Choosing an HF Antenna: Objectives, Site and Constraints

A useful HF-antenna decision begins with the link and the site, not an antenna name. State the frequencies, paths, directions, available height and area, transmit power and duty cycle, receive-noise environment, supports, feedline route, safety obligations and maintenance budget. Only then compare candidate systems.

1. Define the Operating Objective

“Works HF” is not a specification. HF formally spans 3–30 MHz; amateur practice often discusses the 1.8 MHz band alongside HF systems even though it lies in MF. A structure that is electrically large at 28 MHz may be electrically small at 1.8 MHz, so a single label can hide a large change in behaviour.

Turn priorities into thresholds. Examples include “cover these regional paths on 5.3 and 7 MHz,” “place the main lobe within this azimuth sector,” “accept legal-limit SSB with stated duty cycle,” or “improve received SNR from this direction by at least 6 dB.” The numbers must be justified for the station; they are not universal antenna requirements.

2. Survey the Site as Part of the Antenna

Record dimensions in both metres and wavelengths on every required band. Height in metres alone does not predict a pattern. Include sloping terrain, soil, buildings, trees, fences, masts, gutters, photovoltaic wiring, utility conductors and every cable that can carry RF current.

For ground-coupled antennas, “average ground” is only a model assumption. ITU-R P.527-6 models the surface through complex permittivity and conductivity and includes dependence on soil composition, moisture, temperature and frequency. A radial system that performs one way over wet conductive soil may have a different loss and pattern over dry or layered ground.

3. Keep Match, Efficiency, Directivity and Gain Separate

At a named antenna port, a useful equivalent is:

The resistance split is meaningful only when radiation and loss resistance are referred to the same port and current normalization. Under that condition, η_rad can also be written as R_rad/(R_rad + R_loss). It is not generally valid to subtract one convenient measured resistance from another without a complete model.

Directivity D describes pattern concentration. Radiation efficiency η_rad describes accepted power lost before radiation. Gain combines the two. Realized gain also includes mismatch at its stated reference impedance. Feedline and external matching losses require a wider system boundary.

IEEE 145-2025 establishes current antenna terminology. Use its distinctions consistently and state whether a quoted gain is directivity, antenna gain or realized gain.

4. Match the Installed Pattern to the Path

No polarization or antenna family owns “DX,” “regional” or “low angle.” The pattern depends on electrical height, geometry, ground, nearby structures, current distribution and frequency. The useful elevation angles then depend on the path, ionosphere, time and operating frequency.

ITU-R P.533 predicts HF-circuit performance using path, time, frequency, solar conditions, noise and transmitting and receiving antenna data. ITU-R BS.705-2 treats HF antenna patterns, ground effects, surrounding structures and differences between theoretical and installed performance.

Those documents imply a practical sequence:

1. predict or measure the installed three-dimensional pattern on each required band;
2. identify the range of useful path angles and azimuths rather than one “take-off angle”;
3. combine gain over that region with the transmitter, path, receive antenna and noise terms; and
4. test sensitivity to height, ground and nearby-conductor changes.

A low horizontal wire can favour high elevation angles on one band and develop multiple lobes on a higher band. A monopole can provide useful low-elevation radiation only when its return system, loss and surroundings support that pattern. A directional array trades coverage in some directions for gain in others. These are conditional pattern results, not universal rankings.

5. Treat Bandwidth as a Defined Pass/Fail Test

Bandwidth needs a criterion. An antenna may meet a 2:1 SWR limit across a band yet fail a gain, efficiency, pattern, voltage, current or tuner-loss requirement within the same range. Conversely, a narrow impedance bandwidth can be acceptable when fast retuning is available and component stress remains controlled.

Reducing electrical size generally increases stored energy relative to radiated energy and makes bandwidth and loss more demanding. L. J. Chu’s physical-limit analysis establishes idealized Q and bandwidth bounds for antennas enclosed by a sphere. A real compact radiator can sit well above that minimum Q because conductor, loading, dielectric, ground and matching losses add practical limitations.

A matching network can transform impedance at its design frequencies. It cannot erase conductor loss, ground loss or the stored-energy physics of an electrically small radiator. Evaluate the network with the actual complex load, frequency range, component Q, voltage, current, power and duty cycle.

6. Compare Complete Feed Systems

Coaxial cable, open-wire line and ladder line solve different installation problems. None is automatically lower-loss as an installed system.

• Coaxial line provides a defined shielded differential mode and convenient routing, but its attenuation and mismatch loss depend on cable type, length, frequency, temperature and load.
• Balanced open-wire line can have low loss under high mismatch, but nearby conductors, spacing changes, rain, routing and the transition to the tuner can disturb balance and impedance.
• A remote matching network can reduce mismatch on the line, but moves voltage, current, weather sealing, control and maintenance requirements outdoors.

Use manufacturer attenuation data as a baseline, then calculate or measure the installed line with the actual load. Include connector, balun, choke and tuner losses. A station-end tuner can provide a suitable transmitter impedance while a mismatched feedline still dissipates power.

Also separate the intended line mode from exterior or common-mode current. Equal and opposite conductor currents define the intended differential transmission-line mode. Current on the coax exterior, or an unequal residual on a balanced pair, follows a different return path and can alter pattern, received noise, equipment voltage and tuning. Choke location and impedance must be chosen from the installed current paths, not from the antenna name alone.

7. Design Receive Performance Around SNR and Linearity

Maximum terminal voltage is not the same as maximum information. At HF, atmospheric, galactic and human-made noise vary strongly with frequency, location, time and direction. ITU-R P.372-17 provides statistical radio-noise models; those values are planning inputs, not a measurement of a particular garden or rooftop.

Evaluate receive candidates at the receiver input using:

• wanted-signal gain over the required direction and polarization region;
• noise response over the full installed pattern, including local common-mode ingress;
• feedline, transformer, filter and matching loss before the first active stage;
• receiver noise figure only after the external-noise and front-end-loss terms are known; and
• blocking, intermodulation and compression from strong in-band and out-of-band signals.

A smaller or less efficient receive antenna can outperform a larger transmit antenna when its pattern rejects more unwanted noise and its output remains high enough for the receiver noise contribution to stay acceptably small. It can also perform worse when preamplifier noise, overload, feedline ingress or loss dominates. Measure both wanted signal and noise in the same bandwidth and time interval.

8. Make Installability a Technical Requirement

An electrically attractive design is not suitable if it cannot remain safely supported, cleared, maintained and operated. Record wind and ice assumptions, wire tension, mast loads, corrosion, connector sealing, access for inspection, conductor clearances, lightning and surge strategy, protective bonding, RF exposure and local planning or electrical rules.

Do not derive an RF-exposure clearance from power alone. Field distribution depends on frequency, current distribution, geometry, ground, duty cycle and accessible location. ICNIRP’s 2020 RF guidance treats 100 kHz–30 MHz exposure as a near-field compliance problem requiring both electric- and magnetic-field considerations. Applicable national rules and assessed installed fields govern the station.

Lightning protection, RF return, protective earthing and common-mode control are different functions. An antenna counterpoise or choke is not a substitute for coordinated lightning, surge and electrical-safety design.

9. Compare Antenna Families Without Ranking Them Universally

10. Use a Reproducible Selection Workflow

1. Write the mission. List bands, paths, directions, seasons, operating modes and receive priorities.
2. Survey the site. Measure the available three-dimensional geometry and map conductors, ground, supports, people and noise.
3. Set limits. Define power, duty cycle, bandwidth, tuner range, line loss, structure, access, cost and safety requirements.
4. Model complete candidates. Include ground, feedline exterior, mast and major nearby conductors; vary uncertain parameters.
5. Build a loss and stress budget. Name every reference plane and include line, transformer, choke, tuner, conductor and ground losses.
6. Measure the installed system. Record calibration plane, geometry, weather, frequency, power, raw data, repeatability and uncertainty.
7. Compare against the mission. Use gain over the required angular region, SNR, bandwidth, stress, maintainability and safety—not one meter reading.
8. Recheck after changes. Feedline routing, wet soil, vegetation, mast bonding or a new cable can alter the current paths.

IEEE 149-2021 treats antenna measurement as a facility, instrumentation, calibration and uncertainty problem. A field comparison should likewise state the measurand, reference plane, geometry, instrument limits and uncertainty before assigning a performance difference.

Decision Examples With Explicit Boundaries

• Regional service from a modest property: compare horizontal-wire heights and azimuths against the required path-angle range; do not assume that every low wire produces useful NVIS on every band and time.
• Low-band long-distance work: compare installed gain at predicted path angles, including real ground and return loss; do not rank “vertical” against “dipole” without height and radial details.
• One structure for several bands: compare total line-plus-tuner loss, impedance extremes, pattern changes and switching time; bandwidth convenience can trade against pattern control.
• Noise-limited receiving: compare SNR and blocker performance at the receiver input; the antenna with less absolute signal can still win when unwanted noise falls farther.
• Severely restricted space: quantify the compact antenna’s efficiency, bandwidth, tuning, voltage, current and accessibility rather than using a generic efficiency table.

Engineering References

• IEEE 145-2025: Standard for Definitions of Terms for Antennas
• IEEE 149-2021: Recommended Practice for Antenna Measurements
• ITU-R BS.705-2: HF Transmitting and Receiving Antenna Characteristics and Diagrams
• ITU-R P.533: Method for Predicting HF-Circuit Performance
• ITU-R P.372-17: Radio Noise
• ITU-R P.527-6: Electrical Characteristics of the Earth’s Surface
• L. J. Chu: Physical Limitations of Omnidirectional Antennas
• ICNIRP 2020 Guidelines for RF Electromagnetic-Field Exposure

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