Call and
Signal
Menu

Antennas and Feed Lines · Guide

Understanding Loss in Coaxial Feed Lines

A black coaxial cable curves across a plain surface, with a metal connector visible at each end.

A coaxial feed line loses some RF energy between a radio and an antenna. To estimate that loss, find the attenuation rating for the actual cable model at the operating frequency, then scale its stated length to the length of the run. The result is a nominal line-loss estimate—not a guarantee of how much power reaches the antenna. Mismatch and possible connection problems are separate questions, and a good SWR reading at the radio does not settle them.

What Feed-line Loss Measures

The American Radio Relay League (ARRL) describes a feed line, also called a transmission line, as the RF path between the radio and antenna. During transmission, that path carries power toward the antenna; during reception, signals picked up by the antenna travel through the same line toward the radio. ARRL notes that a feed line always loses some energy. Loss matters in both directions, even though examples often describe transmitted power because its starting point is easier to name.

Keep three ideas apart. Transmitter output is power at the radio end of the path. Power arriving at the antenna is what remains after losses along that path. How efficiently the antenna radiates is a different question. A feed-line estimate addresses the path to the antenna; it does not, by itself, answer the radiation question. Likewise, knowing the transmitter’s output does not tell you the power at the antenna without accounting for what happens between them.

Feed-line attenuation is commonly stated as a loss in decibels, or dB, per 100 feet. That is the rating convention ARRL describes. The “per 100 feet” part is essential: a figure for 100 feet is not automatically the loss for a shorter or longer installation. It gives you a length-based figure to scale once you know the route length.

Attenuation and SWR, or standing-wave ratio, answer different questions. An attenuation figure describes loss for a stated length of line under the specification’s conditions. SWR concerns matching in the RF path. It is tempting to read a low SWR at the radio as a broad verdict that the whole path has little loss, but that reading alone does not replace a cable-loss estimate. Nor does a cable’s published attenuation figure describe every matching condition in which the installed line might operate. ARRL’s tuner guidance makes the distinction important: a tuner can reduce SWR on the radio side while mismatch remains on the antenna side.

For a beginner comparing runs, the useful mental model is therefore a path with several distinct questions. What loss is expected from this model of cable over this length at this frequency? Is the installed line operating under conditions like those behind that rating? Is there a separate reason to question the connections? Answering the first question gives a useful baseline. It does not silently answer the others.

Choose the Specification for the Cable and Frequency

Start with the cable model, not a general impression that a cable is “low loss.” ARRL says loss at a given frequency can vary considerably among feed-line types. A number taken from one model’s specification cannot be assigned to another model simply because both are coaxial feed lines. Frequency must also be paired with the number: a manufacturer may give different attenuation ratings for the same cable at different frequencies. Finally, route length determines how much of a per-100-foot rating your run uses. These are three separate inputs—model, frequency and length—not three names for the same effect.

Times Microwave Systems’ LMR-400 datasheet provides a concrete example of how to read a specification. Its frequency row pairs 30 MHz with 0.7 dB per 100 feet, 150 MHz with 1.5 dB per 100 feet, and 450 MHz with 2.7 dB per 100 feet. Each attenuation entry belongs with the frequency in the corresponding position. Thus, if the example run uses LMR-400 at 150 MHz, the relevant listed starting figure is 1.5 dB per 100 feet—not the 30 MHz figure merely because it is smaller, and not the 450 MHz figure merely because it appears in the same table.

The progression also shows why “the loss of LMR-400” is incomplete without a frequency. In these listed entries, the attenuation figure is higher at 150 MHz than at 30 MHz, and higher again at 450 MHz. That observation is about these LMR-400 entries. It does not provide figures for another cable model or establish an installed run’s actual loss. For a different model, the relevant starting point is that model’s own frequency-specific specification. If a comparison involves several operating frequencies, the useful question is what rating applies at each one, rather than which candidate has the most attractive single number.

Read the conditions attached to the figures as carefully as the figures themselves. The LMR-400 datasheet states VSWR = 1.0 and an ambient temperature of 25°C (77°F) for its attenuation values. These are the conditions of the published rating, not a report that a particular station’s coax has a VSWR of 1.0 or is at that temperature. They make the values a conditional baseline for comparison. They do not turn the table into a measurement of a route after it has been installed and connected.

This distinction prevents two common shortcuts. One is carrying a figure across frequencies without checking its paired entry. The other is treating a manufacturer’s matched-condition rating as the complete loss of a real system. The first shortcut chooses the wrong baseline; the second asks that baseline to settle questions it was not given enough information to settle. Before doing any arithmetic, you should be able to name the cable model, the frequency associated with the chosen entry, the entry’s units and the conditions under which it was specified.

Scale the Rated Loss to the Route

Once the model-and-frequency figure is identified, the basic length calculation is straightforward:

Nominal run loss (dB) = rated loss (dB/100 ft) × route length (ft) ÷ 100.

In words, take the applicable loss per 100 feet and multiply it by the fraction of 100 feet in the route. This scales a length-based rating; it does not calculate transmitter output or directly predict power at the antenna. For the LMR-400 examples below, “nominal” continues to mean an estimate based on the datasheet figures stated at VSWR 1.0 and 25°C ambient.

Consider 50 feet of LMR-400 at 150 MHz. The datasheet’s 150 MHz entry is 1.5 dB per 100 feet, so the scaled estimate is 1.5 × 50/100 = 0.75 dB. For 100 feet of LMR-400 at the same 150 MHz, it is 1.5 × 100/100 = 1.5 dB. The 100-foot route uses the whole stated rating length; the 50-foot route uses half of it. In this length-scaling calculation, doubling the length from 50 to 100 feet doubles the numerical dB loss estimate from 0.75 to 1.5. Nothing about the cable model or the selected frequency changed in that comparison.

Now change both the frequency and the route length. At the listed 30 MHz entry, 50 feet of LMR-400 gives 0.7 × 50/100 = 0.35 dB. At the listed 450 MHz entry, 100 feet of LMR-400 gives 2.7 × 100/100 = 2.7 dB. The latter estimate is larger, but this comparison does not show that frequency alone—or length alone—accounts for the difference. It changed two inputs. To isolate length, compare runs at the same frequency as in the 150 MHz example; to understand a frequency change, keep track of which length is being compared. All four results are arithmetic using LMR-400’s specified entries, not measurements of four installed lines.

The route length in the calculation should be the length of cable in the proposed radio-to-antenna path, rather than an unexamined description such as “a nearby antenna.” A planned route may differ from a rough sense of how far apart its endpoints are. Until the length is known, the per-100-foot rating cannot produce a specific run estimate. Likewise, if the actual cable model is unknown, choosing LMR-400’s figure would substitute an example for a missing fact. If the operating frequency is unknown, there is no basis for selecting one of its frequency entries. These are missing inputs, not invitations to guess.

Terminations require a different kind of caution. The multiplication above uses a cable attenuation rating and a length; it contains no input describing the state of a connection. If a termination is uncertain, the arithmetic can still show what the selected nominal cable figure implies for the stated length, but it cannot settle whether that uncertainty adds a problem. Labeling the result “nominal LMR-400 line loss under datasheet conditions” keeps it useful without presenting it as a prediction of antenna-end power.

That is enough arithmetic for a first comparison. You do not need to convert these examples into watts to see that length and frequency change the estimate. More importantly, avoiding an unsupported power figure preserves the boundary between a rated cable calculation and a claim about the complete installed path.

Why a Good Radio-side SWR Can Leave Loss Unresolved

A tuner near the radio illustrates why the place where SWR is observed matters. According to ARRL, the SWR reduction produced by an antenna tuner occurs between the tuner and the radio. The mismatch between that tuner and the antenna can still exist. Consequently, a low radio-side SWR reading after tuning does not establish that the coax on the antenna side is operating under the matched condition assumed by a datasheet attenuation figure. It is information about one side of the tuner, not a loss measurement of the entire feed line.

This does not make the nominal calculation useless. It answers a narrower question: what length-scaled attenuation follows from a model’s rating under its stated conditions? For LMR-400, those stated conditions include VSWR 1.0. ARRL separately cautions that mismatch and high SWR remaining between a radio-side tuner and the antenna can translate into high losses in coaxial cable. Thus, matched-condition attenuation and potential extra loss with mismatch should not be collapsed into one number. The cited information supplies a warning, not a numerical correction that can be added to every run.

Imagine two descriptions of the same proposed route. One says, “Its cable specification implies 0.75 dB for this length and frequency under the stated conditions.” The other says, “The radio shows low SWR with a tuner.” Neither statement proves the other. The first comes from a model-specific attenuation entry and route-length arithmetic. The second may reflect matching on the short path between tuner and radio. To claim low loss along the antenna-side coax, more would need to be known about the condition of that side of the path. This is an explanatory comparison, not a test result or a diagnosis of any particular station.

Tuner location changes which part of the path receives the matching action. ARRL describes remote automatic tuners that do the matching at or close to the antenna. That is materially different from using a tuner near the radio and then assuming its radio-side SWR reading describes the coax beyond it. The placement distinction helps frame a station design question, but it is not a promise that a remote tuner eliminates ordinary cable attenuation, makes every mismatch disappear or fixes an installation fault. ARRL’s description establishes where the matching occurs, not a zero-loss outcome.

Connections belong in the assessment without being assigned a made-up dB allowance. A questionable termination is an installation question; the LMR-400 table does not contain a value for it, and a low reading on the radio side of a tuner cannot identify its condition. If connections are uncertain, retain the nominal line calculation as a baseline while treating the installed path’s total loss as unresolved. That is more informative than either discarding the specification entirely or pretending it covers a connection whose condition is unknown.

A coaxial cable hangs from the center of a dipole antenna, whose two wires extend in opposite directions.

Use the Estimate to Choose a Practical Run

ARRL gives an example that makes the distinction between a simple expectation and antenna-end power concrete: 100 W transmitted into 80 feet of RG-8X coax, feeding an antenna with 1.4:1 SWR at 146.52 MHz, results in 46 W of actual power at the antenna. Keep every condition attached to that example. It concerns RG-8X, that length, that frequency and that stated antenna SWR—not LMR-400, a generic 80-foot cable or every antenna with 1.4:1 SWR. It shows why a seemingly modest SWR figure alone cannot establish that little power is lost before the antenna. It does not provide a basis for assigning the whole difference between 100 W and 46 W to matched-condition cable attenuation.

Cable choice is consequently conditional on the route. ARRL describes an inexpensive line such as RG-58 as usable where only a short length is needed. It also calls 100 feet of RG-58 at about 440 MHz an extraordinarily bad choice. Those statements are not conflicting ratings of RG-58 as universally good or bad. They concern very different lengths, and the warning names an operating frequency. A cable that is workable in a short run should not inherit that verdict when the proposed route becomes much longer at the stated frequency. Nor does ARRL’s warning supply a numerical loss figure to apply to every RG-58 installation.

Nominal loss is not the only route consideration. ARRL notes that some low-loss feed lines are rigid and difficult to bend. A prospective path therefore has to be considered alongside a candidate’s attenuation figure: the attraction of a lower rated loss does not make the cable’s physical characteristics irrelevant to that path. This is a trade-off to recognize, not a universal ranking of cables or a substitute for information about the actual route.

For your own comparison, first fill in the facts that the examples cannot supply for you: Which frequencies will the run serve? What is the actual cable model? How much cable does the intended route use? For each candidate, look for its own attenuation specification at the relevant frequency and note the conditions attached to that specification. Scale the figure to the route length, keeping each result labeled as a nominal estimate. If one candidate lacks a known model, applicable frequency entry or credible route length, its precise-looking calculated loss would be a guess rather than a comparison.

Then ask the questions the multiplication leaves open. Is the coax expected to operate with a mismatch that makes the matched-condition figure an incomplete account? If a tuner is involved, is its low-SWR indication on the radio side while the antenna-side line remains in question? Are any terminations uncertain? Does a physically less flexible candidate suit the proposed path? These questions do not each need an invented dB value to matter. They identify where a choice based solely on the smallest published attenuation entry would overstate what is known.

The useful verdict is conditional: for a known model, frequency and length, the matching datasheet entry can give a clear nominal line-loss comparison. A short run and a long run need not lead to the same cable decision, as ARRL’s RG-58 contrast illustrates. But the calculated dB figure remains a baseline under stated conditions, not proof of antenna-end power, sound connections or low loss beyond a radio-side tuner. Keep that boundary visible, and the specification becomes a practical tool rather than a misleading final answer.