Polar Plot vs Picnic Table

The PERformer design note by KJ6ER describes a portable quarter-wave vertical with its feed point roughly 4–5 feet above ground and two tuned elevated radials separated by 90 degrees. Its model shows a preferred azimuth and proposes aiming the radial sector toward the wanted coverage area. That is a clear, testable engineering proposition.

I am not interested in proving that the antenna can or cannot make contacts. Almost any sensible quarter-wave vertical can make contacts. My question is narrower and harder: does the useful, aimable pattern remain when the design leaves the clean polar plot and lands beside a picnic table?

The Proposition Worth Testing

The linked PERformer note gives enough detail to define the proposition: a resonant, one-band-at-a-time vertical; two tuned elevated radials at 90 degrees; a feed-point choke; and modeled forward gain and front-to-back behaviour at stated elevation angles. Those are model results for the geometry, ground and feed assumptions used in that study. They are not automatically the pattern of every installation carrying the same antenna name.

That distinction is not pedantry. “Front” needs an azimuth-zero convention. Front-to-back ratio needs a stated elevation cut and an exact back direction. Directivity, gain and realized gain are different quantities, and none is interchangeable with radiation efficiency. A pattern can be redistributed toward one bearing without increasing total radiated power.

The current IEEE 145-2025 antenna terminology supplies those distinctions. IEEE 149-2021 makes the matching measurement point: an antenna pattern is a property to be measured under controlled geometry, instrumentation and range conditions.

Two Radials Deliberately Break Rotational Symmetry

Two radial wires separated by 90 degrees do not have the rotational symmetry of three wires spaced by 120 degrees or four wires spaced by 90 degrees. In an otherwise controlled model, that asymmetry can change radial currents and reshape the azimuth pattern. That is the mechanism behind the PERformer proposition.

A more symmetrical radial fan is not magic either. Equal wire lengths and equal angular spacing do not force equal currents when one wire crosses wet ground, another passes a metal leg, and the feed line leaves in a third direction. The electromagnetic boundary conditions include conductor height, soil, nearby objects, feed-line routing, choke impedance and every coupled current path. Geometry gives symmetry only when the rest of the relevant environment respects it.

Rudy Severns, N6LF, explored precisely this sensitivity in Part 1 and Part 2 of his elevated-ground-system study. His models and cited measurements show that sparse elevated radial systems can have unequal, frequency-dependent radial currents; added radials reduce sensitivity, but four physical radials do not guarantee equal current division. That is the useful N6LF context—not a universal verdict for one portable antenna.

Return Current Completes the Antenna

At an isolated two-terminal feed point, current leaving one terminal returns at the other. In a real coax-fed vertical, however, the intended differential-mode current on the centre conductor and inner surface of the shield can coexist with current on the shield’s exterior. That exterior current follows a separate common-mode path through the feed line, radio, battery, operator and capacitance to the surroundings.

Once exterior shield current is significant, “vertical plus two radials” is no longer the complete radiating structure. Moving the coax can then change feed-point impedance, exterior current and pattern. A good SWR at the transmitter does not prove that the feed line stayed out of the antenna.

The open-access measurement study by Constantin and Tamas, Evaluation and Impact Reduction of Common Mode Currents on Antenna Feeders in Radiation Measurements, demonstrates that feeder common-mode current can distort measured radiation patterns and must be evaluated as part of the test arrangement. The practical ARRL common-mode current and choke article also shows why a current probe and a frequency-dependent choke measurement are more useful than assuming a coil has one fixed choking value.

A Choke Defines One Boundary, Not the Whole Park

A feed-point choke inserts a complex, frequency-dependent impedance into one common-mode loop. Its effect depends on the loop impedance, its location, cable length and routing, frequency, nearby conductors and the choke’s own resistance and reactance. It may reduce exterior current substantially; it cannot be treated as a universal off-switch.

Measure exterior feed-line current at several positions and repeat after any choke or routing change. If a second choke is investigated, choose its position from the measured current distribution and the boundary you intend to create—not from a universal distance rule. Ian White, GM3SEK, gives useful construction and verification context in Cost-effective Ferrite Chokes and Baluns.

The Picnic Table Is Already in the Antenna

A clean model is not dishonest. It answers the question encoded in its geometry. The trouble starts when we silently ask it to answer a different question.

A park installation can add a metal or wet wooden table, tripod, mast, guy lines, power lead, transceiver case, operator, shelter, fence, slope and non-uniform soil. Some objects couple weakly; others become an important current path or scatterer. The point is not that “the environment always wins.” The point is that the magnitude and sign of its effect are installation-specific and must be tested.

Recommendation ITU-R BS.705-2 treats ground, site structures and practical variation as explicit influences on HF patterns. Lawrence Livermore National Laboratory’s NEC 5 description and validation material make the corresponding modelling boundary clear: the code can include wires, surfaces, loads, networks, transmission lines and ground, but physical-model error remains when material parts of the real installation are simplified away.

How to Test the Aimable-Pattern Proposition

First write down what “aiming” means. If the operating instruction rotates only the radial sector while the radiator, mast and coax stay fixed, test that exact operation. If the whole antenna is rotated, rotate the complete radiating structure while keeping the measurement range and reference plane controlled. Do not rotate the radial fan in one trial and the coax route, operator and table in another.

A useful test plan includes:

• one frequency, declared radiator and radial dimensions, feed-point and radial heights, wire type and radial angle;
• measured soil or at least recorded ground condition, slope and nearby structures;
• a fixed coax type, length, departure angle and route, with measured choke impedance at the test frequency;
• accepted power at a declared antenna reference plane, rather than transmitter setting alone;
• exterior coax-current measurements at repeatable positions;
• co-polar and, where relevant, cross-polar field measurements over azimuth at stated elevation geometry;
• front, back, azimuth zero, normalization and uncertainty declared before examining the result; and
• rapid A/B/A repeats, followed by repeated placement or a second representative site.

For a local range, use a calibrated source/receiver chain or calibrated remote field probes with stable geometry. For on-air HF observations, ionospheric fading, path changes, receiver AGC and uncontrolled station antennas can overwhelm a few-decibel pattern difference; multiple simultaneous receivers and repeated switching are much stronger evidence than signal reports collected minutes apart.

A Fair Radial Comparison

Compare the 90-degree pair with a 180-degree pair and with three- or four-radial fans only after deciding what stays equal. Radial count, wire length, total copper length, radial height and tuning method are different experimental constraints; they do not all stay constant at once. State the chosen constraint and retune only where the operating procedure permits it.

For every configuration, record complex feed-point impedance, accepted power, exterior feed-line current, pattern at the same frequency and elevation geometry, and an efficiency or gain measurement if the conclusion requires one. A lower SWR is not proof of greater efficiency. A higher peak in a normalized plot is not proof of more radiated power. A larger front-to-back ratio is not automatically a better portable antenna if the wanted paths occupy other bearings or elevation angles.

A more symmetrical radial fan may prove less orientation-sensitive at a particular site. The 90-degree pair may preserve a useful bias. Either outcome is credible when it repeats with declared uncertainty. What is not credible is replacing the measurement with a universal rule that two radials always steer or four radials always win.

What Would Convince Me

I would call the PERformer’s aimable pattern practically useful when rotating the prescribed geometry produces a repeatable change at the declared bearings and elevation, the effect remains larger than the measurement uncertainty and site-to-site scatter, exterior feed-line current stays within the recorded boundary, and accepted power is equalised. I would also want the measured direction to follow the model’s convention rather than a convenient receiver that happened to fade upward.

That is not hostility toward a clever two-radial idea. It is the respect due to a directional claim. Put the picnic table into the test, rotate what the operator is actually told to rotate, and let the current and field measurements decide.

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