Why Some EFHWs Eat Your Signal—and Why I Favour EFOC
Why Some EFHWs Eat Your Signal—and Why I Favour EFOC
A lower-ratio EFOC gives us a less demanding multiband matching problem. The advantage comes from the design—not from declaring every EFHW inefficient.
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.
A friendly SWR reading can hide an unfriendly power budget. That is why I favour the EFOC approach for many practical multiband installations: I would rather make the matching problem less demanding at the antenna than ask a compact high-ratio EFHW transformer to cope with everything from 80 m to 10 m.
This is not an argument against every EFHW. A well-designed mono- or dual-band EFHW can be an excellent antenna. The comparison that matters here is between a high-ratio assembly stretched beyond a sensible operating envelope and an off-centre antenna designed around a more manageable transformation. Their end-access convenience may look similar; the work inside the matching box is not.
The EFOC advantage I am aiming for: A suitable moderate-impedance feed, a lower required voltage ratio, a winding arrangement with less unnecessary complexity, and a return path chosen deliberately. Those are practical opportunities to reduce loss. They are not a claim that an EFOC has no loss—or that a 4:1 label proves a result.
The High Ratio Is a Design Burden, Not a Badge
A common EFHW matching unit transforms a kilohm-class feedpoint impedance toward a 50 Ω feed system. An ideal 49:1 impedance transformation needs a 7:1 voltage ratio; 64:1 needs 8:1. The ratio is legitimate when it suits the load. Trouble starts when the same compact assembly must also handle widely separated frequencies, changing load reactance, substantial voltage and sustained operation.
On the low-frequency side, the winding must provide adequate magnetising inductance. On the high-frequency side, winding resistance, leakage inductance and stray capacitance increasingly shape the response. Ferrite properties and heating enter the problem as well. Mini-Circuits' RF-transformer explanation shows why these are different constraints, not one number called “broadband”.
It is possible to make a good high-ratio network. But a low-power sweep into one resistance does not show what happens with a reactive antenna load, at a different temperature or during a long transmission. The useful design envelope includes the frequencies, loads, power and duty cycle together. That is a much more demanding promise than finding several SWR dips.
Why the EFOC Starts With an Easier Job
The RF.Guru EFOC29 and EFOC17 use a 4:1 UNUN and a specified return-path arrangement. Rather than depend on the very high impedance at the end of a half-wave mode, the off-centre current system provides a different load for the matching network. The aim is an impedance range that suits moderate transformation across the intended bands, with further tuner matching where needed.
A 4:1 impedance ratio corresponds ideally to only 2:1 in voltage. In comparable winding arrangements, that can require fewer high-side turns, less wire and less difficult control of the winding's unwanted inductance and capacitance. Reducing the required output voltage also eases insulation and electric-field demands. This is the positive engineering reason for my preference: avoid an unnecessarily extreme transformation where a different current system can do the job more simply.
The advantage belongs to the whole choice of antenna load and matching network. Replacing a 49:1 box with a 4:1 box on the same kilohm load is not the same design. The moderate-ratio approach works because the feed arrangement changes too.
Fewer Turns Must Not Mean Less Magnetic Margin
There is an important distinction between simplifying a winding and underwinding a core. For a given voltage waveform, flux change is set by voltage integrated over time per turn; core area then determines flux density. Lloyd Dixon's TI magnetics paper sets out that Faraday-law relationship. Holding voltage, frequency and core area fixed while removing turns can increase the flux swing rather than reduce it.
A lower ratio does not automatically lower core flux. If the same primary voltage, primary turns, core and frequency remain, changing the secondary ratio does not by itself reduce the flux they impose. What a lower required ratio can improve is the high-side winding and voltage problem. The designer still has to provide the magnetising inductance, conductor capacity and thermal margin needed at both ends of the operating range.
That is why “fewer turns” is a design opportunity, not a loss measurement. I favour the simpler transformation when the antenna permits it; I do not confuse simplicity with permission to ignore the magnetic circuit.
A Shunt Capacitor Can Improve the Match Without Settling the Loss Question
The compensation capacitor commonly connected across the low-impedance winding of an EFHW transformer is a shunt branch. Its impedance changes with frequency. It can usefully compensate part of the network's reactance, but it also changes branch currents and the operating conditions elsewhere in the circuit. It must tolerate those currents and voltages, and real components have loss.
A better match may allow more power into the system; it does not tell us how that power is divided between useful radiation and dissipation. Loss itself can also make a poor load look better from the transmitter end. The SWR meter cannot decide whether the matching improvement came with an acceptable efficiency trade-off.
This is why our EFHW design position favours intended monoband or bounded dual-band operation, rather than treating an 80–10 m claim as one continuous broadband design. A compensation branch can have a valid engineering purpose. It cannot turn the transformer, wire and installation into a different antenna just because the graph looks tidier. The shunt-capacitor article follows that trade-off in more detail.
The Feedline Can Preserve the Advantage—or Spend It
Transformer loss is only one part of the route from transmitter to antenna. The cable still has conductor and dielectric losses, and mismatch changes the voltage and current distribution along it. If the EFOC arrangement presents a more suitable load to a given coax run on the bands you use, avoiding that extra feedline burden is a real system advantage.
That “if” has a practical meaning: use the installed impedance and the actual cable, not a generic SWR number. A tuner at the shack can make the transmitter happy while the line between tuner and antenna remains mismatched. A better-positioned match can reduce that burden, but the matching network has its own losses and limits.
The return conductor matters too. In an EFOC coax-return installation, a deliberate section of coax exterior belongs to the antenna and its choke marks the boundary. In the two-wire arrangement, both wires supply the intended antenna current and the feedline choke belongs at the feedpoint. Follow the matching manual. The UNUN performs the impedance transformation; the separately specified choke controls the unwanted continuation of exterior current toward the station.
An EFHW also needs a return path. W8JI's end-fed analysis makes that point concrete. A high feedpoint impedance does not abolish the other conductor. The EFOC benefit is not that it escapes current continuity, but that its intended return branch is part of the design from the start.
What an Avoided Decibel Is Worth
Consider two hypothetical feed systems delivering power to the same antenna. Each accepts 100 W at its input; their total dissipative loss is 1.5 dB and 0.5 dB respectively. These are illustrative assumptions, not measurements of an EFHW, EFOC or RF.Guru product. The transmitted fraction is 10−L/10, where L is the loss in dB.
| Hypothetical feed system | Total loss | Power reaching the antenna | Power dissipated before the antenna |
|---|---|---|---|
| More demanding, higher-loss implementation | 1.5 dB | 70.8 W | 29.2 W |
| Lower-loss implementation | 0.5 dB | 89.1 W | 10.9 W |
Avoiding that one decibel puts about 18.3 W more into the antenna for the same accepted 100 W. With radiation efficiency and pattern unchanged, it is also a 1 dB improvement along the path. That is why reducing a matching burden is worth doing: the point is to deliver more of the available power, not to win a contest for the prettiest SWR curve.
The table does not establish which real system achieves either loss. A calibrated transformer measurement under defined loads, together with the actual feedline loss, can establish the relevant budget. Properly configured back-to-back measurements can be useful; treating half a fixture's attenuation as a universal installed-antenna result cannot. Surface temperature alone is not calibrated loss either, and never use touch to investigate an energised RF assembly.
More Watts Delivered Still Need the Right Pattern
It is also possible to improve the feed system and still point a null at the station you want. As a long wire becomes several wavelengths long, its lobes and nulls change. An EFOC does not automatically have simpler high-band lobes, and an EFHW does not automatically waste power because it has them. Pattern is about where the radiation goes; loss is about power that does not become radiation.
A layout that needs less awkward folding or routing may be easier to install well, which can favour a particular EFOC model at a particular site. That does not make its main-wire length the whole antenna: include the other branch and nearby conductors. For a real DX comparison, ask whether the lower-loss feed system also serves a useful installed pattern in the wanted direction.
Reduce the Burden Before Trying to Hide It
For broad multiband use, I favour the EFOC route when it lets the antenna present a sensible load to a moderate-ratio UNUN, keeps the intended return path explicit and avoids unnecessary matching loss. Lower transformation, manageable winding demands and controlled feedline participation are a coherent design argument—not three unrelated SWR tricks.
A band-appropriate EFHW remains a good solution when its high-ratio feed is designed for the actual job. I do not need to pretend otherwise to explain the EFOC advantage. I would simply rather remove an avoidable source of difficulty than compensate around it and hope the missing watts went somewhere useful.
Do not judge an antenna by SWR alone. Reduce unnecessary loss, control common-mode current, and choose the right matching system for the job.
Further Engineering Reading
- Mini-Circuits: How RF Transformers Work and How They Are Measured—winding, magnetic and measurement limits.
- Mini-Circuits: Impedance Matching Devices—the relationship between turns, voltage and impedance ratios.
- Lloyd H. Dixon, TI: Magnetic Field Evaluation in Transformers and Inductors—Faraday's law and voltage per turn, not an HF antenna-product loss rating.
- W8JI: End-Fed Half-Wave Antennas—the complete return-current system and the limits of a match-only inference.
Mini-FAQ
- Are all EFHW transformers inefficient? No. A properly designed high-ratio transformer can work efficiently within its intended frequency, load, power and duty-cycle range. My objection is to treating separated resonant bands as proof that one compact assembly handles every operating condition well.
- Why do you favour the EFOC approach for multiband operation? A suitable off-centre current system can use a moderate 4:1 transformation and a deliberate return path. That reduces the required voltage ratio and can ease winding and parasitic demands, giving the designer a more manageable multiband matching problem.
- Do fewer turns automatically mean lower core loss? No. Core flux depends on voltage per turn, frequency and core area. Removing turns while keeping the same voltage and frequency can increase flux. Fewer high-side turns may simplify a suitable design without proving its total loss.
- Can a shunt capacitor legitimately improve an EFHW match? Yes. A correctly chosen shunt branch can compensate part of a network's reactance. Its effect is frequency-dependent, and its voltage, current and loss limits remain. Better SWR alone does not establish better system efficiency.
- Does an EFOC need a tuner? Some model, band and installation combinations do. Use the specified return path and the actual installed impedance to determine tuner need; a tuner at the shack does not remove loss already occurring in the feed system.
- Does lower feed-system loss guarantee a stronger DX signal? Only with the other relevant factors held equal. More power can reach the antenna, but radiation efficiency and the installed pattern in the wanted direction still determine how much reaches that path.