FD4 and Windom Antennas: Match the Current Path, Not the Name
FD4 and Windom Antennas: Match the Current Path, Not the Name
An off-centre feedpoint can support several useful antenna architectures. The transformer, choke and feed-line route have to match the one actually being built—not a traditional label or a universal impedance recipe.
The FD4 remains useful when its size, bands and installed pattern suit the station. What deserves retiring is the assumption that “Windom,” a particular arm split and a transformer ratio completely specify the antenna. They do not tell us where every ampere returns.
Three Different Systems Hide Behind Similar Names
Loren Windom’s 1929 article described a horizontal wire fed off centre by a single-wire feeder. That feeder was part of the radiating system. A modern FD4 is normally a two-arm, coax-fed off-centre-fed dipole. A third architecture deliberately uses a counterpoise or a declared section of coax exterior as a return branch. These are not interchangeable drawings.
| Architecture | Intended current path | Boundary to control |
|---|---|---|
| Historical single-wire Windom | Horizontal wire, single-wire feeder and station-side return environment form one RF structure | Feeder routing and the station return are part of the antenna |
| Coax-fed FD4/OCF dipole | Wanted differential current flows into one arm and returns through the other | Prevent the coax exterior from becoming an accidental third radiator |
| Deliberately unbalanced offset-fed system | Main radiator plus a named counterpoise or coax-exterior segment | A choke terminates the intentional return branch before the ordinary feed line begins |
The distinction is practical. If the feed-line exterior is meant to be quiet, movement of the coax should not become an informal tuning control. If a defined exterior section is the return branch, choking it at the transformer would remove part of the intended antenna.
Off-Centre Does Not Automatically Mean Unbalanced
Moving a dipole feedpoint away from the centre changes the impedance sampled along its standing-current distribution. It does not repeal the differential-current relationship at a two-terminal feedpoint. In an ideal two-arm OCF dipole, the wanted current entering one arm has an equal return through the other arm even though the physical lengths differ.
Installed balance is a separate question. Unequal height, ground clearance, nearby trees or metalwork, transformer capacitance and a feed line routed close to one arm can couple the two terminals differently to their surroundings. That mode conversion can drive current on the coax exterior. The antenna may then include the feed line, mast and station wiring even though the sketch shows only two arms.
Joeri’s current-path test: do not infer electrical balance from symmetry, asymmetry or a label. Identify the intended pair of conductors, then measure whether a significant exterior return path remains.
A 4:1 Ratio Is Not a 200-Ohm Guarantee
The familiar arithmetic is limited: an ideal 4:1 impedance transformer maps 200 + j0 Ω to 50 + j0 Ω. An installed FD4 does not present 200 + j0 Ω on every desired band. Its feedpoint impedance is complex and varies with wire length, feed position, frequency, conductor diameter, height, ground, nearby objects and any current on the feed line.
A transformer scales the load within its operating limits; it does not remove reactance or create resonance. On a multiband antenna, the same physical feedpoint samples a different current distribution on each band. A ratio useful on one band can present an awkward impedance on another.
Choose the ratio from measured or credibly modelled R + jX at the declared reference plane across all required frequencies. Then include transformer loss, tuner range, feed-line attenuation, voltage and current stress. A low SWR at the transmitter is not by itself an efficiency, temperature or pattern measurement.
Transformation and Common-Mode Control Are Separate Jobs
A transformer transfers the wanted signal between impedance levels. A common-mode choke adds impedance to an exterior-current loop. Some assemblies perform both functions, but the measurements still need to establish both behaviours under the intended load.
- For a genuinely balanced coax-fed OCF dipole, a suitable measured current balun can transform impedance while presenting common-mode impedance at the balanced port.
- For a deliberately unbalanced transformed port or return branch, Joeri’s practical default is a 4:1 UNUN for transformation plus a separately specified 1:1 choke for common-mode control.
- For an installation that does not preserve textbook balance, separating the functions makes the transformer transfer and the choke boundary independently testable. It does not make a weak transformer or choke acceptable.
The 4:1 value in Joeri’s default is conditional on the measured load calling for that ratio. It is not a promise that every FD4, OCF or offset-fed wire wants a 4:1 transformation.
The Choke Position Defines the Return Branch
A choke is finite impedance inserted into a complete common-mode circuit. Its useful position depends on which conductor is meant to radiate, the impedance of the exterior-current path, coupling to ground and nearby structures, and the choke’s complex impedance at each operating frequency.
For a coax-fed FD4 intended as a two-arm dipole, common-mode control normally begins at the feedpoint boundary. For a design that intentionally uses a counterpoise or coax-exterior section, the choke belongs where that declared branch ends. There is no universal quarter-wave distance that works on every band and installation.
A second choke near the station can protect a different boundary, but it cannot erase radiation, pickup or loss on the coax section ahead of it. If changing the feed-line route or touching station-side cables changes the tuning substantially, that is evidence that the installed current boundary needs investigation.
Multiband Behaviour Includes More Than SWR
The FD4 earned its reputation because one long wire can offer usable impedances on several bands. “Usable” still depends on the transformer, feed line, tuner and exact operating frequencies. One fixed arm split does not guarantee every amateur band, and a successful match does not prove that the same current path is preserved everywhere.
As the wire becomes electrically longer, its far-field pattern develops more lobes and nulls. Suppressing feed-line current can make the installed pattern more repeatable, but it cannot turn a multi-wavelength wire into a frequency-independent radiator. An intentional return branch produces another three-dimensional pattern that must be evaluated as part of the antenna.
Receive noise is equally site-specific. Controlling one feed-line pickup route may help, yet local fields can couple directly to the radiator, mast, feed line and station wiring. No balun, UNUN or choke label guarantees a lower noise floor.
The FD4 Is a Choice, Not a Deprecated Law
A well-installed FD4 can be a sensible antenna where a roughly 40-metre wire fits, its band coverage matches the operator’s needs and its pattern serves the desired paths. A shorter offset-fed antenna with an intentional return branch can solve a different space or feedpoint problem. Neither architecture earns a universal ranking from length, SWR or transformer type.
Compare the complete systems at the same reference planes:
- measure complex feedpoint impedance across every required band;
- state whether the coax exterior is excluded or intentionally used;
- measure exterior current at repeatable points while changing feed-line routing;
- verify transformer and choke loss, temperature, voltage and current under representative power, waveform and duty cycle;
- include tuner and feed-line loss rather than comparing transmitter SWR alone; and
- compare received signal, noise and field response with controlled A/B/A tests when those outcomes matter.
Primary and Authoritative Sources
- Loren G. Windom, “Notes on Ethereal Adornments,” QST, September 1929—the original single-wire-fed architecture and its intended feeder-current distribution.
- Roy Lewallen, W7EL, “Baluns: What They Do and How They Do It”—coaxial current modes, imbalance and current-versus-voltage balun behaviour.
- Tom Rauch, W8JI, Windom and Off-Centre-Fed Antennas—historical and modern architectures, feedpoint impedance and common-mode boundaries.
- Tom Rauch, W8JI, Common-Mode Current—installed current paths and the coax exterior as a separate conductor.
- ITU-R Report SM.2158-3—differential/common-mode definitions and mode conversion caused by imbalance.
- C. L. Ruthroff, “Some Broad-Band Transformers”—primary transmission-line-transformer analysis and practical bandwidth limitations.
Joeri’s Bottom Line
The FD4 is not disqualified by its age, and another offset-fed wire is not automatically better because it is shorter. The real dividing line is whether the return path is accidental or designed.
For a deliberately unbalanced transformed port or return branch, I prefer to keep the jobs explicit: use a 4:1 UNUN when the measured load supports that ratio, then place a separately specified choke at the intended antenna boundary. For a genuinely balanced OCF dipole, a measured current balun may be exactly right. Follow the current, choose the architecture, and only then choose the box.
Mini-FAQ
- Is the original Windom the same as a modern FD4? No. The original used a single-wire feeder as part of the RF structure; a modern FD4 is normally a coax-fed two-arm OCF dipole.
- Does an off-centre feedpoint make a dipole inherently unbalanced? No. The wanted two-arm current can remain differential. The environment, transformer coupling and feed-line routing determine mode conversion.
- Does a 4:1 transformer guarantee a 50 Ω input? No. That occurs only for the corresponding load in the ideal model. Real R + jX changes with frequency, geometry and installation.
- Should an FD4 use a current balun or an UNUN plus choke? A measured current balun can suit a genuinely balanced OCF dipole. Joeri prefers a conditional 4:1 UNUN plus separate choke for a deliberately unbalanced transformed port or return branch.
- Where should the common-mode choke go? At the declared antenna boundary: usually the feedpoint when the coax exterior is unwanted, or the far end of an intentional return branch.
- Is the FD4 obsolete? No. It remains useful when its size, impedances and installed patterns suit the station. It should be compared as a complete measured system, not accepted or rejected by name.