RG-58 Coax: When It Is Adequate and When Loss Wins

RG-58 is a family label applied to materially different commercial constructions. A short, named RG-58 cable can be entirely adequate at HF; a long, small-diameter run can waste most of a UHF signal. Select the complete cable assembly from its measured and published limits, not from the family name alone.

RG-58 Does Not Identify One Construction

The historical RG designations distinguish constructions, but current catalogues also use “RG-58,” “RG-58 type” or an RG cross-reference across products that are not electrically or mechanically interchangeable. Three active Belden examples make the problem concrete:

All three are catalogued with RG type 58, yet conductor size, dielectric, shielding, impedance data, attenuation and outdoor suitability differ. Even two cables with the same diameter and connector family need not have the same loss or flexibility. “RG-58” is therefore a starting clue, not a procurement specification.

Loss Requires Frequency and Length

A manufacturer normally publishes matched attenuation in dB per unit length at named frequencies. For a uniform cable, scale the tabulated value to the installed length:

These equations describe a matched line and exclude connector loss. They also show why “low loss” must be quantified: one dB number without frequency and length is incomplete.

The IEC 61196-1-113:2024 attenuation-constant method defines a controlled cable test with declared frequency and test conditions. A datasheet value is therefore a screening input for the named cable, not a guaranteed end-to-end result for connectors, adapters, bends, ageing or installation damage.

Compare Exact Products, Not Family Labels

The table below uses each manufacturer's published nominal or typical values. It is a screening comparison, not a substitute for tolerances, temperature correction or a purchased-cable acceptance test.

Two conclusions matter. First, the RG-58-labelled Belden 7807A has lower published loss than the same-diameter LMR-195 at both shown frequencies. That does not make 7807A universally better; it proves that family names and replacement labels cannot replace part-number data. Second, the much larger LMR-400 has a major UHF loss advantage, purchased with more diameter, weight, bend radius, wind load, connector size and installation effort.

The 900 MHz column is deliberately severe: 100 ft is 30.48 m. Short jumpers tell a different story. Scaling Belden 9201's nominal table to 2 m gives approximately:

That 2 m HF jumper is not disqualified by its RG label. The same small cable over a long UHF run is. For receive systems, translate loss directly into system noise figure when the passive line is ahead of the first low-noise amplifier; 3 dB of preamplifier feedline loss is a roughly 3 dB noise-figure penalty under the usual matched, thermal-equilibrium assumptions.

Shielding, Common Mode and Antenna Pickup Are Different Paths

Shielding effectiveness cannot be inferred from the RG-58 family name or braid coverage alone. Belden 9201 specifies 80% optical braid coverage but does not publish a shielding-effectiveness number in that product sheet. Optical coverage is not itself a dB shielding measurement.

By contrast, the exact Times LMR-195 and LMR-400 data sheets specify shielding effectiveness greater than 90 dB for their tape-plus-braid constructions. That number still belongs to those products and the manufacturer's test basis; it is not automatically the performance of a field-made assembly.

Times Microwave's cable-assembly shielding note explains that transfer impedance and resonant-cavity methods characterize different frequency regions. It also warns that a shield-only transfer-impedance test does not test the connector interface or the cable-to-connector junction. Connector backshells, pigtails, damaged foil, poor braid termination and enclosure seams can dominate a supposedly excellent cable.

The current consolidated IEC 62153-4-7:2021 with Amendment 1:2025 provides a triaxial method for transfer impedance and screening or coupling attenuation of connectors and assemblies, including the cable-to-connector connection. That assembly boundary is the relevant one when the installed connector junction could dominate leakage.

Common-mode current on the outside of a coax shield is not the same as the wanted differential TEM signal leaking through the shield. A current choke addresses the exterior common-mode path. It does not repair poor shield transfer impedance, a leaking connector or receiver overload caused by a strong signal entering through the antenna. Likewise, an SDR does not “expect modern coax”; its overload margin depends on the actual signal spectrum, preselection, gain distribution and front-end limits.

Outdoor Life Is a Jacket and Sealing Question

Outdoor suitability belongs to the exact jacket and construction. The current Belden 9201 sheet lists indoor use, while Belden 8219 explicitly lists outdoor and UV suitability. Times offers LMR-195 and LMR-400 with distinct outdoor PE, watertight direct-burial, riser and general-purpose PVC variants. Those suffixes are functional specifications, not decoration.

• Sun and weather: require the exact jacket's UV, temperature and outdoor rating.
• Direct burial or conduit: use a construction explicitly rated for the moisture exposure; “outdoor” and “watertight” are not interchangeable promises.
• Connectors: use the manufacturer's preparation dimensions and compatible connector. Weather-seal the finished interface without trapping water.
• Mechanics: respect installation and repeated-flex bend radii, pull tension, support spacing and strain relief. Crushing or tight bends can change impedance and loss.
• Inspection: investigate rising insertion loss, unstable SWR, corrosion, jacket damage or water at a termination. A cable analyser can compare the installed line with its commissioning baseline.

“QRO-Safe” Is Not a Cable-Family Rating

RF power handling has at least two different boundaries. Average or continuous power is usually thermal: conductor and dielectric loss generate heat that must escape. Peak power is usually tied more closely to voltage and dielectric breakdown. A peak-power number is not permission to run that many average watts.

The exact cable examples differ substantially:

Belden 9201 publishes a voltage rating but no RF-power table in the cited sheet. Converting that voltage directly into a universal transmitter-power rating would ignore heating, frequency, waveform, mismatch, termination and environment, so this article does not do it.

Times Microwave's high-power coax selection guide explicitly requires derating for ambient temperature, altitude and VSWR. It notes that mismatch creates local hot spots and that peak capability relates to maximum operating voltage. Solar heating, cable bundling, restricted airflow and a hot roof can remove thermal margin that exists in a laboratory rating.

IEC 61196-1-119:2023 likewise treats RF power rating and power withstand as tests of a specified cable or cable assembly at specified frequency, temperature and altitude. A complete station rating must remain inside those test boundaries and the limits of every connector and accessory.

What mismatch changes

For a line with real characteristic impedance and load SWR S, the reflection-coefficient magnitude is:

At 3:1 SWR, |Γ| = 0.5. Local voltage and current maxima can reach 1.5 times their respective forward-wave amplitudes, at different positions along the line. Loss also attenuates both travelling components and turns some RF power into heat. The exact additional line loss and hot-spot location depend on attenuation, electrical length and complex load—not just the scalar SWR at the radio.

When a Small RG-58-Type Cable Is Sensible

• Short HF jumpers: the exact 2 m example above loses only about 0.07 dB at 10 MHz before connector effects.
• Portable or mobile systems: low mass and easy routing may be worth more than a small loss difference, provided flex life, jacket and connector are correct.
• Bench and instrument leads: a named, characterized assembly can be appropriate inside its frequency, power and calibration limits.
• Low-power choke windings: small diameter may fit a core, but observe bend radius and evaluate common-mode impedance and ferrite heating. Straight-cable power data do not rate the completed choke.
• Legacy interfaces: retaining a sound BNC assembly may be better than adding lossy or mechanically poor adapters merely to use a larger cable.

Short jumpers, portable leads, patch cables and low-power choke windings can all be sensible uses when the exact construction stays inside its electrical, thermal, mechanical and environmental limits.

A Selection Workflow That Survives the Datasheet

1. Name the endpoints and operating bands. Include the highest frequency, not just “HF” or “UHF.”
2. Measure the real route. Add service loops and connector transitions; do not compare catalogue loss at an invented length.
3. Set the loss budget. Calculate matched insertion loss at each important frequency, then include connectors and mismatch. On receive, include pre-LNA loss in the noise budget.
4. Define power properly. Record average and peak envelope power, modulation, duty cycle and maximum tune or fault duration.
5. Derate the published table. Apply the manufacturer's temperature, altitude and VSWR method. Ask the manufacturer when assumptions are missing.
6. Specify the environment. Indoor, UV, aerial, direct-burial, fire, temperature, flex and chemical requirements select different jackets and constructions.
7. Check the assembly. Cable shielding does not guarantee connector-junction shielding. Use approved connectors, tools, preparation and weather sealing.
8. Commission and record. Measure insertion loss or one-way transmission where practical, SWR/return loss, and assembly temperature at conservative power. Keep results for ageing comparisons.

Selection Summary

RG-58 is not automatically noisy, unsafe, obsolete or unsuitable for outdoor service. It is also not automatically good enough. Those are properties of an exact cable assembly in an exact system.

For a short HF jumper, a documented RG-58 construction can be efficient and convenient. For a long UHF feeder, small diameter often imposes a serious link-budget penalty, and a larger low-loss cable may repay its mechanical cost. For high power, the manufacturer table, mismatch derating, environment and weakest assembly component decide—not folklore and not the letters printed on the jacket.

Questions, antenna-factor records or height trials to share? Contact RF.Guru.