Wire Antennas in a Small Garden
Wire Antennas in a Small Garden
A small garden does not need a magic antenna name. It needs an honest map of the available geometry, return paths, feed line, losses, safety limits and operating goals.
People often ask me for “a dipole for 40 through 10” or “an 80-through-10 end-fed” that will fit a small garden. I understand the attraction. Pick a familiar label, buy or cut the wire, connect coax and hope the handbook drawing survives contact with the house, fence, tree, wet soil and patio door.
Sometimes it does. More often, the useful design begins one step earlier: what complete current-carrying structure can this site safely support, and what job must it do?
In a garden, you are not installing only a wire. You are installing the wire, feed transition, feed line, return path, supports, nearby conductors, matching network and receiver environment. The garden is not an unfortunate footnote. It is part of the boundary conditions.
Start With the Garden, Not the Antenna Label
Before choosing a topology, draw the site. I want the available endpoints, maximum sag, support heights, feed-line route, building entry, trees, gutters, fences, utility lines and places people can reach. Then I want the operating goal:
- Which bands and portions of those bands matter?
- Is the priority regional coverage, a particular DX direction, receiving signal-to-noise ratio, digital duty cycle, rapid band changes or mechanical simplicity?
- What transmitter power, duty cycle and tuner range are actually available?
- Can a balanced line reach the operating position, or must the installation use coax?
- Where can high RF voltage, strong current, weather exposure and common-mode current be kept under control?
“Small” must also be measured in wavelengths. The same ten-metre span is a severe constraint on 80 m, a shortened or folded geometry on 40 m and a much less restrictive space on 10 m. Height has the same electrical meaning: a support that is low in wavelengths on one band may be moderate on another.
Joeri’s short version: draw the available geometry, choose the wanted current path, then select the feed and matching method. Reversing that order is how a product name becomes an accidental antenna design.
The Handbook Drawing Is a Starting Model
A clean drawing may assume a straight conductor, a defined feedpoint, an ideal source, known ground, no house, no fence and no current on the outside of the feed line. Those assumptions are useful because they let us understand one mechanism at a time.
The installed system adds:
- bends, slopes, unequal leg heights and finite conductor diameter;
- ground with frequency-dependent conductivity and permittivity;
- metal and dielectric structures that can carry induced current or change coupling;
- a feed transition and feed line with loss, transformation and possible exterior current;
- matching components with finite Q, voltage, current and thermal limits; and
- local noise sources coupled through both the antenna’s wanted mode and unintended cable paths.
ITU-R P.527-6 describes the electrical properties of the Earth’s surface as material-, moisture-, temperature- and frequency-dependent. That is why “average ground” is a model input, not a promise about one garden. The effect on an antenna still depends on where its fields and return currents interact with that ground.
Choose a Current Path, Not a Reputation
These familiar wire antennas can all be useful in restricted space. None comes with installation-independent behaviour.
| Topology | Why it may fit | What must be designed and verified |
|---|---|---|
| Inverted-V dipole | One high central support can fit a longer wire into a shorter horizontal span. | Included angle, leg clearance, end height, feed-line departure, environmental symmetry, band-specific current distribution and any required common-mode control. |
| Fan dipole | Several resonant wire pairs can share one feedpoint without traps. | Wire interaction, spacing, support geometry, tuning order, feed transition, current on each branch and final installed impedance on every band. |
| Off-centre-fed dipole | One wire may present usable impedances on several related modes. | Actual complex impedance per band, transformer ratio and loss, feed-line voltage/current, environmental imbalance, exterior-coax current and tuner range. |
| End-fed half-wave system | The feedpoint can sit near a convenient property boundary while the wire follows the available route. | Installed high terminal impedance, transformer loss and stress, a deliberate RF return path, coax participation, choke boundary, wire voltage and multiband current distribution. |
| Non-resonant doublet | One wire plus a low-loss balanced line and suitable tuner can cover several bands without forcing every band to be resonant. | Wire and line lengths, line impedance and routing, tuner range and loss, high-voltage points, balance, common-mode conversion and mechanical access. |
| End-fed non-resonant wire | It may use an awkward perimeter where a centred feed is impossible. | Measured load impedance on every band, intentional counterpoise or return path, transformer or tuner topology, feed-line loss, voltage/current stress and common-mode current. “Random” is not a design specification. |
I do not start this comparison by asking which one is best. I ask which complete system fits the geometry, can be kept safe, has acceptable loss, and can be measured at meaningful reference planes.
Multiband Does Not Mean Broadband
A multiband antenna can have several distinct resonances or several frequencies where a matching network can transform the installed load. That is not the same as maintaining a useful match, efficiency and pattern continuously across the spectrum between them.
Each operating band can excite a different current mode. On one band a folded or bent wire may resemble the intended mode; on another it can develop additional lobes, deep nulls or stronger current on a feed line or support. Transformer loss, tuner loss and feed-line mismatch loss can also change sharply between bands.
So the useful multiband record is not “80–10 m.” It is a band-by-band table containing:
- the complex feedpoint impedance and its measurement plane;
- the matching or transformation used and its insertion loss or thermal evidence;
- feed-line type, length, attenuation and SWR under load;
- exterior-current measurements at relevant cable positions;
- the modelled or measured current distribution and pattern needed for the operating goal; and
- power, duty cycle, weather and access conditions.
Height Changes a Pattern, but Not by Slogan
A low horizontal wire often has a stronger high-elevation component than the same wire at greater electrical height because the direct and ground-reflected fields combine differently. That can support shorter sky-wave paths under suitable ionospheric conditions, but it does not guarantee regional coverage or prove high efficiency.
The result depends on height in wavelengths, polarization, soil properties, terrain, conductor loss, nearby structures and the installed current distribution. A bent 80 m wire over a small garden, a 40 m inverted V and a 10 m dipole at the same physical height are three different electromagnetic problems.
Likewise, “moderate height works DX” is not a performance guarantee. Useful low-angle radiation, path availability and received signal-to-noise ratio all matter. ITU-R P.533 treats HF circuit performance as a function of path, time, frequency, ionospheric state, antennas, power and noise—not antenna height alone.
Resonance, Match and Radiation Are Different Questions
At a declared feedpoint and frequency, input resonance means the input reactance is zero. A match describes the relationship between the presented impedance and the system reference impedance. Neither quantity by itself establishes radiation efficiency, pattern or gain.
A wire near ground or nearby objects can resonate above or below a simple cutting-formula estimate. The direction of the shift is not universal: bends, end loading, dielectric contact, induced currents in nearby metal and changes in the return path can push the input impedance in different ways. Cut long, install the complete system, measure, and trim deliberately—without assuming every frequency shift has the same cause.
A tuner at the shack can present a suitable impedance to the transmitter while the feed line still carries a high SWR. The tuner does not remove the extra line loss, connector stress or high-voltage/current points that arise between it and the antenna. For a non-resonant system, that can be entirely acceptable if the feed line and tuner are chosen for it and the loss and stress remain within limits.
Keep the Feed Line in the Drawing
In normal coaxial transmission, the wanted current travels on the centre conductor and returns on the shield’s inner surface. A separate current can flow on the shield exterior relative to the antenna, mast, station and environment. That exterior current is the common-mode path of concern here.
The coax does not automatically “become the antenna,” and a choke is not automatically the cure. Exterior current requires an excitation and a complete return path. It can arise from feed-transition imbalance, unequal environmental coupling or a design that intentionally uses part of the coax exterior as a counterpoise.
A current choke inserts complex, frequency-dependent impedance into one chosen common-mode loop. Its useful location and required impedance depend on the loop, frequency and intended boundary. A choke cannot create a missing return conductor, guarantee symmetry, remove differential mismatch or replace protective bonding.
Measure before and after. A calibrated RF current probe around the whole coax, used at several positions and frequencies, can show whether exterior current changed. If moving the cable or adding a choke moves the input impedance substantially, the feed line was participating in the installed electromagnetic structure; the new reading is not automatically “better” until the wanted outcome is also measured.
Nearby Objects Are Coupled Conductors, Not Villains
Fences, gutters, downpipes, railings, window frames, sheds, house wiring and plumbing can carry induced current when coupling is significant. They may shift impedance, alter current distribution, reradiate or provide an unintended return path. The sign and size of that change depend on separation, orientation, electrical length, bonding, material and frequency.
Wet branches and foliage can change dielectric loading and loss; wind can change geometry; rain can alter soil and surface conditions. Do not turn those observations into “wet weather always lowers resonance” or “metal always ruins the pattern.” Record the configuration and measure the actual change.
The same discipline applies to noise. A higher receiver meter reading does not identify the coupling path. Compare signal-to-noise ratio in a declared bandwidth with preamplifier, attenuation, AGC and receiver gain held constant. Then switch suspect local devices safely, map exterior-coax current, change routing or use a separate receive antenna to distinguish antenna-mode pickup from cable-borne coupling.
Safety Defines the Usable Geometry
No contact is worth an unsafe wire route. Keep the antenna, supports, feed line and their full wind, sag and failure envelope away from overhead electrical conductors. Follow the utility’s clearance rules and all local electrical, structural, lightning and planning requirements; obtain competent help where the installation approaches a regulated hazard.
Put wire ends, matching units and other possible high-voltage points where people, pets and neighbours cannot touch them during transmission. Keep supports and strain relief within their mechanical ratings, include weather and ice where relevant, and do not tension wire like a guitar string. A tree moves; the wire and support system must accommodate that movement without creating an electrical or mechanical hazard.
RF exposure compliance is installation-specific. Frequency, transmitted power, duty cycle, modulation, antenna gain, pattern, distance, occupancy and averaging rules all matter. The ICNIRP 2020 RF guidelines cover human exposure from 100 kHz to 300 GHz, but the enforceable limits and evaluation method are set by the applicable jurisdiction. Calculate or measure the actual installation and control access; a generic garden checklist is not compliance evidence.
Lightning protection and protective bonding are separate from RF matching and common-mode tuning. A low SWR, a tuner earth terminal or a convenient garden rod does not establish a safe lightning or mains-fault installation.
Install, Measure, Then Trim
This is the sequence I use when the space is tight:
- Freeze the safe geometry. Record coordinates, heights, slopes, bends, support movement and clearances.
- Model the complete likely current structure. Include wire, feed transition, intentional return, relevant feed-line exterior, mast, nearby conductors and an appropriate ground model. Check numerical convergence instead of trusting the first coloured plot.
- Install the actual feed system. Route and secure the line as it will be used; do not tune with a temporary coil of cable beside the feedpoint and then move it.
- Measure complex impedance at stated planes. A reading at the shack includes line transformation and loss. Move the calibration plane or mathematically de-embed the known line when the feedpoint impedance is the question.
- Map exterior current. Measure along the feed line on every band of interest, then install or move a choke only to create a defined boundary.
- Tune in the final environment. Adjust one wire or network variable at a time and recheck every band because branches and modes interact.
- Measure the promised outcome. Match requires impedance data; efficiency, gain and pattern need suitable field methods; receive improvement needs controlled SNR comparisons.
IEEE 149-2021 is the active recommended practice for antenna measurements. It treats the range, instrumentation and evaluation method as part of the result. That principle still helps in a small garden: write down the conditions, use the right instrument for the claimed quantity and include uncertainty appropriate to the conclusion.
Compare Complete Systems Fairly
If you compare two garden antennas on air, keep transmitter power and reference plane explicit, hold receiver bandwidth and gain state fixed, switch quickly, repeat across time and separate signal level from noise level. A remote report includes propagation and the other station. It is useful operational evidence, not a calibrated gain or 3D pattern measurement.
Compare the whole installed systems, including tuner and feed-line losses. A topology that needs more transformation is not automatically worse; a resonant topology is not automatically more efficient. The relevant question is how much power is accepted, how much is lost before radiation, where the field goes, and whether the resulting signal-to-noise ratio serves the operating goal.
Predictable Means Bounded, Not Perfect
A small-garden wire antenna becomes predictable when changes stop being mysterious. You know the intended current path, what the feed line is doing, where the matching loss lives, which objects are coupled, how safety is maintained and which measurement answers each question.
The result may be an inverted V, fan dipole, OCFD, EFHW, doublet or carefully designed end-fed non-resonant wire. The name is the least interesting part. The complete geometry, current distribution and evidence are the design.
That is how HF works in a small garden: not by pretending the garden is free space, but by making the real installation part of the engineering from the first sketch.
Primary Technical References
- IEEE 145-2025: Standard for Definitions of Terms for Antennas
- IEEE 149-2021: Recommended Practice for Antenna Measurements
- ITU-R P.527-6: Electrical Characteristics of the Surface of the Earth
- ITU-R P.533-14: Method for the Prediction of HF Circuit Performance
- Bockelman and Eisenstadt: Combined Differential and Common-Mode Scattering Parameters
- Constantin and Tamas: Common-Mode Currents on Antenna Feeders in Radiation Measurements
- ICNIRP 2020: Guidelines for Limiting RF Electromagnetic-Field Exposure
Mini-FAQ
- Which wire antenna is best for a small garden? — There is no topology-only answer. Declare the safe geometry, bands, coverage objective, feed-line route, tuner, loss budget, return path and measurement plan before comparing complete systems.
- Does a low wire automatically make a good regional antenna? — No. A low horizontal wire may emphasize higher elevation angles, but efficiency, soil, pattern, ionospheric conditions, frequency and local noise still determine useful performance.
- Does a tuner make a non-resonant wire efficient? — Not by itself. A tuner can transform impedance at its reference plane; feed-line, tuner, transformer, conductor and ground-related losses still have to be measured or bounded.
- Does every coax-fed wire antenna need a feedpoint choke? — Not as a universal rule. Measure exterior-coax current and define the intended return path; choose choke impedance and placement for the actual common-mode loop and operating frequencies.
- Why does resonance move after installation? — Bends, height, ground, nearby conductors, dielectric loading, feed-line participation and return-path changes can all alter the installed complex impedance, sometimes in different directions.
- What should I measure before trimming the wire? — Record final geometry and cable routing, then measure complex impedance at a stated reference plane, exterior feed-line current and the outcome you actually want to improve.