique.
How do I test a buried coax without digging it up?
TDR is ideal for buried cable testing. Connect the NanoVNA at one end of the buried run and perform a TDR sweep — any fault anywhere along the buried section appears as a peak at the calculated distance, telling you exactly where to dig. The DC continuity test also works without excavation: a short or open circuit in a buried run is detected from either end. Combine TDR for fault location with DC testing for fault type characterisation before excavating.
What is the best coaxial cable for a permanent outdoor HF installation?
LMR-400 or equivalent (Times Microwave, RF-400) is the best practical choice for most HF installations: low loss, good UV resistance, flexible enough to handle, and widely available with factory-made connectors. RG-213 is adequate and cheaper but higher loss and less UV-resistant. Avoid RG-58 for any permanent outdoor run over 15 m at HF — its higher loss and poorer UV resistance make it a false economy compared to LMR-400.