GazLabs › Homebrew Antennas › Antenna calculators
Antenna Calculators
Starting lengths for the most popular homebrew antennas. Every result is shown in metric and imperial, and the shortening factors are editable so you can match your own wire or tube.
One frequency for every calculator
Each calculator starts at a sensible default for its antenna. Apply one frequency here to compare every design on the same band, then adjust any single card as you like.
Half-wave dipole
Polarisation: horizontal as a flat-top · vertical if hung as a vertical wire
Feed-point impedance about 50–75 Ω at a decent height; a 1:1 current balun is good practice.
Inverted-V dipole
Polarisation: mostly horizontal, with some vertical from the sloping legs
An inverted-V is usually cut a few percent shorter than a flat dipole and presents close to 50 Ω. Keep the ends out of reach: they carry high RF voltage.
End-fed half-wave (EFHW)
Polarisation: follows the wire (horizontal, mixed as a sloper, or vertical)
The end of a half-wave wire sits at a few thousand ohms, so it needs a high-ratio matching transformer. Fit a common-mode choke on the coax, since the system relies on the coax shield and a short counterpoise.
Full-wave loop (square, delta, round)
Polarisation: set by the feed point (middle of a horizontal side = horizontal, middle of a vertical side = vertical)
Perimeter factor 1.0217 is the classic 1005/f (ft) rule. A resonant loop sits around 100–120 Ω; a quarter-wave section of 75 Ω coax brings that down to roughly 50 Ω.
Two-element quad
Polarisation: set by the feed point (bottom side = horizontal, side = vertical)
Classic 1005/f driven and 1030/f reflector (in feet) perimeters. A spacing of 0.15–0.25 λ is typical; feed the driven element at the bottom corner for horizontal polarisation or the side for vertical.
Ground-plane vertical
Polarisation: vertical
Four radials drooping at about 45° bring a ¼-wave vertical close to 50 Ω. A ⅝-wave radiator needs a matching coil or stub at the base, so tune that on the analyser.
J-pole
Polarisation: vertical
Use about 0.95 for bare tube or wire and roughly 0.95–0.97 for window ladder line. The tap position is a starting estimate: slide it for the lowest SWR.
VHF / UHF collinear (stacked half-waves)
Polarisation: vertical
This is the ideal in-phase stack: the radiating sections are separated by phasing sections (coils, stubs or coax sleeves) that turn the currents the right way round. The gain figures come from the pattern simulator in free space and real antennas fall short by roughly 0.5 to 2 dB. Commercial 5/8-wave collinears are built differently, so use the manufacturer's dimensions for those.
On a metal mast? A collinear that is fed at the bottom with no radials uses the coax shield and mast as part of the antenna, which is a common cause of high SWR at height. The mast SWR check shows the effect and the cure.
Coax choke designer (current balun for a collinear or any antenna)
Stops the coax shield acting as part of the antenna
For Messi & Paoloni cables the standard and single static bend radii are the maker's own figures for that cable (for example Airborne 10 is 120 mm standard and 60 mm for a single bend), and repeated bending is 20 times the diameter. For other cables the standard is 10 times the diameter and a static choke coil uses 5 times. A single static bend applies indoors above 15 °C, and a multi-turn coil is a judgement call, so use the standard figure outdoors or in the cold. Diameters for Extraflex Bury, Hyperflex 10 and 13, Pota-Flex and Stealth Flex are catalogue sizes, so overtype the diameter if yours differs. The bend radius rule is a multiple of the cable diameter, and the minimum bend radius in millimetres is calculated from it for the coax you pick. The smallest former is then twice that radius less one cable thickness, because the radius is measured to the centre of the cable. The standard figure is 10 times the diameter and has not changed since 1982. Attenuation testing by Lectromec found coax bent on a mandrel of only 6 times its diameter still stayed within the acceptable limits, so you can enter 6 to relax it, or divide the radius on your cable's data sheet by its diameter to get your own multiple. Cable diameters are typical catalogue values and vary a little between makers, so check your own cable and overtype the diameter if it differs. A coil of coax is an inductor in the shield path, so it only needs to be a high impedance at your frequency. It works best near its self-resonance, where the coil and its own capacitance form a parallel tuned circuit. The numbers assume a close-wound single-layer coil, a Q of 30 and a self-capacitance from Medhurst's formula. Real coax coils usually have more capacitance than a bare-wire coil, so the true peak sits a little lower in frequency. Treat the turns as a starting point and check them with an analyser or NanoVNA, adding or removing half a turn to move the peak onto your frequency.
Fitting it
- Wind the coax neatly in a single close layer on a plastic former such as a pipe or bottle, and tie or tape it so the turns cannot move.
- Put the choke right at the antenna feed point, within a quarter wavelength of it, and keep the coil clear of the mast and other metal.
- For a wideband or HF choke, or where a coil is impractical, use a stack of tube-type ferrite cores or beads (mix 31 or 43 is commonly used at these frequencies), because one core gives only a few hundred ohms. Several chokes in series add together.
- Choosing the mix: mix 31 gives the most impedance from HF up to about 250 MHz, mix 43 is similar but a little lower at 2 m, and mix 61 holds up better at 70 cm and above. For a dual-band colinear, set Band mode to 2 m + 70 cm and the designer compares single-mix stacks with a mixed stack, so each material works on the band it suits.
- Do not coil the cable to cure a poor match in the antenna itself: a choke only stops current on the outside of the coax.
See what a choke does for an antenna on a mast in the mast SWR check.
Quarter-wave sleeve choke (bazooka balun)
A tube over the coax, shorted to the braid at the far end, that stops current on the outside of the braid
The sleeve and the cable braid form a short coaxial line that is shorted to the braid at the bottom end and open at the top, with the open end towards the antenna. A quarter of a wavelength long, that line looks like a very high impedance from the open end, so current cannot travel down the outside of the braid. The designer treats the cable jacket and the air gap as two dielectric layers in series to find the velocity factor, then shortens the quarter wavelength by that factor. It does not model the end effects at the open end, joint resistance or the loss in the tube, so the impedance figures are lossless estimates, and a real sleeve peaks at a few kohm. Cut the tube a little long, check on an analyser or NanoVNA, and trim for the best choking at your frequency.
Building it
- Slide the tube over the coax before the connector goes on, or choose a tube wide enough to pass over it. The shorted end goes at the bottom, away from the antenna, and the open end sits right at the feedpoint.
- Expose a ring of braid at the bottom by removing a short band of jacket, and join the tube to the braid all the way round, soldered if the tube is copper or brass, or with a clamp or a ring of foil for aluminium. Avoid a long pigtail, because that adds inductance and moves the tuning.
- Keep the tube concentric on the cable with foam, PTFE or rubber spacers. If the coax lies to one side the velocity factor and impedance change a little.
- Seal the open top end with self-amalgamating tape or a boot, so water does not collect in the gap, and keep the tube clear of the mast and other metal.
- One tube cut for 2 m is also about three-quarters of a wavelength on 70 cm, so it chokes both bands. Check the table above for the bands you use.
Off-centre-fed dipole (OCFD / Windom)
Polarisation: horizontal as a flat-top · follows the wire if sloped
The antenna is a half-wave wire for the lowest band, fed about a third of the way along. A 4:1 (sometimes 6:1) current balun is the usual match, and a choke on the coax is a good idea. In a free-space model the feed impedance is about 100 Ω on the fundamental and similar on the second and fourth harmonics, but varies with height, so measure the result.
Folded dipole
Polarisation: horizontal as a flat-top · vertical if hung vertically
Two parallel conductors joined at both ends, fed at the middle of one. The feed impedance is roughly four times a plain dipole (about 290 Ω), so a 4:1 balun gives a good match. The default factor suits bare wire or tube; solid 300 Ω twin-lead has a lower velocity factor and needs to be shorter, so cut long and trim.
Extended double Zepp (EDZ)
Polarisation: horizontal as a flat-top
A centre-fed wire 1.28 λ long (0.64 λ each side), fed with open-wire or ladder line to a balanced tuner. The length is not critical for resonance: it is chosen for gain, which is roughly 2 to 3 dB over a half-wave dipole broadside to the wire in the model, but the beam is narrower. The feed point is reactive, so you will need a tuner.
Small magnetic loop
Polarisation: vertical in its own plane when upright · horizontal when lying flat (see the simulator)
A single-turn circular loop tuned with a variable capacitor. The figures use the standard small-loop formulas for inductance and radiation resistance, plus the skin-effect resistance of copper tube. Efficiency, bandwidth and capacitor voltage are estimates: joint and capacitor losses are the usual weak point, so keep the extra-loss figure honest. The formulas are best when the circumference is below about an eighth of a wavelength and are only roughly right up to a quarter.
Three-element Yagi
Polarisation: horizontal or vertical, set by how the elements are mounted
Starting lengths for thin wire elements at close spacing (0.15 λ). In the model this gives about 8 dBi in free space with a feed impedance near 25 Ω, so use a hairpin, gamma or 4:1 balun match. Tube elements are usually trimmed shorter by a percent or two: use the factor, and tune on an analyser. Front-to-back is very sensitive to exact element lengths.
Coax and velocity factor
Velocity factors are typical values; check your cable's datasheet or measure it with an analyser.
SWR and reflected power
Enter the SWR you measured at the antenna (1 is a perfect match). The feedline loss is the matched loss of your whole cable at this frequency; leave it at 0 for an ideal line or a measurement taken at the antenna. A lossy cable hides a poor match: the SWR at the radio looks better than the SWR at the antenna, and the cable absorbs much of the reflected power. Pick a Messi & Paoloni cable and a length to fill the loss in from the catalogue figures (matched loss at 20 °C, interpolated between the listed frequencies and scaled with the square root of frequency outside 28 to 2400 MHz), or choose Other and type the loss from your own cable's data sheet.
Vertical, horizontal or both?
Polarisation is the orientation of the wave's electric field, and it follows the orientation of the radiating part of the antenna: a horizontal wire gives horizontal polarisation, a vertical whip gives vertical. It matters most on line-of-sight VHF and UHF paths, where a mismatched pair of antennas is often quoted as losing up to about 20 dB in the ideal case. Reflections from buildings and terrain usually soften that in real life, but a mismatch is still worth avoiding. On HF skywave paths the ionosphere scrambles polarisation, so height above ground and take-off angle matter far more than which way the wire points.
| Antenna | Polarisation | How to change it |
|---|---|---|
| Half-wave dipole | Horizontal as a flat-top; vertical if hung as a vertical wire | Orientation of the wire |
| Inverted-V | Mostly horizontal, with a vertical component from the sloping legs | Wider apex angle gives more horizontal |
| End-fed half-wave | Follows the wire: horizontal, mixed (sloper) or vertical | Orientation of the wire |
| Full-wave loop and quad | Set by the feed point: horizontal or vertical | Move the feed to a horizontal or a vertical side |
| Off-centre-fed dipole | Horizontal as a flat-top; follows the wire if sloped | Orientation of the wire |
| Folded dipole | Horizontal as a flat-top; vertical if hung vertically | Orientation of the wire |
| Extended double Zepp | Horizontal as a flat-top | Orientation of the wire |
| 3-element Yagi | Horizontal or vertical, set by the elements | Rotate the elements about the boom |
| Small magnetic loop | Vertical in its own plane when upright; horizontal when lying flat | Stand it upright or lay it flat |
| Ground-plane vertical | Vertical | Fixed |
| J-pole | Vertical | Fixed |
| Hentenna | Horizontal when upright; vertical when turned 90° on its side | Rotate the loop in its own plane |
Compromise installation
- Follow the mode you use most. On VHF/UHF, SSB, CW and weak-signal digital modes such as FT8 are conventionally horizontal, while FM voice, repeaters, packet and APRS are usually vertical. Check what local operators in your area use.
- Slant at 45°. Tilting a hentenna, loop or quad to 45°, or using a sloping wire, gives a mixed signal that is theoretically about 3 dB down against either polarisation. That is a modest, predictable loss compared with a full cross-polar mismatch.
- Use two antennas. A vertical ground plane or J-pole for FM and a horizontal dipole or hentenna for SSB, switched at the shack with a coax switch, is simple and cheap.
- Choose loop feed deliberately. A wire loop can be re-fed from the bottom side (horizontal) or a vertical side (vertical) later with little extra work, so you can test both.
- HF: keep it simple. A horizontal dipole or inverted-V, a quarter to half a wavelength up, is a good all-round compromise. A vertical only works well with a decent radial or ground system.
- Mind the surroundings. Keep any antenna clear of metal masts, guttering and wiring, ideally on a fibreglass or PVC mast section, because nearby metal shifts tuning and distorts the pattern. Indoors or in a loft, reflections make polarisation less predictable, so the compromise matters less.
Antenna calculator FAQ
Why is a dipole shorter than half a wavelength?
End effect and the slightly slower wave speed along the wire make a resonant dipole about 5 % shorter than half the free-space wavelength, which is where the shortening factor comes from.
What is velocity factor?
It is the fraction of free-space speed at which a signal travels in a cable. A quarter wave of solid-PE coax is therefore about 66 % of a free-space quarter wave.
Why does the Yagi have such low impedance?
Parasitic elements close to the driven element lower its feed resistance, typically to 20–30 Ω for a close-spaced three-element design. A hairpin, gamma match or 4:1 balun then brings it back to 50 Ω.
Why is a small magnetic loop so narrow-banded?
Its radiation resistance is tiny compared with its reactance, so the Q is very high, often over 1000. That gives a bandwidth of only a few kHz, so it needs retuning when you change frequency, but the same high Q gives good rejection of out-of-band signals.
What SWR is acceptable?
Below about 1.5:1 reflects under 4 % of the power and is more than good enough. Up to 2:1 reflects 11 %, which costs only about half a dB. Most transmitters need protection above roughly 2 to 3:1, and the real concern at high SWR is the stress on the transmitter and the loss in the cable, not the power that bounces back.
Does reflected power get wasted?
Not entirely: reflected power travels back down the feedline and, apart from the cable loss, is reflected again at the transmitter or tuner and sent to the antenna. That is why the loss from a mismatch is small at modest SWR, but a long, lossy cable at a high SWR can absorb most of the power.
Do I need to trim after cutting?
Yes. Calculators give a starting length. Cut slightly long, check resonance with an analyser, then trim a little at a time.
Also see the hentenna calculator and the homebrew antenna guide.