Raised Radials or Raised Feedpoint? Two Different Upper-HF Changes
Raised Radials or Raised Feedpoint? Two Different Upper-HF Changes
On 15, 17 and 20 metres, lifting a radial system and lifting the whole vertical can both improve a station—but for different reasons. One mainly reshapes the return-current problem; the other changes the radiator’s electrical height and its interaction with ground.
When somebody says an elevated-radial vertical beat a ground-mounted one, I want to know what was actually elevated. Were only the return conductors moved above the soil, or did the feedpoint and the entire current distribution move upward too? Without that distinction, an honest result can still produce the wrong explanation.
Joeri’s practical point: radials in the air primarily change return-current distribution, soil coupling and loss. Raising the complete antenna changes electrical height, ground reflection and pattern as well. On upper HF, a few physical metres are a useful fraction of a wavelength, so mixing those two changes can make “height wins” look simpler than it is.
Keep the Two Levers Separate
A vertical monopole needs a second source terminal. On-ground radials, elevated radials, a conductive structure and unintended coax-exterior current can all participate in that return. Changing the radial system can change feedpoint impedance, current sharing, loss and pattern even if the radiator base stays at the same height.
Raising the whole antenna is a different experiment. The radiator, feedpoint and deliberate return move relative to lossy ground and nearby structures. The direct and ground-reflected fields combine differently, and the coax route, mast and support may become new current paths. That can change the elevation pattern as well as efficiency.
| Change | Primary engineering question | What it does not prove by itself |
|---|---|---|
| On-ground radials replaced by elevated radials at comparable base height | Where does return current flow, and how much power is lost in soil and connections? | A lower take-off angle or a higher radiator |
| Complete radiator and radial system raised | How do electrical height, reflection, nearby conductors and the full current distribution change the pattern? | That the radial system alone caused the difference |
| Radial count or length changed | How do coverage, tuning, symmetry and current sharing change? | A universal optimum independent of soil, height and surroundings |
What Elevated Radials Can Do
Surface radials are strongly coupled to earth and normally behave as a distributed conductor-and-soil network. A sparse set can leave substantial return current in lossy soil. Adding conductors changes current density and coverage, but the improvement diminishes as the important near-base region becomes well served.
Elevated radials are more visibly part of the tuned antenna structure. A small, symmetric set can keep more return current on copper and can approach the measured transmission performance of a much larger surface system in a controlled installation. The price is discipline: radial length, height, slope, insulation, surrounding metal and current equality matter.
That is why “four elevated equals sixty-four on the ground” is not a design law. It summarizes one class of results while hiding frequency, soil, geometry, choke placement, radial voltage and measurement uncertainty. More elevated conductors can make a system less dependent on one wire, but no fixed count guarantees balance or low loss.
What Raising the Complete Vertical Can Do
Move the feedpoint and radiator upward and the experiment no longer concerns only radial loss. The height of current-rich conductor changes relative to ground, and so does the phase relationship between the direct field and the field reflected by imperfect earth. The result may move energy between elevation angles; it is not automatically a gain increase in every direction.
The surrounding installation matters just as much. Roof metal, gutters, railings, a tower, guy wires, solar wiring and the feedline can couple to the antenna. A raised system with uncontrolled coax current may appear to gain in one direction because an unintended conductor joined the array. That is pattern change, not free power.
ITU-R BS.705-2 models vertical monopoles with explicit earth systems and separately warns that conductivity, topography and nearby structures change practical patterns. A useful comparison therefore needs the installed elevation pattern or calibrated field data—not an ideal-ground sketch.
Why 15, 17 and 20 Metres Make the Difference Visible
A two-metre physical change is about 0.10 wavelength on 20 metres and about 0.13 wavelength on 15 metres. That is large enough to alter coupling and electrical height, yet it does not create a universal low-angle advantage. The direction and size of the change depend on radiator current, radial geometry, ground and surroundings.
The same physical support height is also a different fraction of a wavelength on each band. A multiband fan adds mutual coupling between radial sets and radiators, so the 15-metre wire does not live in isolation from the 17- and 20-metre conductors. Trim the complete installed system, then verify current and pattern band by band.
Quarter-wave starting dimensions remain only starting dimensions. End effect, conductor diameter, insulation, slope and coupling move resonance. Matching the input after those shifts does not prove that loss or the wanted elevation field improved.
What N6LF’s Measurements Actually Support
Rudy Severns, N6LF, compared surface and elevated radial systems with feedline common mode controlled and the test geometry documented. His surface-versus-elevated experiment showed that a small elevated set can perform very well relative to a much larger ground-surface system under the tested conditions.
His multiband radial experiments are especially relevant here. They tested actual multiband arrangements rather than assuming that one single-band result transfers unchanged. Count, wire length, band, height and coupling stayed attached to the measured result.
The evidence supports two useful conclusions: elevated systems can be efficient with fewer conductors when their geometry and current are controlled, and a dense surface system can also perform very well. It does not support “elevated always wins,” “many short always wins,” or one radial recipe for every site.
Control the Feedline Before Judging the Pattern
The coax exterior is another possible return conductor. If its current changes when the antenna or radials move, then the radiator has changed too. SWR movement with cable routing, strong current below the feedpoint, shack RF or unexplained azimuthal asymmetry are reasons to investigate that path.
Use a choke whose complex common-mode impedance and powered stress are suitable for the installed load, but do not place it by folklore. Its position defines a current boundary, so it must leave the intended return on the antenna side. Measure exterior current above and below that boundary across every operating band.
A Comparison That Answers the Right Question
- Define the objective. Name the band, azimuth, elevation region, bandwidth and power envelope that matter.
- Change one lever. Compare radial type at constant base height, or compare complete-system height with the return geometry otherwise controlled.
- Record the installed circuit. Include radiator, every radial, mast, feedline route, choke, matching network, bonds and nearby conductors.
- Use a declared reference plane. Save calibrated R + jX rather than only an SWR minimum.
- Map current. Measure elevated-radial current sharing and coax- or mast-current changes.
- Hold accepted power constant. Include feedline, matching, conductor, connection and ground-system losses.
- Compare field quickly. Use calibrated fixed receivers or rapid A/B/A switching in the same directions, and log propagation and uncertainty.
Joeri’s Bottom Line
If limited wire and a controlled, inaccessible geometry make elevated radials practical, they can be an excellent return system. If the site favours broad surface coverage and mechanical simplicity, an on-ground field can be the better engineering choice. Neither one wins from the label alone.
On 15, 17 and 20 metres, raising the whole installation can visibly change the pattern because a few metres matter electrically. Just keep the explanation honest: separate what happened to return loss from what happened to radiator height, common-mode current and the reflected field.
Primary and authoritative references
Mini-FAQ
- Are elevated radials always better than on-ground radials? No. A controlled elevated set can be efficient with fewer wires, while a dense surface system can also perform very well. Geometry, soil, symmetry and current distribution decide.
- Does raising radials automatically lower the take-off angle? No. Raising only the return conductors mainly changes coupling and loss. Raising the complete current distribution is a separate pattern change.
- Why is physical height especially noticeable on 15, 17 and 20 metres? The same height is a larger fraction of a wavelength than on lower HF bands, so coupling and the direct-versus-reflected field can change appreciably.
- Are four elevated radials enough? They can work well in a clean symmetric installation, but no fixed count guarantees balanced current, low loss or a stable pattern.
- Can SWR prove which radial system radiates better? No. SWR cannot separate radiation from conductor, matching, connection, soil or common-mode loss. Compare at equal accepted power with current and field measurements.
- Where should the feedline choke go? At the measured boundary that blocks the unwanted coax-exterior path while leaving the intended radial or counterpoise return on the antenna side.