Coaxial Cable Loss Chart by Frequency and Cable Type
Use this coaxial cable loss chart to compare LMR-100A, LMR-195, LMR-240, LMR-400, and RG-58 across common RF bands so you can size cable runs before signal margin disappears.
About half the power is gone after 3 dB of cable loss, before connector mismatch is added.
This chart compares common small and mid-size 50 ohm cables used in RF assemblies.
A run that looks acceptable at VHF can become expensive in loss at Wi-Fi or cellular bands.
Poor stripping, braid control, or torque can erase the advantage of a low-loss cable spec.
Coaxial cable is one of the easiest places to lose RF margin without noticing it until the product reaches test or field deployment. Teams will spend weeks optimizing antenna placement, shielding, and matching networks, then accept a cable choice that burns away 4 dB, 8 dB, or more before the signal even reaches the front end. On a receiver path, that can mean lower sensitivity, more retries, weaker range, or a noisy margin stack that shows up only at temperature or at the end of a long harness route.
The practical problem is that coax loss is not constant. It changes with frequency, cable construction, diameter, dielectric material, braid coverage, and route length. That is why buyers should never ask whether a cable is "good" in the abstract. The useful question is: how much loss will this exact cable add at my actual operating frequency and installed length? For background on transmission line behavior, see coaxial cable and decibel.
This guide gives a practical loss chart for five commonly compared 50 ohm cable families: LMR-100A, RG-58, LMR-195, LMR-240, and LMR-400. The values are typical datasheet figures in dB per 100 feet at +25 C. That makes the table useful for first-pass design, BOM review, and quote-stage discussions on custom RF cable assemblies. It does not replace final validation, because the finished assembly still depends on connectors, launch quality, bend condition, and test method.
"A lot of sourcing teams compare coax by price per foot. RF teams compare it by dB at the working band. The second method is the one that prevents emergency redesign after EMC or range testing."
— Hommer Zhao, Technical Director
Coaxial cable loss chart by frequency and cable type
The table below is the starting point most buyers actually need. It shows why small cable is attractive for packaging but expensive in signal loss as frequency rises. By 2.5 GHz, LMR-100A is near 40 dB per 100 feet, while LMR-400 is 6.8 dB over the same distance. That gap changes enclosure layout, antenna placement, and whether a cable assembly belongs inside a module or out on a feeder run.
| Cable type | OD | 50 MHz | 150 MHz | 450 MHz | 900 MHz | 2.5 GHz | 5.8 GHz | Best fit |
|---|---|---|---|---|---|---|---|---|
| LMR-100A | 0.110" | 5.1 | 8.9 | 15.8 | 22.8 | 39.8 | 64.1 | Very short pigtails, compact internal routing |
| RG-58 | 0.195" | 3.1 | 6.2 | 10.6 | 16.5 | N/A | N/A | Legacy jumpers, low-frequency short runs |
| LMR-195 | 0.195" | 2.6 | 4.4 | 7.8 | 11.1 | 19.0 | 29.9 | Short antenna leads, tighter bends than LMR-240 |
| LMR-240 | 0.240" | 1.7 | 3.0 | 5.3 | 7.6 | 12.9 | 20.4 | General-purpose low-loss jumpers and feeders |
| LMR-400 | 0.405" | 0.9 | 1.5 | 2.7 | 3.9 | 6.8 | 10.8 | Longer runs where every dB matters |
Two points matter immediately. First, diameter is usually buying lower loss. The jump from LMR-240 to LMR-400 adds routing stiffness and panel-space pressure, but it can cut attenuation nearly in half on longer runs. Second, legacy cable types such as RG-58 still have a place, but usually in short jumpers, low-frequency instruments, or constrained repair situations rather than modern multi-GHz links.
Why the loss rises so fast with frequency
RF cable attenuation is not one single mechanism. Part of it comes from conductor loss, which increases as current crowds toward the conductor surface at higher frequency. Part of it comes from dielectric loss in the insulation. The result is that a cable can look efficient at 50 MHz and much less attractive at 900 MHz, 2.4 GHz, or 5.8 GHz. If you want the physics background, skin effect and insertion loss explain the core trend.
This is why small pigtail cable should be treated as a length-limited design tool, not a default procurement habit. An internal GNSS lead at a few inches may perform perfectly on LMR-100A or similar micro coax. Stretch that same logic into a several-foot external run and the link budget changes very quickly. The cable that saved 3 mm of bend space may cost several dB of useful system margin.
"At 900 MHz, people still get away with mediocre cable choices. At 2.4 GHz and above, the cable starts billing you for every lazy decision in layout, routing, and assembly process control."
— Hommer Zhao, Technical Director
How to turn dB per 100 feet into the number that matters
Buyers often read a chart correctly and still make the wrong decision because they do not convert the published figure into installed loss. The math is simple: actual cable loss equals the chart value times cable length divided by 100. So if LMR-240 is 7.6 dB per 100 feet at 900 MHz, a 20 foot run contributes about 1.52 dB. A 35 foot run contributes about 2.66 dB.
That result is only the cable body. You still need to add connector insertion loss, any adapter stack, and any mismatch penalty caused by poor termination or an impedance discontinuity. For example, a nominal 20 foot 900 MHz run on LMR-240 might look like this: 1.52 dB from the cable, perhaps 0.2 to 0.5 dB from connector pairs depending on the hardware, and then more if the assembly has poor return loss. Our return loss calculator is useful when you need to translate reflection performance into a more practical engineering discussion.
Quick sizing examples
- 25 feet at 900 MHz: LMR-400 is about 1.0 dB, LMR-240 about 1.9 dB, LMR-100A about 5.7 dB.
- 10 feet at 2.5 GHz: LMR-240 is about 1.29 dB, LMR-400 about 0.68 dB, LMR-100A about 3.98 dB.
- 6 feet at 5.8 GHz: LMR-400 is about 0.65 dB, LMR-240 about 1.22 dB, LMR-100A about 3.85 dB.
When each cable family makes sense
LMR-100A and similarly small constructions are useful where packaging dominates the decision. That usually means compact pigtails, embedded antennas, short module jumpers, and places where a stiff 0.240 inch or 0.405 inch cable would create impossible routing. The buyer is trading length capability for package freedom.
RG-58 still survives because it is familiar, easy to source, and good enough in some low-frequency and short-run applications. The problem is that many teams keep using it by habit instead of by loss budget. If the link is cellular, GNSS, Wi-Fi, or a longer feeder path, habit is usually the wrong method.
LMR-195 and LMR-240 sit in the most practical middle ground. They are often the right answer for box-level RF harnesses, antenna jumpers, gateway modules, and industrial equipment where the run is long enough that small cable hurts, but the package still cannot tolerate LMR-400-level stiffness. That is also where connector selection starts to matter more, so many buyers pair the loss review with connector lifecycle planning such as obsolete connector replacement.
LMR-400 is usually the value choice once the run gets long enough. Its price per foot is higher and its routing demands are real, but every dB you save upstream can be worth more than the raw cable cost. On vehicle, outdoor, and enclosure-to-antenna runs, a lower-loss feeder can preserve range and receiver margin that would otherwise require a more expensive radio design to recover.
The chart is only valid if the assembly is built correctly
Published cable attenuation does not guarantee finished assembly performance. The full RF path also depends on strip dimensions, braid fold-back control, ferrule compression, dielectric recession, center conductor condition, and connector torque. One bad termination can make a low-loss cable perform like a cheaper one. That is especially true on smaller cable where process windows are tighter.
In automotive and dense box-build work, the cable decision also has to match the connector family and package geometry. A high-density RF layout using FAKRA and mini-FAKRA may accept slightly higher cable loss to gain routing density, service access, or cleaner breakout geometry. The right answer is the one that balances attenuation, assembly yield, and package reality together.
"The cheapest RF cable assembly is the one that meets loss, return loss, and bend requirements on the first build. A cheaper cable that forces rework, tuning, or field failures was never the low-cost option."
— Hommer Zhao, Technical Director
Common mistakes buyers make with coax loss charts
- Comparing cable loss at the wrong frequency band and assuming the ranking will stay the same everywhere.
- Using dB per 100 feet data without converting it to the actual installed length.
- Ignoring connector and adapter losses when the cable run itself already uses most of the margin.
- Choosing the thinnest cable that fits mechanically, then discovering the radio budget cannot tolerate the attenuation.
- Assuming a datasheet number applies to a poorly terminated custom assembly without RF test verification.
FAQ: coaxial cable loss by frequency and cable type
How much signal loss is acceptable in a coaxial cable run?
That depends on receiver sensitivity and system margin, but many practical RF assemblies try to keep cable loss below 3 dB to avoid throwing away roughly 50% of the power before connectors and adapters are counted. At 2.4 GHz, that limit can be reached in well under 20 feet with small cables such as LMR-100A.
Why does coaxial cable loss increase as frequency rises?
Loss rises because conductor loss and dielectric loss both increase with frequency. Skin effect pushes current toward the outer surface of the conductor, so effective resistance climbs as frequency goes up. That is why a cable that looks acceptable at 150 MHz can become a bad choice at 2.5 GHz or 5.8 GHz.
Is LMR-400 always better than LMR-240 or LMR-195?
Electrically, LMR-400 usually wins on attenuation, but it is not always the best assembly choice. Its 0.405 inch diameter and stiffer bend behavior can create routing problems inside compact enclosures, so many buyers accept a 2 to 4 dB penalty to gain easier installation with LMR-240 or LMR-195.
How do I convert dB per 100 feet into loss for my actual cable length?
Multiply the chart value by your run length divided by 100. If a cable is rated at 7.6 dB per 100 feet at 900 MHz, a 25 foot run will contribute about 1.9 dB. Then add connector, adapter, and mismatch loss to estimate the complete channel.
What is the difference between insertion loss and return loss in a coax assembly?
Insertion loss measures how much forward signal is lost through the path, while return loss measures how much energy reflects back because the impedance is not held close to 50 ohms or 75 ohms. A cable can have acceptable published attenuation and still fail the link because poor assembly drives return loss below about 15 to 20 dB in the working band.
Can RG-58 still be used in modern RF cable assemblies?
Yes, but usually for shorter and lower-frequency runs. The Times Microwave selection guide shows RG-58 around 16.5 dB per 100 feet at 900 MHz, versus 3.9 dB for LMR-400. That difference is often too large for antennas, cellular modems, GNSS, and Wi-Fi backhaul unless the jumper is very short.
Need help specifying an RF cable assembly?
If your team is balancing cable loss, connector family, bend space, shielding, and production risk, we can review the run length and target band before you lock the BOM. That is usually cheaper than finding the problem after first article builds.
Data note: chart values are typical published attenuation figures in dB per 100 feet from Times Microwave LMR selection data and LMR-100A datasheet values. Final cable-assembly performance should be verified on the finished interconnect at the working frequency band.