9:1 Long-Wire vs EFHW/OCF: Compare the Complete System
9:1 Long-Wire vs EFHW/OCF: Compare the Complete System
A non-resonant end-fed wire through a 9:1 UNUN and tuner can make contacts across many HF bands. That success is real. What it does not prove is radiation efficiency, a controlled return path or a useful pattern. Compare the whole installed system before ranking it against an EFHW, OCF dipole or EF-OCF.
I have no problem with the 9:1 long-wire as a practical antenna. It is simple, tolerant of improvised geometry and often lets one conductor reach bands that would otherwise need several antennas. My objection starts when “the tuner found a match” becomes “the antenna is efficient,” or when a topology name is used to rank two installations that do not share the same wire, return path, feedline, losses or pattern.
Joeri's practical rule: a 9:1 wire is a general-purpose matching problem, not a guaranteed loss problem. Measure the complex load, transformer, tuner, feedline and return current at declared planes. Then compare accepted power and field performance with the alternative. Resonance, SWR and efficiency remain different questions.
Why a 9:1 Long-Wire Can Work
The wire is not electrically “random.” Its length, height, bends, slope, nearby conductors, ground and return branch form a repeatable impedance and current distribution at each frequency. The word means that the conductor was not cut to one simple resonant mode and is expected to need a matching network.
For an ideal transformer with an impedance ratio of 9:1, the load impedance is referred toward the 50 Ω side by roughly that ratio. That can move a moderate load into the range of a tuner or transmitter. A real broadband UNUN has magnetizing impedance, leakage inductance, capacitance, copper resistance, core loss and a finite voltage/current limit, so the transformation and loss change with frequency and complex load.
The tuner completes the job by cancelling reactance and transforming the remaining resistance at its own port. The radio may then see a low SWR. None of this requires the wire itself to be resonant, and none of it proves how much accepted power reached radiation.
Match, Accepted Power and Radiation Are Different Results
A low SWR at the transmitter says that the impedance at that reference plane is close to the transmitter's target. It does not identify where power was dissipated or which conductors radiated.
Use a complete power budget:
- Transmitter-to-tuner path: connectors, jumpers and any protection or switching devices.
- Tuner: inductor and capacitor loss under the actual resistance, reactance, current and voltage.
- Feedline: matched attenuation plus the additional dissipation caused by the actual standing-wave distribution.
- UNUN: copper, core, dielectric and stray-field loss under the installed complex load.
- Return network: counterpoise conductor, feedline exterior, capacitance, soil and nearby metal.
- Radiating structure: conductor and junction loss plus the power actually radiated into the installed pattern.
Efficiency at one chosen boundary is the radiated power divided by accepted power at that boundary. Move the boundary and the number changes because different losses are included. “Several dB down” is not a property of the words 9:1 or long-wire; it is a measured result for one complete installation and comparison.
The UNUN Does Not Have One Loss Number
A 9:1 label gives a nominal transformation ratio, not a bandwidth, efficiency or power rating. The same unit can be lightly stressed on one band and driven into a difficult voltage, current, flux or temperature region on another.
Judge it with the actual complex loads expected at the antenna port. Measure insertion loss or efficiency in a fixture that reproduces those loads and keeps the reference planes clear. A back-to-back test can be useful for production comparison, but it does not automatically divide into the installed loss of one transformer under a different load.
High resistance can raise winding voltage; low resistance can raise current; a large reactive component can raise circulating energy. Frequency changes magnetizing impedance and parasitic behaviour. Ferrite mix, core volume, winding geometry, conductor, duty cycle, enclosure and cooling all matter. There is no safe universal instruction to “derate the difficult bands” without the temperature, voltage and current limits of the actual assembly.
The Tuner May Be Efficient—or May Become the Main Loss
An efficient tuner with suitable component Q can transform a demanding load with modest loss. A compact internal tuner working at its range limit may dissipate much more. The load resistance and reactance decide the circulating current, voltage and effective network Q; the fact that the tuner completed a cycle does not reveal that loss.
Measure tuner input and output power at the actual load, frequency and power, with uncertainty stated. At high duty cycle, repeat after thermal equilibrium. A cool enclosure is not enough evidence when a coil, switch contact or capacitor carries the local stress.
High SWR on Real Coax Raises Dissipation
Mismatch does not consume power by itself. Real feedline has conductor and dielectric loss, and the standing-wave voltage and current distribution changes how much power that loss removes. The penalty depends on line type, length, frequency, load and the matched-line attenuation.
If the tuner is in the shack, the coax between tuner and UNUN may remain highly mismatched even though the radio sees 1:1. A tuner at or near the antenna port can leave the long station feedline close to its design impedance, but it does not make the tuner, transformer or return network lossless. Compare both layouts at the same accepted-power plane.
An End-Fed Wire Always Has a Return Path
Current leaving the UNUN output must return. Depending on the installation, the return may include a dedicated counterpoise, elevated conductors, capacitance to the surroundings, the outside of the coax, soil and station wiring. If the drawing omits that branch, the installation does not.
Do not call every current on the coax exterior accidental. A declared exterior section can be an intentional branch between the transformer and a choke. Beyond the intended branch, the choke should present enough common-mode impedance for the measured band and current. Its position follows that boundary—not a universal instruction to place it at the box or shack door.
A ground rod is part of the station's bonding and lightning/safety system, not a magic RF sink. Do not create an isolated electrode or defeat required bonding to improve a tuning result. When soil current is deliberately used, its loss and safety consequences belong in the complete design.
Published Length Lists Are Starting Points, Not Proof
Lists of “good random-wire lengths” try to avoid conductor lengths that place an extreme impedance at the transformer on several popular bands. They can be useful search seeds. They are not portable electrical constants.
End effect, height, folding, return-branch length, feedline exterior, soil, nearby metal and tuner reference plane all shift the impedance. Adding or removing a metre can improve one band and make another harder. Select the final geometry from measured complex impedance and current behaviour across the intended bands, not from the neatness of a table.
Resonant and Near-Resonant Antennas Do Not Automatically Win
An EFHW, OCF dipole or EF-OCF may reduce the matching range required on selected bands, which can simplify the loss and stress problem. But resonance does not guarantee low conductor, ground, transformer or common-mode loss. A lossy network can produce an excellent resonant match, and an efficient non-resonant wire with a low-loss remote tuner can outperform it.
Nor is “the same physical space” a sufficient control. Different topologies can place current maxima in different parts of the aperture and create different azimuth and elevation patterns. One may put more field toward a chosen path even with lower total efficiency; another may radiate more total power in unhelpful directions.
| Question | What must be held or measured |
|---|---|
| Which system loses less power? | Same input boundary and accepted power; transformer, tuner, feedline, return-network and conductor loss. |
| Which is more efficient? | Radiated power divided by accepted power at the same declared plane, with uncertainty. |
| Which is stronger on a path? | Same accepted power, time/propagation control and installed azimuth/elevation pattern. |
| Which is easier to use? | Band coverage, tuner range, voltage/current margin, maintenance, supports and common-mode behaviour. |
Build the 9:1 System Deliberately
- Record the whole geometry: main wire, intended return branch, feedline route, choke, ground/bonding and nearby conductors.
- Sweep the complex load: measure resistance and reactance at the UNUN port across every operating band.
- Choose and test the transformer: reproduce the load region, measure loss and temperature, and verify voltage/current margin.
- Put the tuner where it solves the right problem: account for feedline loss on both sides of the selected reference plane.
- Map exterior current: verify the intentional branch and confirm acceptable current beyond the choke.
- Compare restored baselines: hold accepted power and receiver conditions constant, then repeat A/B/B/A field or on-air measurements when propagation varies.
Bottom line: a 9:1 long-wire works because a real conductor, real return path, transformer and tuner form an antenna system. It may be a superb practical choice. EFHW, OCF and EF-OCF alternatives often reduce the matching burden on selected bands, but they do not win by name. Trace the currents, measure each loss at declared planes and compare the installed patterns.
Efficiency, tuner and current-path references
Mini-FAQ
- Is a 9:1 long-wire necessarily inefficient? No. Efficiency depends on the installed wire, return path, UNUN, tuner, feedline and pattern. The ratio alone cannot rank it.
- Does a low SWR at the radio prove good radiation? No. It proves a match at that reference plane. Power can still be lost in the tuner, line, transformer, ground or conductors.
- Does the antenna need a counterpoise? It needs a return path. That may be a deliberate conductor, feedline exterior, capacitance, soil or a combination; name and measure the intended path.
- Where should the common-mode choke go? At the intended end of any exterior-current branch, with adequate measured impedance on each band. There is no universal box-end or shack-entry rule.
- Are published random-wire lengths reliable? They are starting points. Final impedance depends on height, folding, return path, feedline, ground and nearby objects, so measure the installed geometry.
- Is a resonant EFHW or OCF always more efficient? No. Resonance and match do not prove efficiency. Compare complete loss budgets and installed patterns at the same accepted-power plane.