The End-Fed Half-Wave Myth: Why Most EFHWs Are Doing It Wrong
The End-Fed Half-Wave Myth: Why Most EFHWs Are Doing It Wrong
An EFHW can be excellent when it is engineered for declared bands and limits. Most problems begin when a half-wave wire, a 49:1 label and a shunt capacitor are sold as an 80–10 or 40–10 m solution without proving the transformer bandwidth, return path, loss, thermal margin and pattern.
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.
My challenge is deliberately sharp: the common 80–10 m package is not one broadband antenna. It is a set of separated harmonic-region opportunities forced through one transformer, one compensation network and one installed return system across more than three octaves. A 40–10 m version narrows the problem, but still asks one assembly to span more than two octaves.
Mark, the Ham Florida Man, takes up this argument in The MYTH of the All-Band EFHW Antenna. He identifies me, Joeri Van Dooren, ON6URE, and closely reads my companion 80–10 m EFHW article. The central challenge is the same: convenience, contacts and an SWR dip do not prove an efficient all-band system. Here I explain the design choices behind that criticism. Material limits, compensation and the completed antenna have to be distinguished; the critique is not a universal declaration that ferrite or capacitors cannot be engineered usefully.
The point is the compromise: an EFHW is not wrong merely because it is multiband. It is doing it wrong when separated SWR dips are presented as broadband proof, or when one box is assumed to have the same low loss, safe temperature and useful pattern on every labelled band. Engineering the compromise means declaring it, measuring it and rating the completed system.
The Feedpoint Impedance Is Not a Universal Constant
The end of a half-wave-like conductor often presents a high impedance, but the value is complex and installation-dependent. Height, orientation, wire diameter, nearby objects, ground, loading, transformer capacitance and the chosen return path all shift it.
A nominal 49:1 transformer maps 2,450 Ω to 50 Ω only in the ideal real-resistance arithmetic. The installed load may not be 2,450 Ω and may contain substantial reactance. Other ratios can fit other load regions, but no ratio becomes correct because it appears in a product name.
The Return Conductor Never Vanished
Current needs a complete path. At an end feed, the complementary current can flow through an explicit counterpoise, transformer capacitance, the coax exterior, a mast, station wiring and coupling to the environment. Those paths determine terminal impedance and can radiate.
A choke defines a boundary by adding impedance to one continuation path. Put it close to the transformer and the available return structure is short; put it farther down the coax and part of the coax exterior becomes intentional radiator/return structure. Neither position is universally correct. State the intended boundary and verify exterior current along the cable.
Why 80–10 and 40–10 Transformers Are Multi-Octave Compromises
A low-band design needs enough magnetising impedance under its expected voltage. An upper-HF design must control leakage inductance, winding capacitance, transmission-line error and AC conductor/core loss. More turns and higher permeability can help the low end while making the high end harder; fewer turns and a higher-frequency material can improve the upper end while leaving too little magnetising impedance below.
Common Fair-Rite #43 and #52 choices illustrate the trade. Their published permeability and frequency behaviour differ substantially, so the limiting edge depends on the selected core dimensions, turns, winding geometry, parasitics, complex load, power, duty cycle and temperature. There is no generic #43 or #52 recipe that makes an 80–10 m EFHW broadband, and even 40–10 m can be borderline at its low or high edge depending on the complete assembly.
This material-selection problem is one reason RF.Guru uses selected, traceable Würth ferrite materials in its 4:1 and 9:1 matching transformers rather than treating a generic mix number as the design. The material name alone is still not the rating: the finished winding and transformer must be qualified for their declared application. The exact material and proprietary construction remain part of the qualified RF.Guru assembly.
Mini-Circuits’ RF transformer note identifies magnetising inductance at the low end and capacitance, leakage, core and conductor loss at the high end. Fair-Rite’s material pages likewise provide frequency-dependent material data—not a finished amateur-transformer rating.
A Shunt Capacitor Is Engineering—and Never a Free Lunch
A capacitor connected across two declared nodes is a shunt branch. It adds frequency-dependent susceptance and branch current, so it can compensate a known reactive trend and improve a selected match region. It does not manufacture broadband transformation. The same branch can overcompensate another band, move a pole or zero, raise circulating current or voltage, add real-component loss, or create a flattering 50 Ω bench result that does not survive the antenna’s complex load.
A shunt branch is not the same thing as a complete, deliberately solved LC matching network. Measure the complete transformer-and-antenna network with and without compensation across the relevant complex-load matrix. Record transmission or calorimetric loss, input impedance, voltage/current stress and equilibrium temperature. If only SWR improves, the efficiency and bandwidth questions remain open.
Multiple Half-Waves Produce Real, Complicated Patterns
On harmonic bands, a long radiator supports several current maxima and a multi-lobed pattern. That can put useful energy at low angles in some azimuths and deep nulls in others. The result depends on bends, slope, height, ground and return conductors. “More lobes” is neither automatically better nor automatically useless.
NEC can calculate this when the actual wire, ground, feed, loads, networks and transmission lines are modelled. The LLNL NEC record makes those inputs explicit. A wire-only model cannot prove the behaviour of an installation whose coax exterior carries current.
Why I Prefer a Smaller Job for an EFHW
For an end-fed installation centred on one band, I prefer a transformer, radiator and return system designed around that band. A deliberately selected half-wave/full-wave pair is another useful option. The engineering advantage is a smaller frequency and load envelope: the low-frequency magnetising requirement and upper-frequency parasitics no longer have to satisfy an 80–10 m brief at once. That gives the designer more freedom to optimise the wanted bands and demonstrate the result.
Those are design advantages, not automatic efficiency ratings. A full-wave radiator still has a different pattern from a half-wave radiator, and a narrow-band transformer can still be badly built. But choosing fewer bands removes part of the conflict before trying to cure it with winding changes and shunt compensation. That is why I favour purpose-built monoband and bounded dual-band EFHWs—not because a harmonic current distribution is somehow forbidden.
If Broad Multiband Use Is the Goal, Change the Feed Problem
The alternative is not necessarily another high-ratio EFHW box. An off-centre or other deliberately selected feed arrangement can target a lower impedance region and use a more moderate transformation, with a separate choke controlling the unintended current path. For the same accepted power into a resistive load, lower feedpoint resistance means lower terminal voltage and higher current. It changes the design burden instead of merely hiding it behind a better SWR.
An ideal 4:1 impedance transformation corresponds to a 2:1 voltage/turns ratio; 49:1 corresponds to 7:1. The moderate-ratio design asks for less voltage transformation, but the wire and return system must actually present a suitable load. Fewer turns alone do not guarantee lower core loss: voltage per turn, magnetising impedance, material, winding geometry and temperature still matter. A 4:1 box is not a drop-in cure for a kilohm-class end feed.
If a tuner is needed on some bands, that is a declared part of the design—not a defeat. The useful choice is between complete systems with known compromises, including feed-line and tuner loss. My preference is to reduce the transformation problem where the installation allows it, or split the bands into appropriate EFHW designs, rather than call one high-ratio assembly a no-compromise all-band antenna.
RF.Guru’s Practical Design Position
- Prefer a monoband or narrow, harmonically related application when predictable transformer and pattern behaviour matters most.
- Treat a broad multiband EFHW as a measured system compromise, not a no-tuner efficiency guarantee.
- Select transformer ratio from the installed complex load region rather than a universal end-impedance number.
- Define the return path and choke boundary before interpreting SWR or pattern.
- Qualify the exact ferrite-and-winding assembly at representative load, power, waveform, duty cycle and temperature.
- Use compensation when measurements show that it improves the required transfer without unacceptable loss or stress.
A Minimum Evidence Set
| Claim | Minimum useful evidence |
|---|---|
| “Covers the band” | Complex impedance across the whole band at a declared plane |
| “Low-loss transformer” | Relevant-load insertion loss or calorimetry with fixture removal |
| “Handles this power” | Thermal equilibrium, voltage/current and insulation margin at stated duty cycle |
| “Coax is not radiating” | Exterior-current scan over position and frequency |
| “Good DX pattern” | Complete model or controlled pattern/field comparison with ground and feed path included |
The Point of the Provocation
The myth is not that an end-fed wire can radiate or make contacts. The myth is that a wire, a 49:1 label and several SWR dips amount to proof of equally good 80–10 m operation. A match is one result. Transformer heating, return current and the directions in which the antenna sends power are other results, and a contact does not separate them.
I prefer the engineering problem to be smaller and explicit: a monoband or deliberately bounded dual-band EFHW where that fits the station, or a multiband system whose feed arrangement avoids the unnecessary high-impedance transformation burden. A shunt capacitor can help optimise part of a response; it cannot make the trade-offs disappear.
An all-band EFHW can offer useful operating windows, but its convenience has conditions. State those conditions and it is an engineered compromise. Sell the SWR curve as proof that there are none, and the criticism in this title stands.
Engineering References
- Mini-Circuits: RF transformer performance and measurement
- Fair-Rite material 31 data
- Fair-Rite material 43 data
- Fair-Rite material 52 data
- ARRL: installed common-mode current and choke measurement
- Lawrence Livermore National Laboratory: NEC-5
Mini-FAQ
- Is every EFHW badly designed? No. A bounded, measured EFHW can be excellent. The criticism targets conclusions that rely only on a wire length, ratio label and SWR dip.
- Is 49:1 always the correct ratio? No. It is a starting ratio for one assumed load region; the installed complex feedpoint impedance decides the required transformation.
- Can a single transformer cover 80–10 or 40–10 m? It can produce useful windows, but 80–10 spans more than three octaves and 40–10 more than two. That is a measured multi-octave compromise, not proof of continuous broadband, low-loss operation.
- Are compensation capacitors wrong? No. A shunt capacitor can be a valid compensation branch, but it trades frequency response and stress; it creates no free bandwidth and must be validated with the completed network.
- Where is the second conductor? It may be an explicit counterpoise, coax exterior, transformer capacitance, mast, station wiring or environmental coupling—usually a combination.
- Does low SWR prove good radiation? No. It describes match at one plane; loss, current distribution and pattern require separate evidence.
- Why prefer fewer EFHW bands or a moderate-ratio feed design? A monoband or bounded dual-band EFHW reduces the frequency and load envelope the transformer must handle. A feed arrangement presenting lower resistance can reduce the required voltage transformation. Neither advantage removes the need to control return current, component loss and the installed pattern.