Quick Answer: A coaxial cable attenuation figure means nothing without a frequency attached, because loss climbs with the square root of frequency: a cable specified at 5 dB per 100 m at 100 MHz is nearer 15 dB per 100 m at 1 GHz and about 25 dB at 2.4 GHz. Diameter and dielectric decide the rest, and a thicker cable always wins. Size the run from the highest frequency you will carry, then add 3 dB of headroom. Published tables describe a new, dry cable, not a run that has been outdoors for years.
Attenuation is quoted in dB per 100 m, and every catalogue page prints it at one frequency only. That single number says almost nothing on its own, because a coaxial cable loses more as frequency rises and less as its diameter grows. Read the curve rather than the headline figure, and keep 3 dB in reserve for connectors, water and age. A 6 dB mistake separates a working link from one that drops out in rain.
Why a single dB figure is meaningless
Frequency scaling
Skin effect pushes current towards the surface of the conductor as frequency rises, so the useful copper area shrinks. Loss therefore grows roughly with the square root of frequency: four times the frequency, twice the loss. That approximation holds well enough across the terrestrial range, and it is why one figure at 100 MHz cannot describe a 2.4 GHz link.
Diameter and dielectric
A thicker centre conductor and a larger foam dielectric both reduce loss, which is why a 7 mm cable beats a 5 mm patch lead by a wide margin at the same frequency. Air-spaced foam is better than solid polyethylene. Copper-clad steel centre conductors lose a little more than solid copper at high frequency.
Reading a loss table
Vendors print dB per 100 m at fixed frequencies, often 50 MHz, 100 MHz, 400 MHz, 1 GHz and 2 GHz. Compare at the highest frequency your system uses, convert to your actual length, then add connector and splitter loss, which is often the larger number. A quad shield RG6 cable is worth knowing for noisy sites, where a second foil and braid cut ingress.
Worked sums that hold up
A 25 m run of RG6 to a 2.4 GHz receiver: 32 dB per 100 m gives 8 dB, plus 0.5 dB per connector at each end gives 9 dB. The receiver needs a workable level, so 9 dB is usually acceptable and 20 dB is not. Do the sum before buying the cable, not after.
Attenuation at a glance
The table gives typical figures for three common 75 ohm cables, for a new dry cable with good connectors. The same numbers decide an S/PDIF link, as a comparison of coaxial vs optical audio shows.
| Frequency | WF100, dB per 100 m | RG6, dB per 100 m | RG59, dB per 100 m |
|---|
| 100 MHz | 5 dB | 6 dB | 10 dB |
| 500 MHz | 11 dB | 13 dB | 21 dB |
| 1 GHz | 15 dB | 20 dB | 30 dB |
| 1.5 GHz | 19 dB | 25 dB | 37 dB |
| 2.4 GHz | 25 dB | 32 dB | 48 dB |
Designing a run that survives
Start from the highest frequency, the longest length and the number of splits. Allow 0.5 dB per connector as a rule of thumb, about 3.5 dB per two-way splitter, and 3 dB of headroom for water and age. Where the loss still exceeds the budget, fit a larger cable rather than a longer amplifier chain, and seal every outdoor joint with a waterproof F connector instead of tape alone. A 75 ohm system must also stay 75 ohm end to end.
Practical scenarios
- Situation: a 40 m satellite run fails in heavy rain. At 2.4 GHz RG6 loses about 13 dB over that length, so fit WF100 or a larger cable rather than a bigger dish.
- Situation: four outlets from one aerial. Each two-way splitter costs about 3.5 dB, so two of them cost 7 dB before any cable; use one four-way splitter and WF100.
- Situation: a 2 m patch lead measures 6 dB loss at 1 GHz. That is normal for thin coax, so keep patch leads under 1 m and use WF100 for anything longer.
- Situation: an FM aerial fed with 50 ohm cable. Replace it with 75 ohm coax, because the mismatch costs signal and confuses every later measurement.
FAQ
Is dB per metre or dB per 100 m the normal figure?
Catalogues quote dB per 100 m, because per metre figures are awkward decimals: 0.15 dB per metre is 15 dB per 100 m. Divide the per-100 m figure by 100 for one metre, then multiply by your length. Check which unit a table uses before comparing cables.
Does attenuation affect digital and analogue signals the same way?
The cable attenuates both equally, but the visible effect differs. Analogue shows noise and snow as the level falls, while digital holds a clean picture until error correction gives up, then breaks up suddenly. That cliff is why headroom matters more on digital links.
Can an amplifier fix a lossy cable?
An amplifier at the aerial end can recover a weak but clean signal; one at the far end raises noise as well as signal. If the loss comes from a thin cable, a larger cable is the better spend, because it improves the ratio without adding a powered device to maintain.
