A Portable Vertical Write-Up Starts as Engineering

Greg Mihran, KJ6ER, deserves credit for publishing a complete portable concept instead of only a glamour photograph. The 17-page Dominator Halfwave Antenna, February 2025 revision gives parts, dimensions, tuning steps, a 4NEC2 comparison and field SWR captures. That is useful work. My disagreement begins where a narrow observation is asked to carry a broader conclusion than it measured.

KJ6ER’s current antenna site describes the plans as living documents. This response is therefore deliberately tied to the named February 2025 revision. It is a technical reading of those published propositions—not a judgement of Greg, his operating success or anyone’s enjoyment of the antenna.

The Claims I Am Challenging

“No Radials” Is Mechanically Narrow, Not Electrically Empty

A counterpoise and a radial are not perfect synonyms in every antenna text. A radial normally describes a conductor extending from a feed or ground region; a counterpoise often means an artificial ground or return structure, commonly elevated or capacitively coupled to earth. The Dominator’s single linked wire is therefore not a conventional radial field.

But the narrower vocabulary must not hide the circuit. The KJ6ER design connects one transformer terminal to the whip and the other to a linked wire whose published length is about 33% of a wavelength. The paper also says that changing this wire shifts system resonance enough to permit whip lengths around 43% to 45% of a wavelength. That is material electrical participation by definition.

“Low current” does not mean “no effect.” The end region of an end-fed structure can have much less current than the current maximum while still carrying the terminal current that closes the circuit and setting substantial RF voltage. A wire close to lossy ground can change input impedance, dissipate power and contribute a field. Which of those dominates depends on its height, soil, routing, frequency and the rest of the return network.

The full current system can include displacement current to the environment, the counterpoise, transformer capacitance, the tripod or support, and current on the outside of the coax. The radiator label belongs to the resulting current distribution, not to one piece of metal chosen in advance.

The Feed Line Is a Measured Boundary, Not a Promise

The PDF makes a feedpoint choke mandatory so the coax cannot become a second counterpoise. The design goal is reasonable: define the antenna-side return path and reduce unwanted current on the coax exterior. The universal prescription is not.

A choke is qualified by its complex common-mode impedance across the operating band, the installed current and voltage, its self-resonances, cable geometry, dissipation, temperature and insulation margin. A catalogue suppression number alone does not prove that exterior current is negligible in this installation.

NIST’s review of antenna-measurement challenges notes that unwanted common-mode current on a coaxial feed can radiate and change a measured pattern. That gives us a practical test: measure exterior current at several positions, then repeat after a controlled change in choke position, impedance or feedline route. If the current or pattern changes materially, the feed line was participating under the first condition.

Do not prescribe one universal current threshold. State instrument bandwidth and calibration, clamp location, cable route, transmit power and repeatability. The aim is a current small enough for the declared pattern, loss, EMC and exposure requirements—not a slogan that says the coax has vanished.

The 96% Calculation Is Real—and Narrow

For SWR S measured in a real reference impedance under the usual power-wave assumptions:

At 1.5:1, |Γ| = 0.2; reflected incident power is 4%; and net accepted power at that plane is 96%. Keysight’s network-analyser guidance uses the same SWR-to-reflection relationship.

That is mismatch efficiency, not antenna radiation efficiency. The active IEEE 145-2025 antenna terminology standard keeps the relevant quantities distinct. For a declared boundary:

A dummy load can accept almost all incident power and radiate almost none. Conversely, an efficient radiator can have a mediocre match before a network transforms its impedance. SWR cannot tell the two stories apart.

“Structural Efficiency” Needs Its Own Fence

The February PDF reports structural-efficiency values up to 99.5% while separately listing transformer-plus-choke loss of 1.08 dB or 0.51 dB for two versions. Those statements can coexist only if “structural” excludes the matching and choking network. That is a legitimate modelling boundary when it is named clearly; it is not complete-system efficiency.

The distinction becomes visible in a bounded calculation. If the listed insertion losses apply at the actual frequency, power, impedance and temperature, then:

If a 1.5:1 mismatch is then applied at the same consistently defined boundary, the corresponding incident-power ceilings before radiator and ground loss would be about 74.9% and 85.3%. This is an illustration using the PDF’s own component numbers—not a measured verdict on either assembled antenna. The result is valid only if the losses were measured with representative terminations and are combined without double-counting.

The current MyAntennas MEF-130-LP page advertises roughly 0.4 dB insertion loss from 1 to 30 MHz. To use that figure in a system power budget, publish the measurement method, reference planes, load, fixture, uncertainty and power/temperature conditions. A manufacturer component figure is not automatically the loss with one particular complex antenna load.

A Transformer Does Not Move Copper

Page 9 explains the 56:1 device as “looking further down the radiator.” That metaphor crosses the line into a different circuit. The physical feedpoint remains at the transformer terminals. Changing transformer topology or ratio changes the impedance relationship between its ports; it does not move the connection along the whip.

An ideal impedance ratio of 49:1 maps 2450 Ω to 50 Ω, and an ideal 56:1 ratio maps 2800 Ω to 50 Ω. A real HF transformer is a frequency-dependent two-port containing magnetising impedance, leakage, winding resistance, loss, parasitic capacitance and mode conversion. Its input impedance is the result of that network terminated by the antenna’s actual complex impedance.

C. L. Ruthroff’s original transmission-line-transformer paper shows why broadband transformer behaviour belongs to a network, not a scalar turns-ratio label. If changing transformer changes the tuned whip or counterpoise length, the complete network changed; the feedpoint did not walk along the radiator.

The PDF’s 4NEC2 plots use a 2450 Ω reference for SWR and reflection coefficient. That can be a useful ideal high-side design target. It is not the same measurement as 50 Ω S₁₁ at the radio side of a real 49:1 network. Keysight’s de-embedding guide explains that a VNA result belongs to its calibration plane and includes every fixture between that plane and the device unless those effects are characterised and removed.

Whip Length Alone Does Not Name the Antenna

The published field table uses whip-plus-pigtail lengths near 43% to 45% of a wavelength and a return wire near 33%. That does not prove the antenna is, or is not, a half-wave radiator. Physical length is an input; the complex current distribution over the full conductive system is the evidence.

A convincing classification would show current magnitude and phase on the whip, counterpoise, coax exterior, support and any modelled loss branches. It would also disclose how the wire lying on or near earth is represented. The current maximum may indeed fall near the whip’s middle, but that does not make the counterpoise field and ground loss zero.

Here the PDF’s own tuning experience is valuable. The linked wire materially shifts resonance and enables a shorter whip. The technically modest conclusion is strong enough: this is a coupled asymmetric radiator-and-return system that can be tuned on one band at a time. A textbook name should come after the current plot, not before it.

A 4NEC2 Plot Is a Conditional Prediction

Lawrence Livermore National Laboratory describes NEC as a numerical code for currents, fields and patterns of a specified antenna model. The LLNL NEC-5 validation manual explicitly separates numerical modelling error from physical modelling error caused by simplifying the real antenna and its environment.

For the Dominator comparison, publish enough material to reproduce the calculation:

• the 4NEC2 file, expanded NEC deck, engine and version;
• wire coordinates, diameters, conductivity, segmentation and source segment;
• counterpoise height or contact with ground, sag and azimuth;
• ground method, conductivity and relative permittivity;
• tripod, mount, pigtails, coax exterior and nearby conductors included or deliberately omitted;
• transformer equivalent network, loss and high-side reference impedance;
• choke common-mode impedance rather than only a perfect break;
• accepted-, incident- or radiated-power normalisation; and
• segmentation convergence and sensitivity to ground, height, routing and component values.

The February PDF presents plots but not the model deck and raw output alongside them. That prevents an independent check of the power boundary, omitted structures and convergence. It does not make the plot false; it limits the claim to a reported result that has not yet been reproduced.

What the Displayed 15-Metre Numbers Actually Say

The paper’s opening offers up to twice the gain, or +3 dB, relative to a quarter-wave. The 15-metre comparison on pages 15 and 16 is more modest:

• the peak figures in the table are +0.67 dBi for the Dominator and −0.67 dBi for the omnidirectional PERformer quarter-wave—a 1.34 dB difference;
• at 18° elevation, the plotted values are +0.68 dBi and −1.00 dBi—a 1.68 dB difference; and
• the standalone 15-metre Dominator plot on page 4 states +0.60 dBi at 18°, while later pages show +0.67 or +0.68 dBi.

Those are not trivial differences, but that displayed case does not show +3 dB. The several values may reflect different runs, geometry, loss settings, angular samples or rounding. Mapping each figure to its exact file would settle that question. A +3 dB maximum could still exist under another disclosed band or condition; the run demonstrating it needs to be published.

The model’s 18° peak versus 24° for the omnidirectional quarter-wave is also a conditional result. A six-degree modelled shift can be useful, but it inherits the model’s ground, feedpoint height, radial/counterpoise geometry, feedline treatment and angular resolution. It is not automatically the installed angle on every soil.

A Pattern Does Not Assign “Regional” or “Global” Reach

A lower elevation lobe can favour some longer ionospheric paths, while energy at higher angles can favour other path lengths. Neither one angle nor one peak gain assigns a portable antenna a permanent geographical category. Frequency, path, time, season, ionosphere, absorption, noise, polarisation, terrain and the receiving system remain in the link budget.

ITU-R P.533-14, the current in-force method for predicting HF circuit performance, exists because HF reach is a path-and-time problem. Antenna pattern is an important input to that problem—not a substitute for it.

The fairest wording is therefore specific: under the declared 15-metre model, the Dominator concentrates more radiation than that particular omnidirectional quarter-wave model below 30° and has its modelled peak at a lower elevation. “Global” is an operating outcome to be observed across propagation conditions, not a lobe label.

Field SWR Is Field Evidence for Match

The February PDF’s field SWR captures and tuning table are useful. They show that the listed whip and counterpoise settings are credible starting points and that the built system can reach a low input mismatch in the author’s environment.

They do not measure radiation efficiency, transformer temperature, common-mode current, absolute gain or elevation pattern. On-air contacts are equally real and equally bounded: they prove communication happened, but propagation and the distant stations prevent one contact map from becoming a calibrated gain comparison.

The active IEEE 149-2021 recommended practice is the formal framework for antenna-property measurements. NIST’s antenna-measurement uncertainty work also identifies theory, simulation and controlled changes to the measurement system as parts of a defensible uncertainty analysis.

A Field Comparison That Can Answer the Question

For practical portable evaluation, a paired A/B/A design can test the claim without pretending to be an antenna range:

1. A — establish the reference. Record the reference antenna’s complete geometry, ground, feedline, choke, complex input impedance and exterior current. Measure transmitter power at a declared plane and calculate accepted power.
2. B — switch quickly to the Dominator. Preserve transmitter, frequency, receiver, bandwidth and test interval. Correct for network and feedline loss at the same power boundary.
3. A — restore the reference. Confirm that the first condition returns; otherwise fading, drift or an installation change can masquerade as antenna gain.
4. Cross-swap the positions. Repeat with the two antenna locations exchanged to reveal local ground and obstruction bias.
5. Repeat across paths and angles. Use multiple fixed receivers or beacons, several azimuths and many rapid pairs. Report the distribution of paired differences, not only the best report.
6. Measure the mechanisms. Keep complex S₁₁, transformer/choke loss or temperature, and coax-exterior current beside the field records so a level difference can be interpreted.

WSPR or FT8 reports can supply many observations, but the switching schedule, transmitter power, frequency, receiver identity, report quantisation, propagation drift and rejected data must be documented. They can support a field-utility comparison. Absolute gain or a complete elevation pattern still calls for a calibrated method appropriate to IEEE 149.

Where I Land

The Dominator idea remains attractive: one-band-at-a-time tuning, a long telescoping whip, a deliberate return wire and a compact support are practical portable engineering. The February 2025 PDF also gives readers enough detail to build and experiment. That is worth preserving.

My pushback is equally specific:

• “no radials” must not be read as “no return conductor”;
• a counterpoise that moves resonance is part of the antenna system even if its current is below the whip maximum;
• 1.5:1 SWR gives 96% accepted incident power at a plane, not 96% radiation efficiency;
• structural, network, radiation, mismatch and complete-system efficiency need separate boundaries;
• a transformer ratio changes impedance mapping, not physical feedpoint position;
• the displayed 15-metre plots show a useful but smaller advantage than +3 dB;
• NEC gain and angle remain conditional until the model package and sensitivity are disclosed; and
• field SWR and contacts validate tuning and usability, while gain, pattern and efficiency need claim-matched measurements.

Primary Technical References

• Greg Mihran, KJ6ER — Dominator Halfwave Antenna, February 2025 revision
• IEEE 145-2025 — Standard for Definitions of Terms for Antennas
• IEEE 149-2021 — Recommended Practice for Antenna Measurements
• Lawrence Livermore National Laboratory — Numerical Electromagnetics Code
• LLNL — NEC-5 Validation Manual
• NIST — antenna-measurement challenges, including common-mode feed-cable current
• Keysight — VNA reference planes and de-embedding
• ITU-R P.533-14 — Method for the prediction of HF circuit performance
• ICNIRP 2020 — Guidelines for limiting exposure to radiofrequency electromagnetic fields

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