Published on 10/13/2026
Someone runs 30 meters of RG58 up a tower for a 433 MHz link, checks the datasheet, expects about 4.5 dB of loss, and measures 7. The datasheet was not lying. The datasheet describes a new cable, at room temperature, terminated in a perfect 50 ohm system, with no connectors counted. Real installations add loss the table never promised to include.
This article lays out published nominal attenuation figures for a common RG58-class cable, explains the gap between datasheet and installed loss, and shows you how to estimate and verify the real number. A note on evidence: we are not publishing an invented lab dataset here. The figures below are published nominal values from a major manufacturer's datasheet (Belden 8240, a common RG58 C/U type), and the loss mechanisms are standard transmission-line physics. Where numbers are examples, they are labeled as examples.
On This Page
The table below reproduces nominal attenuation from the Belden 8240 datasheet, a widely used 50 ohm RG58 C/U cable. Other RG58 variants differ; always pull the datasheet for the exact part on your reel.
| Frequency | Nominal attenuation (dB per 100 ft) | Approx per 100 m |
|---|---|---|
| 1 MHz | 0.3 | 1.0 |
| 10 MHz | 1.1 | 3.6 |
| 50 MHz | 2.5 | 8.2 |
| 100 MHz | 3.8 | 12.5 |
| 200 MHz | 5.6 | 18.4 |
| 400 MHz | 8.4 | 27.6 |
| 700 MHz | 11.7 | 38.4 |
| 900 MHz | 13.7 | 45.0 |
| 1000 MHz | 14.5 | 47.6 |
Two lessons hide in that table. First, loss grows roughly with the square root of frequency, so doubling frequency does not double loss, but going from 100 MHz to 900 MHz multiplies it by about 3.6. Second, RG58 is a short-run cable: at 900 MHz, 30 m of it throws away more than 13 dB, over 95 percent of your power. Our RG58 vs LMR-240 comparison shows what the same run costs on a low-loss cable, and the RG58 complete guide covers the full spec set.
Key idea: Datasheet attenuation is a nominal room-temperature value for bare cable in a matched system. It is the floor for your real loss, not the estimate.
Nominal means typical, not guaranteed maximum. Manufacturing variation, plating thickness and braid coverage shift the real number a few percent either way. The values are also specified into a matched load: the cable alone, with no connector contributions. And they assume the cable is new, dry and straight. Our guide to reading coax datasheets walks through the other fields that interact with loss, including velocity of propagation and power rating.
| Mechanism | What it adds | What to do about it |
|---|---|---|
| Connectors (2 per assembly) | Typically 0.1 to 0.3 dB each at sub-GHz, more if poorly terminated | Count them in the budget; use quality terminations |
| Temperature rise | Attenuation climbs with temperature; a cable at 60 C can run 10 to 15 percent above the 25 C figure | Derate for rooftop and attic runs |
| Water ingress | Wet dielectric can multiply loss several times over and corrodes braid | Seal outdoor connections; see our waterproofing guide |
| Tight bends and kinks | Kinks create impedance bumps and local loss | Respect minimum bend radius; see our bend radius rules |
| Aging and UV | Jacket cracking, plasticizer loss and braid oxidation creep loss up over years | UV-rated jacket outdoors; periodic verification |
| Mismatch at the load | A high-VSWR antenna adds mismatch loss on top of cable loss | Match the antenna; check system VSWR |
Need a low-loss alternative spec'd for your exact run length and band? We build and sweep it.
That is the practical use of the datasheet: not to predict your exact loss, but to give you a floor and a sanity band. Measure well above the band and you have a fault to find, not a surprise to accept.
Three honest methods, in order of rigor. Best: a VNA sweep of the finished assembly, which gives S21 loss across the band in one pass; every SigmaRF assembly is verified this way before shipment. Field-friendly: a power meter or inline wattmeter at both ends, comparing forward power in versus out. Crude but useful: our multimeter troubleshooting guide catches opens, shorts and gross faults, though it cannot measure loss. If the run is already installed, disconnect both ends and treat it as a component to be tested on the bench, because testing through an antenna confuses cable loss with antenna match.
Mistakes that inflate RG58 loss
What is the attenuation of RG58 at 100 MHz?
Published nominal values for common RG58 C/U (Belden 8240) are 3.8 dB per 100 ft at 100 MHz, roughly 12.5 dB per 100 m. Check the datasheet for your specific part, since RG58 variants differ.
How much loss does RG58 have at 900 MHz?
The same published figures give 13.7 dB per 100 ft at 900 MHz, about 45 dB per 100 m. That makes RG58 a poor choice for long 900 MHz runs; a low-loss cable like LMR-400 cuts that dramatically.
Why is my measured coax loss higher than the datasheet?
Connectors, temperature, moisture ingress, tight bends, aging and antenna mismatch all add loss the bare-cable datasheet figure does not include. Compare your measurement against the datasheet floor plus those additions.
Does coax lose more signal when it gets hot?
Yes. Attenuation rises with temperature, and a cable at 60 C can run 10 to 15 percent above its 25 C specification. Derate for hot installations.
How do I measure the loss of an installed cable?
Disconnect both ends and sweep it with a VNA, or compare power in versus power out with wattmeters. Testing through an antenna mixes cable loss with antenna match and misleads you.
Is RG58 good for long antenna runs?
Generally no above VHF. Its loss at UHF and above makes long runs expensive in dB. Keep RG58 for short jumpers, or step up to a low-loss cable for runs over about 10 m at 400 MHz and beyond.
Datasheet attenuation is a floor, not a forecast. Published RG58 figures tell you what new, dry, room-temperature cable does into a matched load; connectors, heat, moisture and time add the rest. Build your link budget on the datasheet plus the additions, then verify the installed run with a real measurement. When the measured number lands far above the prediction, believe the measurement and go find the fault.
Need assemblies built and verified this way?
SigmaRF builds standard and custom coaxial cable assemblies at our Vadodara, India volume facility and our Cleveland-area precision facility in the USA. Every assembly is VNA-swept for VSWR and insertion loss across 100 MHz to 40 GHz before it ships, and lead times run as short as 2 days.