Antenna Height, Ground Loss and Resonance: Separate the Effects

There is no universal height such as 0.2 or 0.25 wavelength at which a wire antenna suddenly becomes efficient, stable or suitable for DX. The outcome depends on the complete geometry, polarization, electrical height, feed and return paths, soil model, nearby conductors and the performance metric being discussed. Treat “resonant,” “matched,” “efficient” and “low-angle” as different claims.

Five Quantities That Should Not Be Collapsed into One

At a stated reference plane, antenna input impedance is:

Zero reactance does not require Rin = 50 Ω, so resonance does not guarantee a low 50 Ω SWR. Conversely, a tuner can present 50 Ω to the transmitter while the antenna itself remains reactive. The accepted power is divided among radiation and losses in conductors, dielectrics, ground and any components included inside the chosen system boundary. A matching network adds its own loss.

Directivity describes how radiated power is distributed by angle. Gain also includes radiation efficiency. Realized gain additionally includes mismatch at the stated port. A lower take-off angle, low SWR or sharp current maximum is therefore not an efficiency measurement. IEEE Std 145 defines these antenna terms separately for precisely this reason.

What Height and Real Ground Actually Change

Height must be expressed electrically: h/λ. A wire 10 m above ground is about 0.06 wavelength high on 160 m, 0.13 wavelength on 80 m and 0.25 wavelength on 40 m. The same installation can therefore have very different impedance and elevation patterns on different bands.

Ground interaction has at least two distinct outcomes:

• Reflection and interference: the direct and ground-reflected fields combine by angle. Conductivity, relative permittivity, polarization, incidence angle, terrain and height affect the result. A pattern change is not automatically dissipative loss.
• Absorption: finite-conductivity, lossy ground can dissipate near-field energy or power carried by a ground-return system. The loss mechanism and its importance depend on the antenna geometry and current path; “keep the current peak away from dirt” is not a general loss equation.

ITU-R P.527 models the Earth's electrical properties with conductivity and complex permittivity, including layered-soil and penetration-depth effects. Soil type alone is not enough: moisture, temperature, frequency, depth and seasonal change matter. Labels such as “average ground” are modelling assumptions, not site measurements.

ITU-R BS.705 likewise treats HF pattern calculations as models with explicit antenna and ground assumptions. A model can compare designs usefully, but a result obtained over one uniform half-space should not be advertised as the performance of every installation.

Horizontal Dipoles, EFHWs and Doublets

A thin, straight fundamental half-wave dipole has a current maximum near its centre and current minima near its ends. An end-fed half-wave has the same broad standing-wave form on its wire, but its feed transformer, short return conductor, support structure and coax exterior complete the real system. On harmonic operation, multiple current maxima and pattern lobes appear.

Lowering a horizontal antenna generally changes its feed impedance and moves more radiation toward higher elevation angles, but the amount depends on its shape and ground. The reduction in low-angle field can be important even when conductor loss remains small. That is a pattern result, not proof that most accepted power is heating the soil.

A doublet is not permanently “non-resonant.” Its input reactance and resistance vary with frequency, length, height and routing; it may pass through resonances on some bands. Balanced line can reduce feed-line loss at high SWR, but line length, characteristic impedance, conductor loss, balance and tuner loss still belong in the efficiency budget.

No single target such as 0.2 wavelength proves efficiency for a dipole, EFHW or doublet. If the objective is a particular low-angle field, state that elevation angle and azimuth, then compare realized gain for the exact installation. If the objective is heating loss, calculate or measure accepted and radiated power separately.

Quarter-Wave Verticals and Radial Systems

A ground-mounted quarter-wave monopole has large current near its base and needs a defined return path. Current flowing through lossy earth can add series loss, reduce radiation efficiency and broaden the feed-point match. That makes the connection, radial field, soil and feed-line common mode part of the antenna—not an accessory after the fact.

There is no universal rule that 16 radials are a minimum, 120 are always best, or any four elevated radials are sufficient. The useful count depends on radial length, wire available, burial or surface contact, soil, frequency, radiator size, desired efficiency, symmetry and isolation from the feed line. The familiar 120-radial reference belongs to specified broadcast-style ground systems; it is not a proof that every amateur installation should or can use that geometry.

Rudy Severns, N6LF, tested these questions rather than reducing them to a count:

• His 7.2 MHz experiments used a 34 ft vertical, four 35 ft radials and deliberate common-mode isolation while radial height was varied. Moving an otherwise similar system between elevated and earth-contact conditions produced materially different loss in that test.
• His surface-radial experiments found that, for a fixed amount of wire in the tested system, more shorter radials could outperform a few long ones. The excess-loss effect associated with individual radial resonances became much smaller by about 16 radials in those measurements.

Those are valuable primary experiments, not universal constants. Elevated radials require defined geometry, tuning, clearance and feed-line isolation. Ground-level radials need good base connections and a design appropriate to the soil and available wire. Raise the feed point only as part of a complete elevated system; mast height by itself does not create a low-loss return.

Loops Do Not Receive a Ground-Loss Exemption

“Closed current path” does not make a loop ground-insensitive. A full-wave loop can be horizontal, vertical, triangular, rectangular, distorted or multi-band; feed point and installation change its current distribution and polarization. A low horizontal loop commonly develops a high-angle pattern, while a vertical loop's low-angle pattern depends strongly on height, orientation, feed point and ground.

Likewise, calling current “distributed” does not prove low loss. A loop at 0.05 wavelength above real ground can still couple strongly to the ground, shift impedance and lose low-angle field. State the geometry, frequency and soil assumptions, then calculate pattern and efficiency rather than assigning every loop a “low” or “moderate” sensitivity label.

Short, Off-Centre and Transformer-Fed Antennas

Names such as EFOC, random wire and Rybakov describe families, not complete electromagnetic specifications. A 29 m off-centre-fed wire with a 4:1 transformer does not inherit a universal “semi-resonant” impedance or ground-loss class. Its wire division, routing, transformer topology and loss, counterpoise, coax length, choke position, bands and surroundings all matter.

On a short non-resonant vertical, the radiator may have low radiation resistance and substantial reactance. Loss in a transformer, loading or ground return can then consume a large fraction of accepted power. Feed-line common-mode current is not inevitable: it appears when the coax exterior participates in an inadequately defined or unbalanced return path. A suitable counterpoise or radial system plus measured choking can make the current path reproducible.

A low SWR after matching cannot reveal those losses. A lossy ground, component or cable can make an antenna easier to match while reducing radiated power.

A Reproducible Way to Compare Height

1. Define the objective. Specify band, mode, azimuth, elevation angle, coverage area, allowed footprint and whether the metric is efficiency, realized gain, field strength or match.
2. Draw the entire RF system. Include wire diameter and conductivity, insulation, transformer or tuner, feed line, choke, mast, supports, radials or counterpoise and nearby conductors.
3. State the ground model. Record conductivity and relative permittivity, frequency, assumed layer depth and terrain. Run at least plausible dry and wet cases if site data are unavailable.
4. Move one variable at a time. Sweep electrical height while retaining the same conductor, feed, return and ground. Retune only if that is explicitly part of the comparison.
5. Inspect more than SWR. Compare input impedance, accepted power, component and ground loss, current distribution, common-mode current, elevation pattern and realized gain at the relevant angles.
6. Validate the installation. A calibrated VNA can verify impedance, but not radiation efficiency by itself. Add common-mode-current checks, temperature inspection, and repeatable field-strength or pattern measurements where practical.

Practical Decision Table

Primary Sources Checked

• IEEE Std 145-2025 — Standard for Definitions of Terms for Antennas
• ITU-R P.527-6 — Electrical characteristics of the surface of the Earth
• ITU-R BS.705-2 — HF transmitting and receiving antenna patterns and diagrams
• N6LF — Experimental Determination of Ground System Performance for HF Verticals, Part 3
• N6LF — An Experimental Look at Ground Systems for HF Verticals

Questions, antenna-factor records or height trials to share? Contact RF.Guru.