Full-Wave Antennas: Geometry, Feedpoint and Pattern Decide

An exactly one-wavelength straight dipole is difficult to feed at its centre because that point approaches a current node. Other structures containing roughly one wavelength of conductor can place the feed near a current maximum, close the path into a loop, couple conductors as a folded dipole or use length for directional travelling-wave operation. Geometry, feedpoint and feed system define the result.

“Full Wave” Is a Length, Not a Topology

Electrical length describes phase change along a conductor or current path. It does not by itself specify the conductor shape, feed position, current distribution, impedance or radiation pattern.

That is why “it uses one wavelength of wire” cannot predict whether an antenna is easy to match, efficient, directional, quiet on receive or suitable for a particular path.

Why the Half-Wave Dipole Is Convenient

A thin half-wave dipole in free space has a current maximum near its centre and current nodes at its ends. The ideal feed resistance is about 73 Ω; practical trimming to remove reactance and the influence of height, diameter and surroundings commonly move the input elsewhere in the broad 50–75 Ω region.

Its ideal free-space directivity is about 2.15 dBi. The broadside figure-of-eight pattern is easy to understand, and a balanced feedpoint can be connected to coax through suitable common-mode control. These properties make it convenient—not uniquely capable of efficient radiation.

The Rohde & Schwarz Antenna Basics guide separates directivity from gain and emphasizes that radiation resistance must be associated with a stated location, commonly the feedpoint or a current maximum.

The Centre-Fed Full-Wave Dipole

For an ideal thin straight dipole of exactly one wavelength, the standing-wave current has:

• current nodes at both open ends;
• another current node at the centre; and
• current maxima approximately one quarter wavelength in from each end.

Putting the feed terminals at the centre therefore puts them at a voltage maximum and current minimum. In the infinitesimally thin, exact-length model, terminal current tends to zero and the centre-feed radiation resistance tends to infinity. A real wire has finite diameter, a feed gap, losses, supports and imperfect electrical length, so the measured impedance is finite—but it can still be thousands of ohms and can change rapidly with frequency and surroundings.

Virginia Tech’s current Radio Systems Engineering text explicitly distinguishes the centre terminals from terminals placed at a current maximum. MIT antenna notes illustrate the full-wave current distribution and why the feed reference matters.

Connecting this high-impedance point directly to 50 Ω coax creates severe mismatch. The transmitter may reduce power, feedline attenuation rises because forward and reflected waves increase conductor and dielectric loss, and high RF voltage stresses insulation and matching components. That is a feed-system problem, not proof of poor radiation efficiency.

A high feed resistance does not automatically mean high loss. Radiation resistance and loss resistance must be referred to the same feed current. The useful comparison is accepted power and realised gain after matching, feedline, conductor and environmental losses are included.

The Full-Wave Dipole Pattern

A straight full-wave dipole still has broadside maxima in the ideal free-space model. Its beam is narrower than that of a half-wave dipole, and its ideal directivity is about 3.8 dBi rather than 2.15 dBi—roughly 1.65 dB more directivity.

That number is not guaranteed installed gain. Directivity describes how radiated power is redistributed in angle. Gain includes radiation efficiency, and realised gain also includes mismatch at the stated reference plane:

The full-wave dipole concentrates somewhat more power broadside and less elsewhere; it creates no free power. The analytical pattern for a finite centre-fed wire is developed in this University of Texas treatment of linear antennas.

Moving the feedpoint near one of the current maxima makes impedance transformation more practical without fundamentally changing the ideal standing-wave distribution. In a real installation, however, the asymmetric feed, transformer, feedline exterior and nearby conductors must be included because unwanted common-mode current can alter pattern and impedance.

Close the Wire into a Loop and the Circuit Changes

A resonant one-wavelength loop is not a straight full-wave dipole bent until the ends touch. Closing the conductor changes its boundary conditions and current distribution.

Near the fundamental loop resonance, common square, circular and triangular loops often have feed resistance somewhere around 100–150 Ω. The exact impedance depends on shape, conductor diameter, feed gap, feed position, height, ground and nearby objects. A geometric perimeter of exactly one free-space wavelength is not a universal cutting length; trim or model the loop in its installed environment.

A full-wave loop normally has two broadside lobes perpendicular to its plane. Ideal free-space directivity for a circular ring is about 3.5 dBi, compared with 2.15 dBi for a half-wave dipole. The difference is modest and shape dependent. The Virginia Tech resonant-loop study shows why shape, circumference and feed geometry must be stated before quoting impedance or gain.

Horizontal Full-Wave Loops

A horizontal loop well below half a wavelength over real ground often favours high elevation angles. That can suit near-vertical-incidence skywave and regional HF work when the ionosphere supports those paths.

“Low loop equals NVIS” is still too simple. Very low height can increase ground interaction and loss; the exact elevation pattern depends on height in wavelengths, soil, shape and feedline current. NVIS coverage additionally depends on frequency relative to ionospheric critical frequency, absorption, time and path—not antenna angle alone.

A Naval Postgraduate School study modelled a one-wavelength horizontal quad loop for NVIS and medium-range HF communication. It supports the use case, not a universal height or coverage promise.

Raise the loop and lower-angle lobes become more important. Use it on higher bands, where its perimeter is multiple wavelengths, and the azimuth and elevation patterns develop additional lobes and nulls. Such a loop is not omnidirectional across all bands.

Multiband use can work well with low-loss balanced line and a suitably rated balanced matching system. A shack tuner can provide a comfortable transmitter impedance, but it cannot recover power already dissipated in a long, high-SWR coax run.

A horizontal loop may receive less local noise in one installation because of pattern, polarisation, balance or reduced feedline pickup. It is not inherently “quiet.” A loop with common-mode feedline current can collect plenty of household noise.

Vertical Delta and Quad Loops

Turn the loop vertically and its broadside lobes can support useful DX directions. The loop plane sets the two main azimuth headings, while feed position strongly influences polarisation.

• Feeding the centre of a horizontal side of a square or rectangular loop commonly favours horizontal polarisation.
• Feeding the centre of a vertical side commonly favours vertical polarisation.
• A bottom-centre-fed delta commonly favours horizontal polarisation.
• A lower-corner or appropriate quarter-wave-offset delta feed commonly favours vertical polarisation.

These are design starting points, not guarantees. Shape, height, feed gap, unequal current, ground and feedline common mode can create mixed polarisation and distorted patterns.

A closed loop has a two-terminal conductive antenna path and does not need a radial system in the way a ground-mounted monopole does. That does not make it immune to common-mode current. If the feed transition is unbalanced or the environment is asymmetric, the coax exterior can still become part of the antenna. Use suitable common-mode impedance and verify outside-coax current.

The earth remains part of the boundary. A vertical loop near lossy soil can experience pattern, feed-impedance and efficiency changes even though no radials are required.

Full-Wave Loops as Beam Elements

A cubical quad uses a driven full-wave loop by design. A somewhat larger loop can serve as a reflector and smaller loops as directors. Delta-loop arrays use the same broad principle with triangular conductors.

Forward gain and front-to-back ratio come from the coupled element currents and their phases, not from the word “full wave.” Element spacing, circumference, conductor diameter, boom, feed and surrounding structure determine the array.

The ARRL General Class study guide introduces quad and delta-loop arrays and the effect of feed position on polarisation. ARRL’s HF loop antenna collection provides practical examples.

The trade-off is often mechanical: more conductor, larger spreaders, greater wind and ice load, more support points and a larger turning radius. Electrical merit does not solve an unsafe structure.

The Folded Dipole Proves That Wire Counting Fails

A conventional folded dipole uses two closely spaced half-wave conductors joined at their ends. Its total conductor length is roughly one wavelength, but its end-to-end span and pattern remain those of a half-wave dipole.

For equal conductor diameters and close spacing, the feed impedance is approximately four times that of the corresponding ordinary dipole—often near 300 Ω in the ideal free-space case. Unequal diameters, spacing, height and nearby structures change the transformation ratio.

The same MIT antenna material explains the coupled-conductor relationship. Total copper length is plainly not enough to classify the antenna.

Long Wires, V-Beams and Rhombics

Antennas several wavelengths long deliberately use phase progression along their conductors to create directional lobes. Standing-wave long wires and unterminated V-beams are frequency sensitive; terminated travelling-wave designs can provide broader impedance and pattern behaviour but dissipate some accepted power in the termination.

That resistor loss is intentional and must be included in efficiency. The directional pattern may still provide useful realised gain in the wanted direction, but “low SWR” does not make termination power radiated power.

The U.S. Army HF antenna manual documents field-expedient long-wire, V and half-rhombic systems. Their successful use makes the broader point: passing half a wavelength does not make a conductor defective.

Matching Does Not Erase Feedline Loss

Any of these antennas can be efficient and still present an inconvenient port impedance. Matching and loss must be separated:

1. Antenna impedance is the complex terminal ratio at the defined feed plane.
2. Mismatch describes power not accepted at that plane for the connected source or line.
3. Feedline attenuation dissipates power as waves travel in both directions on a mismatched lossy line.
4. Matching-network loss dissipates power in conductors, capacitors, cores and contacts.
5. Radiation efficiency compares radiated power with accepted power after antenna-system losses at the chosen boundary.

Place the impedance transformation where it yields the lowest safe total loss. That may mean a feedpoint network, an intentional transmission-line transformer, or low-loss balanced line to a remote tuner. It does not mean every high impedance is harmless: voltage rating, current, insulation, arcing, common mode and bandwidth still matter.

How to Compare Full-Wave and Half-Wave Systems

1. Name the topology. “Full wave” is incomplete; specify straight dipole, loop, folded conductor or travelling-wave wire.
2. Fix the reference plane. Compare equal accepted power at the antenna, or account for every upstream loss.
3. Use the full pattern. A single broadside report cannot reveal side lobes or nulls.
4. Match polarisation. Cross-polarisation can overwhelm a modest directivity difference.
5. Control height and ground. Equal wire height, equal feedpoint height and equal current-centroid height are different tests.
6. Measure feedline current. Common mode can turn the coax into an uncontrolled extra element.
7. Include matching and line loss. A low SWR at the transmitter is not an efficiency measurement.
8. Report uncertainty. Propagation, station AGC and one receiving direction do not constitute an antenna range.

For numerical models, include the feedline exterior, balun or choke impedance, real ground and nearby structures. Check segmentation and compare the model with measured feed impedance and current before trusting tenths of a decibel.

Design Decisions at a Glance

Choosing a Full-Wave System

A centre-fed, exactly one-wavelength straight dipole is normally a poor direct load for 50 Ω coax because the centre is near a current node. That is a valuable and specific conclusion.

Move the feed toward a current maximum, close the wire into a loop, couple a second conductor into a folded dipole, or use several wavelengths as a directional travelling-wave structure, and the impedance and pattern change. None of those changes repeals conservation of energy; they change current distribution and the way power is coupled into space.

Judge the complete antenna by installed realised gain in the required directions, feed and matching loss, voltage and current stress, common-mode behaviour, mechanical safety and suitability for the path. The wavelength is not good or bad. The geometry, feedpoint and installation decide.

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