Designing a 160/80 m EF-OCF: Flattop or Inverted-L?

My reason for considering this architecture is practical. A full low-band wire may fit the electrical job but not the available supports. Moving the feedpoint toward one end and bending part of the long branch downward can make the installation possible. The price of that convenience is that the off-centre split, return conductor, matching network, choke boundary and bend all become explicit design variables.

Start With the Two Required Current Modes

A useful first model begins with a total conductor path near a half-wave at the chosen 160 m frequency. In free space, the half-wave reference is c/(2f); the installed resonant length is changed by conductor diameter and insulation, height, bends, end capacitance, ground, nearby structures and the return path. A handbook shortening factor is a starting cut, not a finished dimension.

At roughly twice the frequency, the same path can support a full-wave mode near 80 m. That does not guarantee that the two resonances land in the desired parts of both amateur bands. End effects are frequency dependent, the bend couples differently to ground on each band, and the matching and return structures add their own reactance. Treat 160 m and 80 m as two installed modes that must both be designed.

The higher mode also contains more current maxima and a phase reversal along the wire. Its feed impedance and radiation pattern can therefore differ sharply from the 160 m fundamental even when both bands produce a usable SWR.

The Off-Centre Split Is a Design Variable

There is no universal 70/30 split for this antenna. Moving the feedpoint changes the current and voltage presented to the matching network on both bands. It also changes how much conductor belongs to the long branch and how much belongs to the explicit return branch.

A candidate split should be evaluated against the complex impedance on 160 m and 80 m, the current and voltage at the feedpoint, the intended return geometry and the pattern created by both branches. A point that gives a convenient resistance on one band may present large reactance, voltage or common-mode current on the other.

Off-centre geometry by itself does not make every OCF antenna common-mode. A conventional two-wire OCF dipole can still carry equal and opposite terminal currents in its intended differential mode. The one-end EF-OCF architecture discussed here is deliberately asymmetric because the two branches are implemented differently and may use the coax exterior as one radiating branch. Differential transformation and the downstream common-mode boundary must therefore remain separate design functions.

Flattop: Use the Horizontal Aperture Deliberately

A flattop keeps most of the current path horizontal. It is attractive when two suitably high supports and a clear span are available. On the 160 m fundamental, a simple horizontal half-wave begins with a broadside pattern, but its elevation response depends strongly on height in wavelengths and ground. At the modest heights available on many properties, a large high-angle component is common.

On 80 m, the full-wave mode adds lobes and nulls. Feed asymmetry, the return branch, unequal support heights, bends and nearby conductors can rotate or fill those lobes. The word “flattop” describes the drawing; it does not certify a pattern.

The flattop is usually the cleaner modelling problem because the main branch has fewer abrupt changes in height and orientation. Its practical disadvantage is the long clear span and the mechanical load placed on two supports.

Inverted-L: Trade Horizontal Span for Vertical Aperture

An inverted-L uses a vertical section and turns the remaining conductor horizontally. That can fit a long low-band wire on a site with one strong support and less horizontal room. The vertical and horizontal sections are not two separate antennas: their fields combine according to current magnitude, phase, orientation, height and ground reflection.

The bend does not guarantee a low take-off angle. If a current-rich region falls in the vertical section, that section can contribute useful vertically polarised radiation. If the current maximum falls well into a low horizontal section, high-angle radiation and environmental coupling may dominate. On 80 m, additional current maxima and reversals make the answer even more installation specific.

Choose the vertical height and bend position by plotting current magnitude and phase on both bands. Then inspect the complete azimuth and elevation patterns. A feedpoint match cannot reveal whether the available aperture is being used in the desired direction.

The Return Branch Is Part of the Radiator

An end-fed off-centre wire is a two-terminal RF system. The shorter branch may be a separate conductor, or the exterior of a declared coax section may be used intentionally between the transformer and a choke. In the latter arrangement, differential current continues inside the coax while the outside of the shield carries the intended antenna current.

The choke is the intended end of that exterior-current branch. It creates an impedance boundary; it is not a perfect wall. Its position and common-mode impedance affect resonance, feed impedance, current distribution and pattern. Map exterior current on both bands and confirm that the station-side coax, mast and wiring no longer become uncontrolled extensions of the antenna.

The Transformer and Choke Have Different Jobs

The transformer handles the differential impedance presented by the two antenna branches. When the measured complex load supports it, a nominal 4:1 UNUN can be a sensible starting network. A 4:1 label does not promise 50 Ω on either band, cancel reactance or establish balance.

The separate 1:1 choke addresses current on the outside of the coax and other connected conductors. Specify it by complex common-mode impedance, voltage, current, temperature and bandwidth at the chosen position. A low differential SWR is not evidence that the choke has created the intended boundary.

Design both components for the actual complex loads. Transformer magnetising current, leakage, winding capacitance, conductor and core loss, and the choke's own resonance all change with frequency. No power rating can be transferred from a core name or turns count to the completed, enclosed assembly without electrical and thermal qualification.

Commission the Geometry Before Trimming the Final Length

• Fix the site model: record support heights, bend position, horizontal direction, conductor, insulation, nearby metal, buildings, trees, ground and every intentional return conductor.
• Choose candidate branch splits: calculate or model complex feed impedance, current and voltage on both bands rather than copying one percentage.
• Include the real feed structure: model the transformer terminals, enclosure, intentional coax exterior, choke position, downstream coax route and station bonds.
• Install long: leave practical trimming margin, but trim against both target modes and the final geometry rather than a universal cut table.
• Measure at declared reference planes: record complex impedance at the antenna terminals and after the matching network, then separate mismatch from network and feedline loss.
• Map exterior current: check the intended branch and the station-side feedline at repeatable positions on 160 m and 80 m.
• Check the operating result: compare accepted power, thermal behaviour, current distribution and the pattern or field paths that matter for the station.

Choose the Shape From the Site and the Required Pattern

Neither shape is inherently more efficient. Conductor and ground loss, matching-network loss, return-current control, feedline loss and current distribution decide the accepted-power budget. Geometry then decides where the radiated power goes. Compare the complete installed systems rather than ranking the names.

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