Handbook, chapter 10. The feeder, SWR and matching
The cable from the transceiver to the antenna is not a wire. A wire has a resistance you can measure with a meter; a cable has a characteristic impedance, which a meter cannot see at all.
Fifty ohms in coax is not the resistance of its centre conductor. It is the ratio of voltage to current in the wave running along it, and it is set by the geometry: the diameters of the conductor and the braid and the dielectric between them. You cannot measure it with a meter — the meter shows a few ohms in the conductor and infinity between conductor and braid, and neither of those numbers is fifty.
What actually happens in the line
As long as the load equals the characteristic impedance, the wave reaches the antenna and all of it goes in. The moment the impedance differs, part of it is reflected back and adds to the wave running forward. Peaks and troughs of voltage appear along the cable and stay where they are. Hence the name: a standing wave.
SWR is the ratio of a peak to a trough, and nothing more. At 50 ohms the line is flat and the SWR is one. At 150 ohms the voltage varies threefold — an SWR of 3.
⚠️ What «came back» means. In the drawing the load is purely resistive and the line lossless, and the figure labelled «back» is the power the antenna did not accept on the first pass. In a lossless line it returns to the transmitter and part of it goes out on the air again, so the real loss is less than that number. The real damage is described below.
A real cable, through its loss, also «improves» the SWR at the input, which is exactly why it has to be measured at the antenna rather than at the transceiver: there it always looks better than it is.
How much SWR actually takes
This is where beginners take fright for nothing. The loss from the reflection itself is small:
| SWR | Power reflected | In decibels | In S-points |
|---|---|---|---|
| 1.5 | 4 % | −0.18 dB | 0.03 points |
| 2.0 | 11 % | −0.51 dB | 0.09 points |
| 3.0 | 25 % | −1.25 dB | 0.21 points |
The last column is the one the table is here for. One S-point is six decibels. So even an SWR of 3.0 costs a fifth of a division — the other station's needle will not move, and you will never hear it by ear.
An SWR of 1.5 does not need curing. At all. The effort spent turning it into 1.2 is better put into half a metre of height.
The two real problems are different, and neither is about signal level. First: a modern transceiver reduces its own power when it sees an SWR above two, to protect the output transistors. Second: the reflected wave travels back along the cable and is lost in it a second time, so a poor cable with a poor SWR eats far more than the table suggests.
The meter in the shack reads better than reality
The trap everyone falls into, and it is arithmetic.
The SWR meter sits by the transceiver while the reflection is born at the antenna. Before it gets back to the meter the reflected wave travels the whole cable again — and is attenuated in it exactly as the forward wave was. The meter sees it already weakened and honestly computes from what it sees.
Take a real cable. Thirty metres of RG-58 on twenty metres is about two decibels one way, so four for the round trip:
| SWR at the antenna | What the meter in the shack shows |
|---|---|
| 2.0 | 1.54 |
| 3.0 | 1.94 |
So an antenna that really has a three looks in the shack like a perfectly respectable two. The longer and worse the cable, the more prettily the meter lies — and in the limit, on a very long cable, any antenna reads almost one, because nothing comes back.
Two practical conclusions follow. A perfect one on the meter is a reason to check not the antenna but the cable and the meter itself. And where you can, measure SWR at the antenna, not in the room.
Why a tuner does not cure a bad antenna
A tuner is a matching device, and it does exactly one thing: it shows the transceiver fifty ohms at its input.
It sits in the shack, while the reflection happens at the antenna. Everything between the tuner and the antenna carries on as before: the standing wave in the cable stays, the loss in it stays. The transceiver is happy now and gives full power — it is just that part of it heats the cable.
So the right order is: the antenna first, then the cable, and only then the tuner — and that is for working a neighbouring band, not for fixing the main one.
Loss: the thing remembered last
Cable loss grows with frequency. The same RG-58 that takes almost nothing at 7 MHz eats a noticeable share at 145 MHz — and takes it twice, on transmit and on receive.
A practical rule: on HF almost anything will do, on VHF it will not. Thirty metres of thin cable to an antenna at 430 MHz turns a hundred watts into a few tens, and no amplifier puts that right.
Connectors: which goes where
The question «which cable and which connector» comes up before the question about SWR, and the answer is nowhere to be found.
| Connector | Where it is used | What to know |
|---|---|---|
| PL-259 / SO-239 | HF transceivers, most amateur gear | Has no characteristic impedance of 50 Ω of its own. On HF that does not matter; above 200 MHz it does |
| N | VHF and anything outdoors | Stands up to damp and to high frequencies. If the antenna is on a mast, this is the one |
| BNC | test instruments, handhelds | Bayonet, mates with a quarter turn. Convenient, but not for outdoors |
| SMA | handheld radios, SDR receivers | Small, threaded. Comes in normal and reverse — easy to confuse, and they do not mate |
The main rule: fewer adapters. Each one is a little loss, a little mismatch and one more place for water to get in.
Next — chapter 11, the transceiver: which of those knobs are actually worth turning.