Does an Inverted-L EFHW Have a “Direction”?

An inverted-L looks as though it ought to point along its horizontal wire. Sometimes the installed pattern does favour one azimuth. Sometimes it is broad, skewed or full of lobes that move from band to band. The direct answer is: the antenna can have direction, but end-fedness alone does not define it.

Before asking where it points, say which pattern cut you mean. Is the question about azimuth at 10° elevation, a high-angle azimuth cut, the elevation of maximum radiation, polarization, a null or maximum realized gain? An antenna has a three-dimensional vector pattern. One compass arrow cannot describe it.

End-Fed Is a Feed Description, Not a Pattern

In the familiar thin, straight half-wave approximation, conductor current is roughly sinusoidal: small at the open ends and largest near the middle. Moving the feed from the centre toward an end changes the feed-point impedance and matching problem. If geometry, environment and conductor-current distribution were otherwise identical, the feed-point label would not create a new far-field direction by itself.

A real EFHW installation is not otherwise identical. The matching transformer, its capacitance, the deliberate return conductor, the feed-line exterior and nearby structures can change current magnitude and phase. Bending the wire changes mutual coupling and its relationship to ground. The correct comparison is therefore not “end-fed versus centre-fed” in the abstract, but two fully declared current distributions in two fully declared installations.

For a passive linear antenna, reciprocity links its transmitting and receiving directional properties at the same frequency and polarization. The same installed current-path details that alter its transmitted field also alter what it receives.

The Bend Does Not Create Two Independent Antennas

The vertical and horizontal sections are parts of one continuous conductor. Current does not reach the corner and restart with a new amplitude or phase. The bend, total electrical length, conductor diameter, loading, transformer and environment establish one distributed solution.

Every small current element contributes a far-field vector whose amplitude, phase and polarization depend on the current at that location and the viewing direction. The observed field is the vector sum over the entire radiator and every other conductor carrying material RF current. That is why “vertical part equals DX” plus “horizontal part equals NVIS” is too crude: the two contributions can reinforce, cancel or produce different polarization components in different directions.

The vertical section still has useful intuition. A short vertical current element has a null along its own axis, so it does not radiate most strongly straight upward. A straight horizontal half-wave wire in free space has a broadside maximum and end-on nulls. But an inverted-L is neither isolated element. Its current is not uniform, the corner couples the sections and ground modifies both contributions.

Ground Is Part of the Pattern Calculation

The image-antenna picture is a useful explanation over an ideal conducting plane. Real ground is lossy. Its reflection magnitude and phase depend on frequency, incidence angle, polarization, conductivity and permittivity. The direct field and ground-reflected field combine differently at each elevation angle.

Height must therefore be stated in wavelengths at the frequency being analysed, together with the ground model. Raising the horizontal wire changes phase relative to its reflected field; changing vertical-section length changes where the large-current regions sit in height. Neither action guarantees a lower or higher takeoff angle without the rest of the model.

Recommendation ITU-R BS.705-2 calculates HF patterns with height, ground and element geometry as explicit inputs and separately warns that terrain, conductivity and nearby structures influence practical patterns. That is the right discipline even when the garden antenna is much simpler than a broadcast array.

The Straight-Wire Broadside Rule Is Only a Starting Clue

For an isolated straight horizontal half-wave wire, the free-space pattern suggests broadside preference and end-on nulls. If that wire runs east–west, north and south are the first azimuths worth examining in that idealized model. They are not guaranteed main-beam bearings for an inverted-L.

The vertical section can fill or skew the horizontal section’s nulls. Unequal height, sloping terrain, a building, mast, fence, tree moisture, transformer enclosure, return conductor and coax route can break symmetry. A real installation may retain a broadside tendency, rotate it, weaken it or create several comparable maxima.

Describe a prediction with its boundary: frequency, wire coordinates, feed and return geometry, ground parameters, nearby conductors, polarization and elevation cut. Without those inputs, “point the wire east–west for north–south” is a sketch, not an installed-pattern result.

The Fundamental Can Be Broad—or Not

On a frequency where the installed conductor supports a half-wave-like current mode, there is usually one dominant current maximum along the wire. That often produces a relatively simple pattern compared with higher modes. It does not guarantee an omnidirectional or merely subtle azimuth pattern.

The vertical-to-horizontal length ratio matters because it locates high-current conductor in different orientations and at different heights. An L with a short vertical drop and long high horizontal run is not the same radiator as one with most of its high-current region vertical. The bend position must be included in the current plot, not treated as a cosmetic routing choice.

Matching also does not prove the pattern. The same SWR at the shack can accompany different current on the intended wire, coax exterior and losses. Use impedance as a feed-system check; do not use it as an azimuth measurement.

Higher Modes Add Current Reversals and Lobes

On higher resonant bands, a multiband EFHW can support modes with additional current maxima and phase reversals. Fields from those regions add in some directions and cancel in others, commonly producing more azimuth and elevation lobes.

Do not assign the pattern from the nominal band labels alone. End effect, bend geometry, transformer reactance and compensation, conductor loading, ground and the return path shift resonances and currents. A wire sold or cut for 40 m is not automatically an exact 1λ radiator on 20 m, 1.5λ on 15 m and 2λ on 10 m in the installed electromagnetic system.

This is where horizontal orientation may become operationally important—but only after the full-band pattern is calculated or measured. A lobe on one band can coincide with a null on another, and the relevant azimuth depends on the elevation angle and polarization used for that path.

High-Angle Coverage Is Not a Fixed Height Table

Near-vertical-incidence skywave operation requires useful radiation at high elevation angles and an ionosphere capable of returning the operating frequency over the desired short path. Antenna height influences high-angle radiation through ground interference, but it is not the only variable and does not guarantee coverage.

The horizontal-section length and height, vertical-section current, ground, loss and feed-line exterior all contribute. A fixed 0.1–0.2λ prescription cannot state the maximum, null depth, azimuth uniformity or efficiency of an unidentified inverted-L. Calculate the 3D pattern for the actual geometry, then evaluate the high-angle region required by the path.

A dogleg or multiple wire directions may change azimuthal nulls, but they also change current distribution, impedance and coupling. Treat the revised layout as a new antenna model rather than assuming that bends simply “fill in” coverage.

The Coax Exterior Can Become Another Pattern Element

At an ideal coaxial port, wanted transmission-line current flows on the centre conductor and inner surface of the shield. The shield exterior can carry an additional current relative to the surroundings. An end-fed matching network must exchange current with a second terminal through some combination of deliberate conductor, transformer capacitance, enclosure, feed-line exterior, station wiring and earth coupling.

If the coax exterior carries material current, include its exact route in the antenna. Its field is not random: for a fixed installation it combines deterministically with the wire’s field. It feels unpredictable when cable length, route, bonds or attached equipment change and the common-mode solution changes with them.

A choke adds complex impedance to one selected common-mode path. There is no universal rule that every inverted-L needs one choke at the transformer and another farther down. Measure exterior current at several cable positions, define the boundary you want and test a choke whose impedance is suitable over the bands and levels involved. Then repeat the current and pattern/SNR measurements. A choke that changes the return system may also change feed impedance and the intended-wire current.

Model the Installation, Not a Symbol

A thin-wire method-of-moments model such as NEC is a useful starting point. Burke and Poggio’s NEC-2 formulation solves conductor current and derives near and radiated fields for arbitrary wire structures in free space or over ground. Its value here is not a pretty 3D plot; it is the ability to inspect why that plot exists.

Include, as far as the method permits:

• the exact vertical, horizontal and sloping wire coordinates and conductor diameter;
• the feed location, matching-network approximation and declared return conductor;
• realistic ground conductivity and permittivity, with terrain limits noted;
• the coax exterior as a wire on its installed route when common-mode current is plausible;
• masts, gutters, fences or other nearby conductors capable of carrying induced current; and
• loss, accepted power and available source power, with one normalization stated for every comparison.

For each operating frequency, inspect complex segment current before examining the pattern. Then save the full 3D gain pattern and, when matching is included, the realized-gain pattern. Keep gain comparisons at equal accepted power and realized-gain comparisons at equal available source power. Save relevant azimuth cuts at declared elevation angles, elevation cuts at declared azimuths and polarization components. Change one variable at a time: bend location, height, ground, coax route or choke impedance.

Numerical models have limits. Transformer parasitics, ground-contact wires, complex terrain and small construction details may need a better model or measured equivalent. The model remains a hypothesis until the installed current and field behaviour support it.

Measure Direction Without Measuring Propagation

A calibrated antenna-pattern measurement controls geometry, distance, polarization, input power, reflections and reference antennas; IEEE 149-2021 describes that discipline. A garden HF installation rarely offers a full far-field range, but the same logic improves field tests.

First verify the feed system: calibrated impedance at a declared plane and a repeatable scan of coax-exterior current. For pattern evidence, use a stable source or receiving site, fixed power and receiver settings, declared polarization and geometry, and rapid repeated measurements across the azimuths that matter. Keep surroundings unchanged and report uncertainty.

WSPR, RBN and contact maps are valuable operational evidence, not direct pattern plots. Transmitter population, receiver sensitivity, propagation, time, path polarization and reporting thresholds are uneven. Compare configurations with the same band, power, schedule and rapid A/B/A changes over many samples. A persistent difference can support an aiming decision; a single opening cannot locate a lobe.

Choose Orientation From the Objective

On the fundamental band, begin with the geometry that safely fits the site and places the important current regions where the model predicts useful elevation and azimuth coverage. On higher modes, check every used band because lobes and nulls move. For high-angle work, inspect the full high-elevation region rather than one nominal height ratio.

If a specific azimuth matters, rotate the complete model relative to the site and compare realized gain at the required elevation and polarization. Then verify the installed result. If the pattern remains broad, choosing a quieter cable route and less-coupled position may deliver more than turning the horizontal wire a few compass degrees.

So, does an inverted-L EFHW have a direction? It has a three-dimensional installed pattern. That pattern may contain useful azimuthal preference, but the only honest compass answer comes after the currents, ground, frequency, feed line and measurement cut are declared.

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