End-Fed Antennas Explained: My Practical Multiband Choice
End-Fed Antennas Explained: My Practical Multiband Choice
One convenient feed position, three different antenna families. For most multiband end-fed installations I start with EF-OCF—but the reason is its manageable matching problem and deliberate current paths, not a promise that one family always wins.
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 convenient part of an end-fed wire is obvious: the feedpoint can sit near a building, tree, mast or operating position while the far end is supported elsewhere. The difficult part is less visible. The source still needs a complete RF circuit, and part of that circuit may be the coax exterior, a counterpoise, a mast, nearby structures or distributed capacitance to the environment.
My practical default: for a garden or portable station wanting several HF bands, I usually start with an end-fed off-centre-fed arrangement when the site can accommodate both intended branches. It lets me aim for moderate transformation, keep the return path explicit and accept sensible tuner matching where needed. That is a useful starting design—not a fixed ratio or an efficiency guarantee.
“End-Fed” Is a Feed Description
A centre-fed dipole makes its two intentional arms easy to see. At an end-fed feedpoint, the long wire is obvious but the second current path is often omitted from the drawing. It has not vanished. The return may be a deliberate wire, a declared section of coax exterior, a metal support, enclosure capacitance, ground coupling, station wiring or several paths in parallel.
Roy Lewallen, W7EL, separates the wanted differential current inside coax from imbalance current on the outside of the shield in Baluns: What They Do and How They Do It. That distinction matters here: an intentional length of coax exterior can be part of the radiator, while current that continues past the intended boundary can alter the pattern, carry RF into the station or collect local noise.
The label therefore begins the investigation; it does not finish it. Two installations using the same length of wire and the same nominal transformer can behave differently when height, slope, soil, nearby conductors, coax route or choke position changes.
Three Useful End-Fed Families
| Family | What defines it | Main engineering question |
|---|---|---|
| End-fed half-wave (EFHW) | The wire is operated near a half-wave mode on a chosen band and may support useful higher modes on other bands. | What complex impedance appears at the actual feedpoint, and what return path completes the source circuit? |
| End-fed off-centre-fed (EF-OCF) | The feed system deliberately creates unequal radiating branches, often including a defined exterior-conductor section. | Where is the intended branch boundary, and does current remain controlled beyond it on every band? |
| Non-resonant end-fed wire | The wire length is chosen for installation and tunability rather than one exact resonant mode. | Can the tuner and matching network transform the full R + jX range with acceptable component, feedline and ground-system loss? |
These categories overlap in casual language, which is why “random wire” is a poor electrical specification. A wire is not random to the current flowing on it. Its length, routing and return structure set repeatable impedances and current modes, even when they were not deliberately chosen.
The EFHW: Convenient, but Not a One-Number Load
An EFHW is normally fed near a current minimum and voltage maximum of a half-wave mode. Its feedpoint impedance is therefore high compared with a centre-fed half-wave dipole, but it is not a universal resistance. End effects, conductor diameter, height, slope, coupling, transformer leads and the return structure all contribute resistance and reactance at the terminal pair.
Higher-frequency operation can use other wire modes, but harmonic arithmetic alone does not certify a useful multiband antenna. Every band has its own feedpoint load, wire-current distribution, exterior-current path and far-field pattern. A low SWR at the transmitter only reports the impedance seen at that reference plane; it does not prove transformer efficiency, feedline efficiency, common-mode control or a useful radiation pattern.
The matching network must be designed around measured complex loads, not a memorised ratio. The transformer’s magnetising behaviour, leakage coupling, winding capacitance, conductor loss, core loss, voltage stress and thermal rise change with frequency and load. Compensation can improve a measured part of the response, but it is not magic and must not be used to hide dissipation behind a flatter SWR curve.
I choose an EFHW when a particular band—or a deliberately designed harmonic pair—is the main job and the available support suits its radiator and return path. Then the high-impedance feed is a specific problem worth solving. I am much less enthusiastic about asking one compact transformer to serve every higher-band mode and treating several SWR dips as proof that the whole arrangement works equally well.
The EF-OCF: My Multiband Starting Point
An end-fed off-centre-fed arrangement can present a more moderate load than feeding extremely close to the voltage maximum. The important feature is not a fixed tap percentage; it is the complete two-branch current geometry. One branch may be the long wire and the other a deliberately bounded counterpoise or coax-exterior section.
This is the reason for my preference. If the intended bands present loads that suit a moderate transformation, I do not have to solve the same high-voltage, high-ratio matching problem as at the end of a half-wave mode. In an appropriate topology, the smaller required voltage ratio can also simplify the high-side winding and its parasitic effects. Mini-Circuits' transformer guidance explains the underlying winding and magnetic constraints. The design still needs adequate magnetising inductance, core area and current capacity; a lower ratio alone does not establish lower core loss.
For a fixed multiband station, that gives me a useful compromise between installation convenience and matching demands. For portable work, a repeatable long-branch/short-branch layout gives me something deliberate to deploy rather than letting the remaining coax choose the antenna. I would rather do a little tuner matching on a difficult band than insist that every band must look perfect through one heavily stretched matching network.
This is where transformation and choking must be treated as separate jobs. An UNUN can transform an intentionally unbalanced, measured load. A separately characterised 1:1 choke can define where current on the exterior branch should stop. Putting the choke directly beside the transformer is appropriate only when another intended return already completes the antenna there; otherwise it changes the circuit being matched.
A suitable current balun can also be a valid choice when its load and common-mode requirements are within its capabilities. My preference for separate transformation and choking keeps the two jobs explicit; it does not mean a current balun stops working merely because a practical antenna is asymmetric. Supports, routing and nearby structures still determine the installed current paths.
The site condition matters: both intended branches must fit with suitable clearance, and their routing must remain reasonably stable. If the only available route pushes the short branch against metalwork or makes the station wiring its substitute, the EF-OCF default no longer solves the problem. “Practical” does not mean “install it anywhere and forget the other conductor”.
The Non-Resonant Wire: Flexibility Has a Price
A non-resonant end-fed wire can be routed as a sloper, inverted-L or irregular wire and matched over several bands. That flexibility can be valuable where one resonant geometry will not fit. It also means the matching system may encounter very low resistance, very high resistance or large reactance on different bands.
A tuner can transform impedance; it cannot guarantee that accepted power reaches the wanted wire mode. Loss can accumulate in the matching network, feedline, ferrite, ground or return structure. Avoid wire lengths that place an extreme load outside the tuner’s characterised range, and confirm the completed system under power rather than relying only on a low-power analyser trace.
I use this alternative when routing freedom matters more than preserving a defined resonant or off-centre geometry—for example, a temporary site where the available support dictates the wire. With a suitable tuner and deliberate return it can be useful and efficient on selected bands; “non-resonant” is not another word for “lossy”. The trade-off is that the tuning range and loss budget deserve more attention as the load changes.
Low SWR Is a Starting Point, Not an Efficiency Meter
A broad low-SWR response can result from a genuinely well-behaved system, but loss can also make impedance look less demanding. The measurement is incomplete until the reference plane, calibration, cable length and matching state are declared. Keysight’s VNA calibration guidance explains why the calibrated reference plane belongs at the terminals whose impedance you intend to report.
For a useful comparison, record complex impedance before and after the matching network, estimate or measure feedline loss, inspect core and conductor temperature after a representative transmission, and map current on every intentional and unintended branch. De-key and isolate the transmitter before handling hardware or moving probes; do not use touch to diagnose an energised antenna. Compare transmitted results at equal accepted power. Field strength or receiver SNR can then answer whether one installation delivers a better wanted result under the same path and measurement conditions.
Installation Geometry Changes the Answer
Height is not one variable. Record the terminal height, average wire height, height of the current-rich regions, distance to conductors and effective electromagnetic ground. An inverted-L, sloper and flat-top made from the same wire can produce different terminal impedance and pattern because the current-bearing sections occupy different positions and orientations.
The Numerical Electromagnetics Code can model wires, ground, networks, transmission lines, currents and patterns. The model must include every conductor that materially carries current—including a coax-exterior branch when it is part of the radiator. A model that stops at the transformer cannot predict radiation from a missing return path.
Likewise, a free-space plot is not an installed-site guarantee. ITU-R BS.705-2 distinguishes theoretical antenna patterns from performance affected by practical surroundings. Treat take-off angle and lobe claims as geometry-specific until the full installation has been modelled or measured.
Choose the Job, Then Check the Installation
- Several HF bands from a repeatable garden or portable layout: my starting choice is EF-OCF, provided the intended short branch fits and the band-by-band load supports moderate matching. The attraction is the manageable current system, not a no-tuner promise.
- One priority band or a planned harmonic pair with a suitable support: a purpose-designed EFHW can be the more focused answer. Give its high-impedance feed, voltage clearance and return conductor the attention they need.
- An awkward temporary route that will not accommodate either defined layout: a non-resonant wire with a suitable tuner and deliberate return may make the best use of the space. Choose the length and matching location to avoid loads the network cannot handle well.
Having chosen the architecture, draw the whole return, measure R + jX at a fixed reference plane and choose the matching network from that load. Verify its loss and operating limits, place the choke at the intended boundary and check current on both sides. For important comparisons, use controlled A/B/A or A/B/B/A observations. These checks put the chosen design to work; they do not replace the reason for choosing it.
Practical Conclusion
I like end-fed antennas because they solve real installation problems. I do not like treating “end-fed” as proof of a particular impedance, transformer ratio, efficiency or pattern. The useful antenna is the entire installed current path: wire, return, matching network, choke, feedline and surroundings.
For most multiband end-fed setups with room for the complete current system, EF-OCF is still my preferred starting point: a more manageable transformation and an explicit branch/choke boundary are useful advantages. I choose the EFHW when a focused band plan makes its high-impedance feed worthwhile, and the non-resonant wire when the site demands maximum routing freedom. The default is clear; the installation decides whether it fits.
Mini-FAQ
- Does an end-fed antenna work without a counterpoise? It always needs a return path. That path may be a deliberate wire, a defined coax-exterior section, a mast, distributed capacitance or several conductors in parallel.
- Is one transformer ratio correct for every EFHW? No. The correct network depends on the measured complex load, frequency range, return geometry, power, voltage stress and acceptable loss.
- Does low SWR prove that an end-fed antenna is efficient? No. SWR describes the impedance match at one reference plane; it does not separate radiation from transformer, feedline, conductor or ground-system loss.
- Should the choke sit directly beside the transformer? Only when the intended antenna return is complete there. If a declared coax-exterior section is an intentional branch, the choke belongs at that branch’s verified boundary.
- Why do you usually start with EF-OCF for multiband use? When both intended branches fit, its off-centre current system can make moderate transformation and deliberate return-path control practical across several bands. That is my default design preference, not a guarantee of more gain or less loss at every site.
- Can a tuner make any wire a good multiband antenna? A tuner can transform impedance within its range, but it cannot guarantee low loss, controlled common-mode current or a useful pattern.