Building a Wire Antenna from Scratch: Wire, Insulators, Chokes, Feedline and Modeling Before You Cut
- Replies 2
- Views 37
- Created
- Last Reply
A better way to browse. Learn more.
A full-screen app on your home screen with push notifications, badges and more.
This thread is for the practical decisions that come up when you build a wire antenna from scratch rather than buying one. The example is a simple one: a half-wave dipole for 40 m, possibly extended later to a fan dipole or a ladder-line-fed doublet. The same questions apply to verticals and trap designs, so let's go through them in order.
Start with the wire. For a permanent span, 14 AWG hard-drawn copper or copper-clad steel is a common choice because it resists stretching. Soft-drawn copper stretches over time and the antenna slowly gets longer and resonates lower. Stranded insulated wire is easier to handle and coil, and 16 to 18 AWG is fine for shorter spans or portable use. Insulation adds capacitance along the wire, so an insulated dipole typically resonates a few percent lower than the bare-wire formula predicts. The usual starting point is about 468 divided by the frequency in MHz, which gives feet for the whole dipole (about 142.6 divided by f for metres). Treat that as a starting length only. Cut the wire long, then trim in small steps while measuring.
Insulators matter more at the ends than most people expect. Each end sees high RF voltage and mechanical load. UV-stable plastic, glazed ceramic, or properly rated acrylic or polycarbonate work well. A knotted loop of rope is not an insulator once it is wet and dirty. Use non-stretch rope such as polyester, or Dacron-type line, for supports, and consider a pulley or a weighted line at one end so tree movement does not snap the wire. At the center, relieve the strain so the coax weight does not pull on the solder joints. A solder joint on stranded wire becomes brittle at the edge of the solder, so make the mechanical connection first and seal it against moisture afterwards.
Then there is the feedline problem. A dipole is a balanced antenna, and coax is unbalanced. Without a choke or balun, current can flow on the outside of the coax shield, which distorts the pattern, lets RF into the shack and makes SWR readings change when you move the cable. A 1:1 current balun at the feedpoint is the usual fix. A choke made from several turns of the coax itself through a suitable ferrite core, typically type 31 or 43 material for HF, is a common alternative. The ferrite mix needs to suit the band, because the choke has to present high impedance at your operating frequency, not just something that looks like a lot of ferrite. For ladder-line feeds into a tuner, the balun usually goes at the tuner end, and the line should leave the antenna as symmetrically as possible.
On the vertical side, the key point is that the radial system is half the antenna. A quarter-wave vertical over a few elevated radials behaves very differently from the same element sitting on a handful of short wires lying on the soil. Elevated resonant radials, ideally at least four, are a reasonable compromise, and drooping them downward moves the feedpoint resistance toward 50 ohms. Ground-mounted systems usually want many more radials, and length matters less than count. Fiberglass poles make good vertical supports, but check that the pole is not carbon fiber, since carbon is conductive and can interact with the antenna.
For measuring, an antenna analyzer or a VNA at the feedpoint gives the clearest picture. Measuring SWR at the transmitter end with long, lossy coax can hide a bad match because the cable loss improves the apparent SWR. Also remember that a low SWR does not mean an efficient antenna. A dummy load has an excellent SWR and radiates nothing.
Before cutting any wire, you can model the design in EZNEC or MMANA-GAL. Both help you find approximate resonant length, feedpoint impedance, take-off angle and the effect of height over ground. The usual pitfalls are too few segments, a poor ground choice, and wires meeting at awkward angles. MMANA-GAL has an average gain test that is a useful sanity check on whether the model is behaving. Neither program knows about common-mode current on your feedline unless you add it to the model, so a clean-looking pattern is not a guarantee.
What are you building first, and which of these steps has caused you the most trouble: the mechanical side, the feedline, or getting the model to match reality?
Link to comment
https://www.hamradiobase.com/forums/topic/10834-building-a-wire-antenna-from-scratch-wire-insulators-chokes-feedline-and-modeling-before-you-cut/Share on other sites