Delta Loop vs Dipole Directionality

The usual question is whether a full-wave delta loop is “more directional” than a half-wave dipole. My answer is deliberately inconvenient: not from the names alone. The two antennas distribute current over different geometries, but directionality belongs to the complete installed field—not to the words triangle, loop or dipole.

Directionality Needs More Than One Number

A radiation pattern is a frequency- and polarization-specific function of direction. One azimuth cut cannot describe the elevation pattern; one elevation cut cannot reveal every azimuth null; a co-polarized plot does not show the cross-polarized field. Directivity, gain and realized gain are different again because efficiency and input mismatch enter at different stages.

This matters when someone reports that a loop is “more directional.” They may mean a deeper null in one horizontal cut, a stronger signal in one compass direction, a lower elevation maximum, a different polarization, or simply a better complete link on that day. Those are not interchangeable claims.

The Far Field Comes From Distributed Current

Every conductor segment carrying RF current contributes a field with magnitude, phase, direction and polarization. The far field is the vector sum of those contributions. Alter the wire route, feed location, support, loading, nearby metal, feed-line exterior current or ground interaction and the sum can change.

A thin, straight, centre-fed half-wave dipole in free space gives the familiar broadside maximum and axial nulls. That is a useful reference, not a promise for a bent dipole over real ground. A resonant one-wavelength loop in a declared plane and mode commonly produces broadside lobes, but its exact pattern follows the non-uniform current around the perimeter. A triangular loop is not electromagnetically identical to a circular or square loop merely because the circumference is similar.

The phrase “single element” also does not set a universal directivity ceiling. Electrical size and current phase along the structure matter. What is safe to say is narrower: neither an ordinary dipole nor an ordinary single full-wave loop becomes a steerable high-gain array simply because of its name.

A Closed Wire Still Supports Standing Waves

The dipole has open ends, so current approaches zero at those ends in the ideal thin-wire model. A loop closes on itself except at the driven gap, so it has a different boundary condition and a periodic conductor path. It still supports a standing-wave current distribution. “Current circulates continuously” is not enough to calculate the field, and the small-loop magnetic-dipole picture must not be pasted onto a resonant loop whose circumference is near a wavelength.

The driven gap is a real electromagnetic port. Voltage and current at that port depend on perimeter, conductor diameter, shape, frequency, feed position, nearby structures and the selected mode. Moving the feed can change the impedance transformation seen at the port; in an imperfectly symmetric installation it can also change which currents are excited on the loop, feed system and supports.

Feed Position Does Not Carry a Universal Polarization Label

A vertically mounted delta loop contains sloping or vertical conductor components and at least one horizontal component. Each can contribute to the radiated field. Feed position can change the magnitude and phase of the resulting currents, so corner feed and centre-of-side feed often produce different polarization mixtures in practical models.

That observation is useful, but the slogans “corner feed is vertical” and “bottom-centre feed is horizontal” are not complete specifications. Apex direction, perimeter, wire height, feed-line route, symmetry, ground, mast and nearby conductors all matter. If polarization matters, inspect the complex currents and calculate or measure both co- and cross-polarized patterns.

Feedpoint Height Is Not Radiating Height

A delta loop can place its feedpoint near a lower corner while much of the current-carrying wire extends higher. A centre-fed horizontal dipole places its feedpoint at the physical centre, but the pattern still comes from current along the complete wire. Comparing only feedpoint heights therefore creates a false equivalence.

A fair geometry comparison records at least top height, bottom height, footprint, wire orientation, total conductor length and current distribution. You may instead choose equal mast height, equal highest point, equal wire budget or equal available garden. Each is a legitimate engineering constraint, but each answers a different question.

The often-repeated low-angle story needs the same care. A vertically polarized component and a horizontal dipole component reflect differently from real ground. Soil conductivity and permittivity, incidence angle, terrain and height in wavelengths shape the elevation interference. A low horizontal dipole often favours higher elevation angles than a taller or vertically polarized installation, but no fixed ranking survives every band, soil and geometry.

The Feed System Can Become Part of Either Antenna

A nominally balanced radiator driven through coax needs a transition that controls unwanted exterior current when that current is significant. The need is not proved by the antenna label, and a choke is not automatically successful because it is present.

Common-mode current can flow on the outside of coax, a mast, support wire, enclosure or station wiring. Those currents may tilt polarization, fill an intended null, skew an azimuth pattern and change both transmitted and received noise behaviour. The result is the pattern of the complete current-carrying structure.

Measure exterior current over the operating band, install an appropriate common-mode impedance where the current-path model calls for it, and measure again. Include the feed line and mast in the numerical model until measurements show that their currents are negligible for the claim being made.

Compare Equal Questions, Not Convenient Installations

A loop may win because it uses the available support height better, offers a convenient low feedpoint, produces a useful polarization mixture, or avoids lossy loading. A dipole may win because it is simpler, easier to balance, easier to model, or puts its broadside response exactly where needed. None of those outcomes proves a universal shape hierarchy.

Model First, Then Measure the Claimed Quantity

LLNL’s Numerical Electromagnetic Code can solve currents, near fields and radiation patterns for declared wires, conducting surfaces, loads, networks, transmission lines and ground. Model the dipole and loop with the geometry you can install, not with unrelated catalogue sketches. Check segmentation and convergence, and inspect the current on every conductor before trusting the colour plot.

A defensible pattern measurement uses a suitable range or a valid near-field-to-far-field method. Record the coordinate system, frequency, polarization, power reference, alignment, cable route, reflections, dynamic range and uncertainty. IEEE 149-2021 treats antenna measurement as a controlled procedure, while NIST’s antenna-measurement work makes the uncertainty sources explicit.

On-air reports can still be valuable. Use rapid switching or simultaneous receivers, hold transmitter and receiver states fixed, repeat across paths and time, and record wanted signal and noise separately. Call the outcome what it is: a complete installed-system observation, not an isolated 3D-pattern measurement.

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