Multiband Antennas Without a Radial Field
Multiband Antennas Without a Radial Field
Some antennas genuinely need no monopole-style radial field. None can operate without a complete RF current path.
“No radials” is useful only when it means no ground-mounted monopole radial field. It must never be read as “no return current.” A dipole has another arm, a loop continues around its conductor, an end-fed wire uses an intentional or incidental counterpoise, and a monopole needs a conductive reference structure. If we cannot identify that path, the installation is using something we forgot to draw.
My selection rule: do not ask whether the antenna has visible radials. Ask which conductors carry equal-and-opposite feed current, where current can convert to common mode, what part of the input resistance is loss, and whether the installed pattern serves the intended paths.
The Missing Conductor Is Usually Hiding in Plain Sight
At the feedpoint, RF current cannot simply leave one terminal and vanish. The complete electromagnetic circuit can include conduction current in metal and soil plus displacement current through capacitance to the surroundings. The practical return structure may be obvious, distributed or accidental:
- a buried, on-ground or elevated radial system;
- the second arm of a dipole-derived antenna;
- the remaining conductor of a closed loop;
- a deliberate counterpoise;
- the exterior of a coax shield, mast, support or guy;
- station bonding, protective-earth wiring and connected equipment; or
- capacitance from the antenna and its hardware to earth and nearby structures.
These paths can exist in parallel. Their currents divide according to impedance, and those impedances change with frequency. That is why changing coax length, moving a choke or touching mounting hardware can shift the feedpoint measurement of an antenna advertised as “radial-free.”
A Monopole Needs a Conductive Reference
The ideal quarter-wave monopole over a perfect conducting plane is a useful model. A real installation replaces that perfect plane with radials, mesh, a vehicle body, an elevated counterpoise, metal roofing, earth or some combination. Current returning through lossy soil and junctions contributes loss resistance; current on asymmetrical structures can also alter the pattern.
Rin = Rradiation + Rconductor + Rground + Rnetwork + other coupled-loss terms
The terms are not always separable by one feedpoint measurement, but the equation exposes the trap: a convenient input resistance can include loss. A matching network may transform the sum to 50 Ω without identifying how much accepted power reaches the far field.
Rudy Severns, N6LF, measured real HF vertical ground systems while controlling feedline common mode. His QEX radial-screen experiments show that radial length, count, soil, moisture and layout interact. A sparse on-ground system could be particularly sensitive, and the best length in one test was not a universal free-space quarter wavelength. His later elevated-radial study shows why a small symmetrical elevated set can perform well under controlled conditions yet become sensitive to asymmetry, nearby conductors and changing surroundings.
Those are experimental boundaries, not a magic radial count. An on-ground field, a few elevated conductors and a metal vehicle body are different antennas. The right design follows the actual soil, height, symmetry, available wire and operating bands, then survives measurement at the site.
Dipoles and Loops Avoid a Monopole Radial Field
A centre-fed dipole supplies two intended conductors at the feedpoint. It does not need a monopole ground screen. That does not make every installed dipole perfectly balanced: unequal arm geometry, ground coupling, a nearby mast and an asymmetrical feed transition can drive current on the feedline exterior.
An off-centre-fed dipole is intentionally fed where the two sections present different geometries relative to ground and the feedline. Its transformer can provide impedance transformation, but current balance and common-mode suppression remain separate questions.
A full-size closed loop also provides a continuous intended conductor rather than returning current through a ground radial field. On multiple bands, however, its current distribution, feed impedance and radiation pattern change with electrical length. A loop is not automatically balanced merely because the wire closes, and a low loop can couple strongly to unequal surroundings.
These topologies are honest ways to avoid a conventional radial field. They do not support a universal claim of higher efficiency, lower noise, better DX or a fixed take-off angle. Those outcomes belong to the installed current distribution and environment.
End-Fed Does Not Mean One-Terminal
An end-fed half-wave or non-resonant end-fed wire still has two transformer or matching-network terminals. On the antenna side, return current may use a deliberate counterpoise, the coax exterior up to a choke, capacitance to nearby structures, or several of these paths at once. The current ARRL EFHW kit makes the point physically by providing a separate counterpoise connection.
If the coax exterior is deliberately allowed to form part of the antenna-side counterpoise, its length and route belong in the design. The choke then defines a boundary after that section. If the exterior current continues through the station, the microphone cable, control wiring and protective earth may join the antenna whether the drawing shows them or not.
Transformer, Choke and Tuner Have Different Jobs
| Part | Primary job | What it cannot prove |
|---|---|---|
| Impedance transformer | Changes the impedance relationship between its ports | That return current is controlled, loss is low or the pattern is useful |
| Common-mode choke | Presents impedance to unwanted current on the feedline exterior | That the antenna has an adequate intended return conductor |
| Antenna tuner | Transforms the impedance at its own reference plane | That upstream feedline, transformer, ground or termination loss is small |
A choke is not a missing radial in a ferrite enclosure. It can move the common-mode boundary and force the antenna to use the intended conductors on its side, but it cannot make the current-continuity requirement disappear. Its impedance also changes with frequency, construction, cable routing and temperature.
ITU-T K.136 defines converted common-mode current as current produced by cable or network unbalance and requires unwanted current on the outside of shielded transmission lines to be controlled during RF measurements. That is a useful laboratory discipline for antennas too: if exterior current is not controlled or recorded, the feedline is an uncontrolled conductor in the test.
Multiband Operation Moves Every Boundary
A conductor that is electrically short on one band may be resonant or several wavelengths long on another. The coax exterior, mast, counterpoise and station wiring therefore do not keep the same impedance or current distribution across HF. A choke that provides a strong boundary on one band may be less effective on another. A tuner setting changes the impedance at its plane but not the physical current paths beyond it.
The radiator’s pattern changes too. Higher-order modes can create additional lobes and nulls, while unintended current on a vertical feedline or mast can tilt or fill them. “Multiband” should mean that the complete installation meets separately stated criteria on each band, not merely that an analyser finds several SWR dips.
Loss Can Make an Antenna Look Easy to Match
Loss often broadens an impedance curve and lowers its peaks. That can make a system easier for a transmitter or tuner while reducing the fraction of accepted power that is radiated. Terminated antennas make this trade deliberately; poor ground contact, lossy loading networks and unintended structure currents may make it accidentally.
A low SWR is therefore a feedpoint result, not an efficiency certificate. IEEE Std 145-2025 keeps impedance, radiation efficiency, directivity, gain and realized gain as distinct quantities. The distinction matters most when comparing unlike topologies: match, loss and pattern must be reported separately at declared reference planes.
Choose the Topology by Its Intended Current Path
| Topology | Intended return structure | Installation question |
|---|---|---|
| Ground-mounted monopole | Radials, mesh, conductive body and earth coupling | How much accepted power is lost near the base and in soil? |
| Elevated monopole | Elevated counterpoise or radial conductors | Are currents symmetrical on every band, and what nearby metal disturbs them? |
| Dipole or off-centre-fed wire | Two intended radiator sections | Does unequal coupling convert current onto the feedline or support? |
| Closed loop | The continuous loop conductor | How do higher modes, feed asymmetry, height and surroundings change current and pattern? |
| End-fed wire | Deliberate counterpoise, controlled coax-exterior section and environmental capacitance | Where is the choke boundary, and what changes when cable length or routing changes? |
Space may make one topology more practical than another. A loop or dipole-derived antenna can remove the need for a ground radial field; an end-fed wire can put the feedpoint in a convenient place; a monopole can provide a useful vertically polarized pattern. The correct choice is the one whose complete current path, loss budget and pattern can be made repeatable at the intended site.
How to Test a Supposedly Radial-Free Installation
- Draw every conductor. Include both radiator terminals, coax shield exterior, mast, guys, tuner chassis, station bonds, protective earth and nearby metal.
- Declare the reference plane. Feedpoint S11 and shack-end SWR are different measurements when the feedline has loss or common-mode current.
- Map exterior current. Use a characterized clamp current probe at several positions along coax, mast and bonds on every operating band.
- Move one boundary deliberately. Change choke position or add a known counterpoise, then restore the baseline. A repeatable change identifies a current path more convincingly than one isolated reading.
- Separate match from loss. Record transformer, tuner, termination and feedline temperature or loss under the intended power and duty cycle.
- Check the pattern that matters. Model the complete conductors and verify important directions with repeatable field measurements or simultaneous/rapid-switched receive comparisons.
My Practical Conclusion
I have no objection to an antenna that operates without a conventional radial field. Dipoles, loops and carefully bounded end-fed systems can do exactly that. My objection is to the phrase “no radials” being used to conceal the other half of the RF circuit.
If the return structure is intentional, measurable and stable across the required bands, the installation has a defensible design. If the coax, mast, station wiring or soil is silently doing the work, the antenna may still make contacts—but its tuning, loss and pattern belong to those accidental conductors too.
Mini-FAQ
- Can an antenna operate with no RF return path? No. The complete path may use intended conductors, earth coupling, displacement current or incidental structures, but feed current cannot leave one terminal without an electromagnetic return.
- Does a dipole need ground radials? Not as a monopole ground screen. Its two radiator sections provide the intended differential-current path, although imbalance can still drive current on the feedline exterior.
- Does an end-fed antenna need a counterpoise? It needs an antenna-side return path. That may be a deliberate counterpoise, a controlled section of coax exterior and distributed capacitance; omitting a named wire does not remove the path.
- Can a common-mode choke replace radials or a counterpoise? No. A choke presents impedance to exterior-feedline current and defines a boundary. The intended antenna-side return current still needs somewhere to flow.
- Does low SWR prove that no radial system is needed? No. SWR describes impedance at a reference plane. Ground, transformer, tuner, feedline and termination loss can all produce a convenient match while reducing radiated power.
- Will the same hidden return path behave on every HF band? Usually not. Its electrical length, coupling and resonance change with frequency, so current, loss, choke effectiveness and radiation pattern must be checked band by band.