Most amateurs think about feedline once. They buy a roll of RG-58 from Amazon, run it from the shack to the antenna, and never think about it again. That’s fine on 40m with a 30 ft run. It’s a mistake on 2m with a 100 ft run, where the same cable will eat about 6 dB of your transmit power before it reaches the antenna — three quarters of your signal gone before it ever sees the radiator.

I wanted a calculator that made this trade-off visible — not just at HF, not just for coax, and not assuming everyone runs a perfect 1:1 match.

The matched-loss model

Cable loss has two components. Conductor loss scales with the square root of frequency — it’s the skin effect, where high-frequency current crowds onto the outer surface of the centre conductor and the inner wall of the shield. Dielectric loss scales linearly with frequency — it’s the energy absorbed by the plastic insulation between the two conductors as the field flips back and forth.

That gives a clean two-term model:

loss per 100 ft (dB) = k₁ × √f + k₂ × f

Each cable has its own k₁ and k₂. RG-58 has high values for both (lossy on every band). LMR-400 has much lower k₂ thanks to a foam dielectric. Heliax hardline has the lowest of both because the centre conductor is thick copper tube and the dielectric is mostly air.

The constants in the calculator are fitted to manufacturer data — Times Microwave for the LMR series, Belden for RG-58 and RG-8X, CommScope for Heliax, SSB Electronic for Aircell 7 and Ecoflex 10/15, and ARRL Antenna Book Table 24-1 for the open-wire and ladder lines. They’re typical of new, dry cable in still air. Connectors add a fraction of a dB per pair. Wet ladder line adds a lot.

SWR makes it worse, but not the way most people think

A 3:1 SWR at the antenna doesn’t just mean “you’re losing some power to reflection.” Reflected power doesn’t vanish — it travels back down the cable, hits the transmitter, and most of it gets sent back up again. That’s another pass through the cable, picking up more loss along the way. On a lossy line, those extra passes add up fast.

The standard formula for total loss including SWR effects is:

ρ = (SWR − 1) / (SWR + 1)
a = 10^(matched_dB / 10)
total_dB = 10 × log₁₀[(a² − ρ²) / (a × (1 − ρ²))]

It’s not pretty, but it’s the textbook ARRL form. When SWR is 1.0, ρ is 0 and the formula collapses back to the matched-line loss. As SWR climbs, the extra loss grows — and it grows faster on a lossier cable.

There’s a related effect that catches people out. Long lossy cable runs flatten the SWR seen at the radio. The reflected wave gets attenuated coming back, so the SWR meter shows something closer to 1:1 even though the antenna is badly mismatched. This is exactly why a coil of RG-58 makes a passable dummy load — it’s not matched, it just absorbs everything before the reflection can return. The calculator surfaces this as “SWR at transmitter” so you can see the gap between what your radio sees and what’s actually happening at the feedpoint.

Why ladder line still matters

The hidden value of ladder line — 450Ω window line or DIY 600Ω open-wire — is that its matched loss is so low that even a 4:1 SWR barely costs you anything. Run the numbers in the calculator: 100 ft of 450Ω window line at 14 MHz with a 4:1 SWR shows about 0.2 dB total loss. The same length of RG-58 at the same frequency and SWR is closer to 3 dB.

Half an S-unit gets dismissed by people who only chase strong signals. Tune across the band sometime and look at where stations actually sit on the meter — most are within a few S-units of the noise floor, not above S9. Losing 2 S-points doesn’t matter to the kilowatt station running stacked Yagis; they engineered themselves into your S-meter on purpose. For everyone else, a few dB of feedline loss and a touch of QSB is the difference between completing a contact and never quite making it through.

This is why experienced operators ( and my legendary friend GM3TEN ) love tuned doublets fed with open wire. You could feed a single antenna on every band, accept whatever wild SWR appeared on the line, and let an antenna tuner at the shack handle the match. The line itself didn’t care.

The line itself didn’t care, but the tuner does. Open-wire and ladder line are balanced, and they hand you whatever impedance the antenna gives them on each band. Getting from there to a 50Ω unbalanced rig needs a matching network — typically a balanced antenna tuner, or a 4:1 balun feeding a regular tuner. Every component in that chain dissipates some of the power as heat. A link-coupled or LC tuner is far more efficient than a broadband ferrite balun, but nothing is lossless. Run serious power and a thermal imaging camera pointed at your tuner is sobering — the energy you “saved” with low-loss line ends up cooked off in the matching network if the design isn’t right for the band. The honest move is to engineer the system around the bands you care about most and accept that the others will be a bit ropey.

One correction on “the line itself didn’t care” — GM3TEN rightly pointed out that a mismatched feeder doesn’t just carry the load impedance unchanged, it transforms it. The impedance seen at the input of a transmission line depends on the load impedance, the line’s characteristic impedance, and the electrical length of the line. The extreme case makes this obvious: a quarter-wave line feeding an open circuit looks like a short circuit at the transmitter end, and vice versa. This isn’t just a curiosity — it’s exploitable. A carefully chosen length of 75Ω coax can match a 150Ω antenna to a 50Ω transmitter with no tuner at all (worked example here). Quarter-wave shorted stubs make sharp notch filters — useful when running two transceivers at the same site and needing to suppress cross-band interference without a commercial duplexer (stub filter examples here). Credit to GM3TEN for the nudge.

The other catch is that ladder line hates being near anything — wet wood, metal masts, your house — and the published figures assume free space and dry conditions. Treat the numbers as a best case.

The cable comparison panel

The most useful part of the calculator turned out to be the comparison panel. It shows the same frequency, length, power, and SWR across all coax types side by side — from RG-174 at the lossy end through to Heliax at the other. The difference at HF is modest. The difference at VHF is dramatic. Drop the frequency to 7 MHz with 100 ft and most cables are inside about a dB of each other. Bump it to 146 MHz and watch RG-58 fall off a cliff while LMR-400 barely flinches.

That’s the kind of comparison that’s hard to internalise from a single number, but obvious once you see it laid out.

What it doesn’t model

The calculator assumes new cable in still air. It doesn’t account for temperature drift (loss increases a few percent at higher conductor temperatures), connector losses (typically 0.1–0.3 dB per pair), water ingress in old coax (potentially several dB), or the proximity effects on ladder line. For ham use, those caveats are footnotes — the headline numbers are accurate enough to make the right cable choice.

It also stops at single-band loss. Multi-band antennas fed through a tuner present different SWRs at different frequencies, and the loss varies accordingly. The calculator handles each band independently — pick the band you care about and run the numbers.

Try it

The calculator is live at skipzone.co.uk/tools/coax-loss-calculator. Try the same setup on RG-58 and LMR-400 at 2m, then compare both to 450Ω window line on 80m with a 4:1 mismatch. The point of the tool isn’t to give you a number — it’s to show you, at the bands you actually operate on, which trade-offs are worth caring about and which aren’t.