NECtacy in the Park: When Model Precision Outruns Evidence
NECtacy in the Park: When Model Precision Outruns Evidence
Greg Mihran, KJ6ER, published a practical elevated quarter-wave portable antenna and a detailed 4NEC2 case for it. The useful question is not whether NEC belongs in the park. It is which conclusions follow from the declared model—and which still belong to the installed antenna.
RF.Guru working definition: Common-mode current is the non-cancelling phasor-sum current in a specified set of conductors, evaluated at a defined cross-section and using a declared current-direction convention. In the intended differential transmission-line mode, the outgoing and return currents are equal and opposite, so their phasor sum is zero. When they do not cancel, the remaining current must close through another reference or return path—such as the outside of a coax shield, a mast, equipment chassis, station wiring, nearby structures, earth, the operator, or distributed coupling through the environment.
This broader working definition is especially useful in practical antenna systems. On transmit, non-cancelling current on the outside of the coax can make the feedline and connected structures part of the radiating antenna system unless that path is intentional, clearly defined and properly controlled—for example by providing the required return path and placing a suitable common-mode choke at the correct boundary.
The February 2025 PERformer (Portable, Elevated, Resonant) Antenna document gives readers dimensions, a parts list, field-tuning guidance, SWR photographs and modelled patterns. That is exactly the kind of openness that makes an engineering discussion possible.
The design is a one-band-at-a-time elevated quarter-wave vertical for 40–6 metres. Its feedpoint is roughly 4–5 feet above ground, with two tuned elevated radials normally separated by 90°. KJ6ER presents the 90° arrangement as a directional option and the 180° arrangement as the more omnidirectional alternative.
I like the mechanical idea. Elevated tuned radials can reduce the loss associated with a sparse ground-coupled return system, and a telescopic radiator makes deployment quick. My concern begins when the precision of one model run starts sounding like a portable-site guarantee.
The engineering boundary: NEC predicts the currents and fields of the geometry, materials, ground, sources, loads and networks that were actually entered. A field deployment adds whatever the model omitted: real soil, uneven terrain, tripod, stakes, feed line, choke, operator and nearby conductors. A precise output remains conditional on that boundary.
What the PERformer Document Asks the Numbers to Carry
The document joins several different propositions:
- the elevated system is described as more than 90% efficient, while a “typical” ground-mounted quarter-wave with four ground-coupled radials is assigned 37%;
- the 90° radial pair is credited with roughly 0.3–0.5 dBi maximum gain and 3–4 dB front-to-back ratio;
- a 15-metre run lists +0.31 dBi peak gain at 24° elevation, 3.37 dB front-to-back ratio and a reflection magnitude displayed at −51.2 dB;
- the 90° and 180° radial arrangements are compared with angle-specific differences of 0.98, 2.36 and 5.27 dB; and
- 1.5:1 SWR is translated to 96% “power throughput efficiency,” alongside the broader proposition that resonance makes an antenna best and most efficient.
Those values are not impossible. They simply belong to different quantities and evidence classes. Keeping them separate makes the model more useful, not less.
NEC Solves the Declared Electromagnetic Problem
The Lawrence Livermore National Laboratory description of NEC is admirably concrete: the code models wires and conducting surfaces, sources, homogeneous ground, loads, networks and transmission lines, and can output currents, near fields and radiation patterns. That list also tells us what must be declared.
For a portable vertical, a reproducible model record should include:
- the complete geometry and source position, including both radials and their height profile;
- wire radii, conductivity and any lumped or distributed loss;
- ground formulation, conductivity and relative permittivity;
- solver and version, segmentation, junction treatment and convergence checks;
- frequency sampling and the normalization used for pattern comparisons;
- the modelled feed line, mast, tripod, choke or an explicit statement that they are outside the boundary; and
- the exact input deck and output run tied to every published plot.
The linked 16-page PERformer revision publishes useful geometry and output pictures, but it does not include the NEC input deck needed to reproduce the displayed runs. That means the plots can be read as documented predictions, not independently rerun results.
Convergence matters as much as presentation. A polished polar plot can still move when segment density, source segmentation, junction geometry, ground parameters or an omitted conductor changes. The LLNL NEC-5 Validation Manual exists because computational results need verification against known cases and sensitivity to implementation—not because colourful output is self-validating.
The Efficiency Comparison Needs One Shared Boundary
Radiation efficiency is the fraction of accepted power that becomes radiation:
ηrad = Pradiated / Paccepted
Paccepted = Pradiated + Pconductor + Pground + Pnetwork + …
A model may report radiation efficiency for its declared conductors and ground. It cannot include an unmodelled choke, connector, feed-line exterior, wet tripod interface or contact resistance. “System efficiency” is therefore meaningful only after the system boundary is named.
The greater-than-90% elevated case and the 37% four-ground-radial case are not a portable design rule by themselves. A fair comparison needs the same frequency, radiator, conductor properties, accepted-power plane, soil, installation envelope and loss model. Four short lossy wires on difficult soil are not a universal stand-in for every ground-mounted quarter-wave. Likewise, two elevated tuned radials do not guarantee one efficiency across bands and sites.
Rudy Severns, N6LF, showed why the useful conclusion is conditional. His controlled elevated-ground-system work examines radial height, number, tuning and ground interaction under declared conditions. It supports elevated radials as a serious design option. It does not turn one radial count or one efficiency figure into a universal constant.
At 1.5:1 SWR, the 96% Is Real—and Narrow
For a real reference impedance and a one-port measurement at a declared plane:
|Γ| = (SWR − 1) / (SWR + 1)
At 1.5:1, |Γ| = 0.2, reflected incident power is |Γ|² = 0.04, and mismatch acceptance is 1 − |Γ|² = 0.96.
The corresponding mismatch loss is approximately −10 log10(0.96) = 0.177 dB.
So 96% of incident travelling-wave power crosses that reference plane under the stated assumptions. The calculation does not say how the accepted power divides among radiation, ground, conductors, choke, feed line or other losses.
The −51.2 dB reflection result displayed for the 15-metre model is even narrower. Interpreted as the magnitude of the reflection coefficient in decibels, it corresponds to |Γ| ≈ 0.00275 at that exact simulated frequency. It is an exceptionally close simulated match. It still does not measure radiation efficiency or absolute field strength.
Resonance is similarly specific: it means the input reactance is zero at the chosen reference plane. A resonant structure can be lossy; a non-resonant structure can be efficient when a low-loss matching network transforms its impedance. “No external tuner required” is a useful deployment feature, not an efficiency proof.
Gain, Front-to-Back Ratio and RDF Answer Different Questions
Using the terminology boundary formalised by IEEE 145-2025:
Gain(θ,φ) = ηrad × Directivity(θ,φ)
Realised gain(θ,φ) = (1 − |Γ|²) × ηrad × Directivity(θ,φ)
Front-to-back ratio compares two directions in the pattern. A 3.37 dB ratio is approximately a 2.17:1 power ratio between the declared front and back samples. It does not say whether the front increased, the back decreased, or both. It therefore cannot be added casually to forward gain.
Receiving directivity factor, often discussed as RDF in amateur receiving work, is another pattern-derived quantity. It can be useful for ranking rejection of spatially distributed noise under a defined pattern, but it is not transmitter gain and it does not include site noise, polarisation mismatch or receiver overload.
The PERformer’s asymmetric radial geometry can skew the modelled azimuth pattern. That is a legitimate hypothesis. The important follow-up is whether the installed current distribution reproduces the predicted skew when the feed line, choke, support and environment are present.
Small Delta Tables Need the Run Behind Every Cell
The comparison between 90° and 180° radial spans lists differences at selected elevation angles. These are model deltas, not automatically field advantages. Before interpreting them, readers need to know:
- whether both patterns were normalized to the same incident, accepted or radiated power;
- whether conductor and ground loss changed between configurations;
- whether the feed line and common-mode boundary were identical;
- whether each delta is a change in absolute gain or a comparison against a deeper null in one pattern; and
- how the result moves with soil, height, radial sag, routing and nearby conductors.
A 5.27 dB difference at one upward angle can be mathematically correct while saying little about peak gain, useful path gain or total radiated power. A selected angle may sit near a null in one case. The full, equally normalized patterns matter.
This is where uncertainty stops being academic. NIST’s antenna-measurement uncertainty guidance identifies alignment, mismatch, multipath, instrumentation and repeatability as contributors to the result. In a park, soil and setup repeatability join that list. If the claimed advantage is about 1 dB, the experiment needs to show that its combined uncertainty and deployment variation are smaller than the effect.
A Field Test Must Measure the Claimed Quantity
SWR photographs show that the physical antenna could be adjusted to a useful input match at the analyser plane. Successful contacts show that communication paths closed. Neither measurement yields radiation efficiency, absolute gain or front-to-back ratio.
IEEE 149-2021 treats radiation pattern as a fundamental antenna property and describes the facilities and instrumentation needed to measure antenna properties. NIST also warns that the antenna-to-range interface can alter an omnidirectional antenna’s measured pattern when common-mode feed-cable current is not controlled.
A practical PERformer comparison can still be done without pretending a picnic table is an anechoic range:
- declare the tested frequency, radial geometry, height, soil state, coax route, support and choke;
- calibrate the reference plane and record complex input impedance, not only the lowest SWR number;
- measure exterior coax current at repeatable locations and repeat after moving the feed line;
- hold accepted power constant when comparing 90° and 180° configurations;
- use rapid A/B/A switching or repeated rotations against stable reference signals;
- measure the stated forward and rear directions, relevant elevation coverage and polarisation;
- cross-swap positions or repeat at several comparable sites; and
- publish the distribution and uncertainty, not only the best cycle.
For efficiency, add an accepted-power and radiated-power method appropriate to the frequency and antenna. For absolute gain, use a calibrated reference and a geometry that supports the measurement. For a model-validation claim, compare more than one observable: impedance, currents and pattern are much stronger together than any single trace.
Where I Land
The PERformer is a credible portable antenna concept. It is compact, tunable, mechanically clear and built around elevated return conductors that deserve serious consideration. KJ6ER’s willingness to publish dimensions, plots and field traces gives builders a useful starting point.
The named efficiency, gain and angle-specific pattern values belong to the displayed model until the input deck, sensitivity analysis and claim-matched field measurements carry them across the boundary. The SWR traces establish match at their measurement plane. They do not certify radiation efficiency, absolute gain or a compass-heading advantage.
That is my entire argument: use NEC hard. Publish the deck. Perturb the inputs. Measure the installed currents. Then take the polar plot to the picnic table and see which parts survive.
Primary Technical Sources
- Greg Mihran, KJ6ER: PERformer Antenna, February 2025 revision
- Lawrence Livermore National Laboratory: Numerical Electromagnetics Code v5.0
- Lawrence Livermore National Laboratory: NEC-5 Validation Manual
- IEEE 145-2025: Standard for Definitions of Terms for Antennas
- IEEE 149-2021: Recommended Practice for Antenna Measurements
- NIST: Antenna Measurement Challenges and Common-Mode Feed-Cable Current
- NIST: Estimating Uncertainties in Antenna Measurements
- Rudy Severns, N6LF: Experimental Determination of Ground-System Performance for HF Verticals, Part 1
Mini-FAQ
- Is NEC unsuitable for portable antennas? — No. NEC is valuable when the geometry, ground, materials, sources and networks are declared. The output predicts that model, not every future deployment.
- Does 1.5:1 SWR mean 96% antenna efficiency? — No. It means 96% mismatch acceptance at the stated reference plane. Radiation efficiency separately compares radiated power with accepted power.
- Can two elevated radials produce directionality? — Yes. Their geometry can skew the current distribution and pattern. The installed magnitude and repeatability need feed-line-current and field measurements.
- Is front-to-back ratio the same as forward gain? — No. Front-to-back ratio compares two directions. It can improve because the front rises, the back falls or both; absolute gain needs efficiency and calibrated field evidence.
- Do SWR photographs validate a NEC pattern? — They validate the measured input-match behaviour at their reference plane. Pattern, gain and efficiency require different measurements.
- What would make a modelled one-decibel advantage persuasive? — Publish the input deck, convergence and sensitivity checks, then show repeated calibrated A/B/A field measurements with accepted power, current paths and uncertainty controlled.