F4VUX's 160/80 m Inverted-L in the Ardennes
F4VUX's 160/80 m Inverted-L in the Ardennes
Marc, F4VUX, installed a long end-fed Inverted-L beside his house in the French Ardennes. The wire produced distinct 160- and 80-metre impedance responses in a difficult physical setting. That makes the installation worth studying—but an SWR trace alone cannot tell us how much power was radiated, where it went or how the return current divided between the intended conductor, the soil, the feedline exterior and the building.
What I like about this installation is not a heroic SWR number. It is the way a familiar low-band geometry meets a very real site: limited support height, a low feedpoint, a nearby wall, uncertain earth parameters and enough wire to operate in two very different electrical modes. Those facts give us an engineering case, not a universal recipe.
The central distinction: Marc's measurements describe the impedance presented at one measurement plane under one set of conditions. They do not, by themselves, establish antenna efficiency, low-angle gain, NVIS coverage, ground quality or common-mode suppression.
The Installation Marc Built
The retained field record describes an approximately 8 m vertical rise followed by roughly 69.5 m of horizontal wire. The feed region was about 25 cm above ground, close to a house wall. A short conductor of about 1 m connected the feed region to a ground stake, while a common-mode choke was placed approximately 5 to 10 m along the coax.
The ground was described as acidic sandy or loamy soil over weathered schist and granite. That is useful site context, but geological description is not a measurement of RF conductivity or relative permittivity. Both quantities vary with frequency, moisture, temperature, composition and depth. The surrounding building, buried services, vegetation and terrain can also change the installed system.
At 3.6 MHz, an 8 m vertical rise is only about 0.096 free-space wavelength; at 1.8 MHz it is about 0.048 wavelength. Those ratios help us picture the geometry, but they do not isolate an 8 m radiator. Current flows on the complete wire and its return structure, and the bend couples the vertical and horizontal sections into one antenna.
The SWR Record Is Useful, but It Is Not a Field-Strength Record
The installation record reported these approximate SWR values:
| Band | Frequency | Reported SWR | What it establishes |
|---|---|---|---|
| 80 m | 3515 kHz | 1.1:1 | A close match at the measurement plane at that frequency and under those conditions |
| 80 m | 3800 kHz | 3.6:1 | The impedance changed substantially across the band |
| 160 m | 1800 kHz | 2.8:1 | A moderate mismatch at the measurement plane |
| 160 m | 2000 kHz | 6.6:1 | A larger mismatch toward the upper end of the measured range |
The field record does not define whether those values were measured at the feedpoint, after the transformer, at the far end of the coax or through a tuner. It also does not state the calibration plane, cable type and length, loss, weather or soil moisture. That missing context does not make the readings useless. It limits the conclusions we can attach to them.
A low SWR means that the impedance at the stated reference plane is near the system impedance. It is not an efficiency measurement. Transformer loss, feedline loss, soil loss and current flowing on lossy structures can all make an SWR curve look smoother. Conversely, a useful radiator can present a mismatch that a suitable, low-loss matching system handles safely.
Ground Quality Cannot Be Read from the Curve
It is tempting to call rocky or sandy ground “poor” and then explain every measurement from that label. The RF system needs actual electrical parameters. Conductivity and permittivity influence fields in and close to the earth, while terrain affects the far-field pattern. Neither can be inferred reliably from soil colour, acidity or the apparent sharpness of an SWR minimum.
A dissipative return path can broaden an impedance response, but so can transformer behaviour, feedline attenuation, environmental coupling and multiple interacting resonances. A broad or convenient match is therefore not proof that the soil is helping, and a narrow match is not proof of high efficiency.
If ground loss matters to the conclusion, measure or bound it. Record moisture and weather, identify the soil layers that carry current, compare a declared return system under repeatable conditions and use a model whose ground assumptions are stated. The model is a hypothesis until field measurements support it.
The Ground Stake Is One Connection, Not the Whole Return Path
An end-fed wire still needs a complete current loop. In Marc's installation, current could divide among the short stake connection, displacement current into the surroundings, the exterior of the coax, building wiring and other nearby conductors. A 1 m connection to a stake does not make the rest of those paths disappear.
The resistance and inductance of the stake connection, the soil immediately around it and the coupling to the house all vary with frequency. Protective earthing and lightning bonding also have safety purposes that must not be altered just to improve an antenna measurement.
The practical test is a current map. Measure common-mode current on the coax exterior near the feed region, immediately before and after the choke, farther along the route and at the station entry. If possible, also measure current in the intentional return conductor. Repeat on both bands and at several frequencies, because a useful choke boundary on 80 metres may behave differently on 160 metres.
The Choke Defines a Boundary Only When the Current Data Agree
A choke located 5 to 10 m from the feedpoint can leave that length of coax exterior electrically available on the antenna side. That may be intentional, but the length is not a universal counterpoise formula. The choke's complex common-mode impedance and voltage, current and thermal limits must suit the installed load over both bands.
Impedance transformation and common-mode suppression are separate jobs. The matching network transforms the differential load presented to the feedline. The choke impedes current on the feedline exterior at its installation point. A good transmitter match does not prove that the choke is effective, and an effective choke does not make an unsuitable transformer lossless.
Nor can the choke be said to decouple the house wiring merely because the SWR is stable. Coupling can occur through the antenna's near field, feedpoint capacitance, bonding conductors and coax exterior on either side of the choke. Current measurements and a controlled before-and-after field or interference test are needed.
The Pattern Must Be Earned with Geometry and Measurement
The horizontal section is low in wavelengths, especially on 160 metres, so substantial high-elevation radiation may be plausible. That does not prove NVIS communication. An NVIS path also depends on ionospheric critical frequency, absorption, operating frequency, path length, power, noise and receiver performance.
The vertical rise may contribute useful lower-angle field, but an 8 m vertical section does not guarantee a DX take-off angle. The current magnitude and phase along the complete conductor, the return structure, height, ground parameters, terrain and nearby building determine the installed pattern. On the higher current modes of the wire, several lobes and nulls can form, and their directions need not align with the desired path.
That is why I separate an impedance success from a radiation claim. To compare patterns, use a full-geometry model with measured or bounded ground, then verify selected bearings and elevation objectives with controlled field measurements. Keep accepted power constant, use fixed receiver settings or calibrated signal records and alternate configurations quickly enough that propagation changes do not dominate the result.
What This Field Case Can and Cannot Establish
| Observation | Reasonable conclusion | Conclusion that still needs evidence |
|---|---|---|
| The long Inverted-L produced measurable responses on 160 and 80 m | The installed system has useful impedance structure on both bands | Efficiency, safe power or superiority over another antenna |
| SWR was lowest near 3515 kHz in the recorded 80 m points | The measurement-plane impedance was closest to the line impedance there | Best field strength, best DX frequency or best NVIS frequency |
| The feedpoint sat low beside a wall | The building and local return environment deserve attention | That the wall has no effect or that the system is balanced |
| A choke was installed down the coax | An intentional common-mode boundary was attempted | Suppression across both bands or isolation of house wiring |
| The site was described as sandy or loamy over weathered rock | The soil context should be recorded and measured | A numerical RF ground quality or a loss value |
A Measurement Programme That Would Close the Gaps
- Freeze the geometry: record wire lengths, height profile, bend, feedpoint, coax route, choke position, stake and nearby conductors.
- Declare reference planes: calibrate the VNA at the plane being reported, or characterise and de-embed the intervening cable within a stated uncertainty.
- Record complex impedance: save resistance and reactance across both bands, not only SWR points, with tuner state and weather documented.
- Measure return current: map coax-exterior and intentional-return current before and after the choke at representative frequencies.
- Account for loss: measure or bound feedline, transformer, choke and ground-system dissipation, including equilibrium temperature at the intended power and duty cycle.
- Test the field: compare equal accepted power by bearing with rapid A/B/B/A switching, simultaneous receivers where possible and a restored baseline.
- Qualify coverage claims: relate regional or distant results to the ionospheric state, receiver noise and path geometry rather than to SWR alone.
Low Feedpoints Need Deliberate Safety Engineering
A feed region only about 25 cm above ground is accessible. End-fed matching networks and wire ends can carry high RF voltage, while return conductors and coax shields can carry appreciable current. Weatherproofing does not make an energised enclosure safe to touch.
Keep people and animals outside the controlled area, provide mechanical strain relief, protect terminals against accidental contact and maintain clearances from the wall, vegetation, utilities and combustible material. Commission at low power first, then increase power in controlled steps while watching temperature, arcing, connector heating and unexpected exterior current. RF exposure assessment must use the actual power, duty cycle, accessible geometry and installed pattern.
Bottom line: Marc's Ardennes installation is a valuable field case because it exposes the real system: long wire, low support, uncertain ground, nearby building, return conductor, coax and choke. Its recorded SWR points show where the system matched at one plane. Current, loss, pattern and field measurements are what turn that match into a defensible performance result.
Primary technical references
- Recommendation ITU-R P.527-6 — Electrical characteristics of the surface of the Earth
- Recommendation ITU-R BS.705-2 — HF transmitting and receiving antenna characteristics and diagrams
- Recommendation ITU-R P.533 — Method for the prediction of HF circuit performance
- Rudy Severns, N6LF — Experimental determination of ground-system performance, Part 1
- Roy Lewallen, W7EL — Baluns: What They Do and How They Do It
- ICNIRP — Guidelines for limiting exposure to electromagnetic fields, 100 kHz to 300 GHz
Mini-FAQ
- Does the 1.1:1 reading prove high efficiency on 80 metres? No. It establishes a close match at the measurement plane; feedline, transformer, ground and environmental losses still have to be measured or bounded.
- Does rocky or sandy soil automatically mean poor RF ground? No. RF behaviour depends on conductivity, permittivity, frequency, moisture, temperature and layered structure, not on a geological label alone.
- Is the 1 m wire to the stake the complete counterpoise? Not necessarily. Current can also divide onto the coax exterior, surrounding capacitance, bonding conductors and nearby structures.
- Does a choke 5 to 10 m from the feedpoint isolate the house? Not by placement alone. Its complex common-mode impedance and stress limits must be suitable, and currents should be measured on both sides across both bands.
- Does the low horizontal wire guarantee NVIS? No. The installed pattern must support high-elevation radiation, while the ionosphere, frequency, absorption, noise and link budget must support the path.
- What measurement would add the most confidence? A declared-reference-plane impedance sweep combined with a band-by-band return-current map, loss and thermal tests, and controlled equal-accepted-power field comparisons.