Many people refer to an end-fed wire of any length as a long wire antenna. This is incorrect. Just because a length of wire might be considered to be long because it’s the length of your garden, it’s not necessarily a long wire antenna.
Long-Wire Antennas (Longer Than One Wavelength)
A long wire, in the strict sense, is at least one wavelength long at the operating frequency, and often several. At that length it stops behaving like a dipole and develops gain and directivity along the line of the wire, which makes it a cheap way to get some gain if you have the space.
How the pattern changes with length
The longer the wire, the more lobes it has and the closer the two main lobes swing towards the line of the wire. A half-wave dipole radiates broadside in a figure-of-eight; a resonant wire N half-waves long has N lobes on each side, 2N in total, and the main lobes strengthen as they narrow.
Approximate figures for a resonant (standing-wave) long wire, based on the ARRL Antenna Book:
| Length (λ) | Main lobe angle from wire | Approx. gain over dipole |
|---|---|---|
| 1 | ~54° | ~0.5 dB |
| 2 | ~36° | ~1.5 dB |
| 3 | ~29° | ~2.3 dB |
| 4 | ~25° | ~3 dB |
| 6 | ~20° | ~4 dB |
| 8 | ~17° | ~5 dB |
Gain builds slowly, so you need several wavelengths before it is really worthwhile, and the wire has to be aimed, roughly, where you want to work.
Resonant vs terminated
- Unterminated (resonant/standing-wave): the far end is left open. This is the usual amateur version. Current reflects off the open end, so the pattern is bidirectional, with lobes towards both ends.
- Terminated (travelling-wave): a non-inductive resistor of roughly 400–600 Ω goes from the far end to ground. The reflected wave is absorbed, the pattern becomes unidirectional towards the terminated end, and the antenna becomes very broadband. The cost is that a good share of your power is burnt in the resistor.
On LF/MF, the low-slung receive version is the Beverage antenna, excellent for top-band and 80 m receiving.
Feeding it
- End-fed resonant wire: the end is at a voltage maximum, so impedance is high, typically low thousands of ohms. A 49:1 or 64:1 unun, or a parallel-tuned L-match/ATU at the feed point, handles this.
- Non-resonant "random" long wire: a 9:1 unun plus an ATU is common. Choose a length that is not close to a multiple of a half-wave on any band you intend to use, or the tuner will struggle.
- Ground/counterpoise: an end-fed wire needs something to push against: a ground system, radials or a counterpoise. Without one, the feeder braid and the shack become part of the antenna, giving RF in the shack. A common-mode choke at the feed point helps a lot.
- Centre-feeding with open-wire line (doublet-style) also works well, but the current distribution and pattern differ from the end-fed case.
Multiband behaviour
The same wire gets electrically longer as frequency rises: a wire 1λ on 40 m is roughly 2λ on 20 m and 4λ on 10 m. Pattern, lobe angles and gain all change band by band, so it may be a near-dipole on one band and a sharply lobed antenna on another.
Doing more with long wires
- Height: sets the vertical radiation angle, as with any horizontal antenna. Higher gives lower angles, better for DX.
- Sloping: sloping the wire, or terminating it, tilts the pattern towards the low end.
- V-beam: two long wires fed in antiphase, apex angle about twice the main lobe angle. The lobes add along the bisector, giving useful gain.
- Rhombic: four long wires in a diamond, usually terminated. Very high gain and broadband, but needs a lot of land.
Practical points
- Long spans build up a lot of static: fit a static bleed resistor or RF choke to ground, plus proper lightning precautions.
- Use hard-drawn copper or copper-clad steel so the wire doesn't stretch.
- Allow for sag and wind load at the supports.
I learned a lot from the old timers back in the 60s and 70s. They would chat away to each other on the 80 metre amateur band about aerials and aerial tuning units and I’d listen intently and take notes.
There’s more on end-fed aerials.