EF-OCF: When the Coax Exterior Is the Counterpoise
EF-OCF: When the Coax Exterior Is the Counterpoise
In this one-end EF-OCF architecture, the outside of the coax shield between the transformer and a deliberately placed choke is the shorter radiating branch. The choke establishes the intended far boundary of that branch. That explicit current path is the real design advantage; it is not a promise that every EF-OCF must outperform every EFHW.
I do not describe the coax as a counterpoise to excuse accidental RF on a feed line. I describe it that way because, in this architecture, a controlled section of shield exterior is intentionally recruited as the second antenna branch. If that conductor is missing from the drawing, the antenna description is incomplete.
Joeri's design position: when the installed complex load calls for the ratio, a 4:1 UNUN performs the differential impedance transformation. A separately specified 1:1 choke controls the common-mode boundary. Between those two devices, the coax exterior may be an intentional branch; beyond the choke, exterior current should be measured rather than assumed away.
End-Fed Does Not Mean One-Conductor
Current cannot leave a source terminal without a return path. At an end-fed antenna feedpoint, the return may be a dedicated wire, a radial or counterpoise system, the exterior of the feed line, a mast, station wiring, displacement current through the surroundings, or a mixture of these paths.
An EFHW and an EF-OCF can both look mechanically simple because only one long wire is obvious. Neither becomes electrically complete by hiding the other current path. The useful comparison therefore includes every conductor carrying current—not just the visible radiator length.
In the EF-OCF arrangement discussed here, the return path is made deliberate. The main wire is one asymmetric branch. The outside surface of a chosen coax section is the other branch. Its far end is defined by a common-mode choke. That is why I call the coax exterior the counterpoise or short OCF branch.
Three Currents Can Exist on One Coaxial Cable
The wanted transmission-line mode consists of equal and opposite currents on the centre conductor and the inside surface of the shield. Those fields are largely confined within the coax. A separate current can flow on the outside surface of the shield. Roy Lewallen's balun experiments describe that exterior current as imbalance or common-mode current and show why it can radiate independently of the internal pair.
In an ordinary feed line, we usually try to suppress that exterior current. In this EF-OCF section, we deliberately use it as an antenna current up to a chosen boundary. Calling it intentional does not change the physics: the route, height, surrounding conductors, soil, mast and choke impedance all determine how much current flows and where it goes.
Keep the modes separate: the UNUN transforms the differential load seen between its antenna-side terminals. The choke adds impedance to the exterior-current path. A good differential SWR does not prove that the exterior-current boundary is working.
The Choke Defines a Boundary, Not a Perfect Wall
Moving the choke changes the geometry and electrical loading of the intentional exterior branch. It can therefore change feed impedance, current distribution and pattern. This is not a coax-length trick applied after the antenna; the choke position belongs to the antenna design itself.
A real choke has finite, complex and frequency-dependent common-mode impedance. It does not make the downstream coax perfectly quiet on every band. Current can pass through the finite impedance or be re-excited farther along the line by coupling to the radiator, mast or station wiring.
That is why a physical distance alone cannot certify the boundary. Measure exterior current before and after the choke, farther down the cable and at the station. Repeat on every operating band and after changing the cable route or nearby conductors.
The 4:1 Ratio Follows the Installed Load
Moving the source away from the extreme end of a wire can place selected current modes in a more moderate impedance region. When measurements support it, a nominal 4:1 impedance transformation can be a practical starting point. It does not guarantee a 50 Ω input across multiple bands.
The real transformer sees a frequency-dependent complex load. Magnetising impedance, leakage inductance, winding capacitance, conductor and core loss, load phase, frequency, power, duty cycle and temperature all affect loss and stress. A lower numerical ratio is an engineering opportunity, not proof of a cooler or more efficient assembly.
RF.Guru keeps transformation and common-mode control separate because the two jobs require different evidence. That default suits the normally unbalanced field installation and can also serve a genuinely balanced case only when the combined UNUN-plus-choke network is deliberately arranged and verified as a hybrid interface.
Why This Architecture Can Be a Better Starting Point
The EF-OCF can be preferable when it turns an accidental return path into a declared antenna branch. You can route that branch, choose its boundary, measure its current and keep the station side out of the radiating system. An EFHW can also be built with a deliberate return structure, but too many installations leave that job to whatever coax, mast and wiring happen to be connected.
An off-centre source may also present a less extreme load than an extreme-end feed on selected bands. If completed-network measurements confirm lower transformer loss or stress under the actual loads, that is a real advantage. It is not available from the labels 4:1 and 49:1 alone.
The final comparison is directional. The system that places more accepted power into the useful lobe can be better for a path even if its feedpoint SWR is less attractive. Conversely, a controlled return path can still produce the wrong elevation or azimuth pattern. Compare complete accepted-power patterns, not only transformer ratios or shack SWR.
The Coax Route Is Part of the Radiation Pattern
Once the coax exterior carries intentional current, it is a radiator. A vertical drop, a horizontal run, a sloping section and a cable laid near soil are different antenna geometries. They can show similar impedance while producing different current distributions and fields.
Internal coax velocity factor is not the cut-length rule for this exterior mode. The exterior current interacts with the jacket, soil, mast, supports and nearby conductors. Establish the useful branch length from the installed impedance and current distribution, then include the complete conductor geometry in the model.
Burial or enclosure is not automatically forbidden, but it changes dielectric loading, loss and coupling. If the intended branch is buried, run through conduit or placed against metal, that condition must be designed and measured rather than treated as invisible feedline routing.
A Separate Wire Is an Alternative Branch
A dedicated counterpoise wire can replace the intentional coax-exterior branch. In that arrangement the choke may belong near the feedpoint so the downstream coax is excluded as much as practical. The wire still needs a defined route, current map and pattern model.
A separate conductor does not automatically increase efficiency, clean the azimuth pattern or remove environmental interaction. It changes the geometry. Whether the change helps depends on conductor loss, ground coupling, current division, choke performance and the intended radiation direction.
Compare EF-OCF and EFHW at the Same Boundary
| Evidence | What to record | What it can establish |
|---|---|---|
| Complete geometry | Main wire, return branch, transformer, choke, coax route, mast, ground and nearby conductors | Which conductors belong to the antenna and pattern model |
| Complex load | Resistance and reactance at the antenna-side transformer plane on every band | Whether the chosen ratio and compensation suit the installed modes |
| Network loss and stress | Completed transformer and choke under representative complex loads, power and duty cycle | Actual dissipation, temperature, voltage and current margin |
| Exterior-current map | Current along the intended branch, both sides of the choke and at the station | Whether the designed common-mode boundary exists |
| Accepted-power pattern | Validated full-geometry model or calibrated restored-baseline field comparison | Efficiency and directional performance relevant to the operating path |
Use the same source plane, accepted power, site and environmental state for both antennas. A tuner finding low SWR, a receiver sounding quieter or a transformer staying cool at one power level cannot rank complete systems by itself.
The intentional coax branch carries RF: route it away from people, accessible metal, control wiring and equipment. The choke boundary does not replace RF-exposure assessment, protective bonding, lightning protection or safe station disconnect procedures.
Bottom line: in this EF-OCF, the outside of the coax between transformer and choke is a deliberate counterpoise and radiating branch. That makes the return path visible and controllable. It can be a better system than an EFHW when the measured load, completed networks and installed pattern support the comparison—not because the antenna name guarantees it.
Current-path, balance and modelling references
Mini-FAQ
- Is the coax really the counterpoise in this EF-OCF? The outside of the declared coax section carries intentional antenna current between the transformer and choke. The internal differential mode remains the feed line.
- Does every EF-OCF require a 4:1 UNUN? No. Use that ratio only when the installed complex load and completed network support it. The choke remains a separate function.
- Where should the choke go? At the intended end of the exterior-current branch, verified by current measurements and installed behaviour on every operating band.
- Is the coax perfectly quiet after the choke? No. Choke impedance is finite and current can be re-excited farther along the route, so measure downstream current and at the station.
- Does a separate counterpoise wire always work better? No. It creates a different branch geometry whose loss, coupling, current distribution and pattern must be verified.
- Does this architecture always beat an EFHW? No. It can be better when the deliberate return path, transformer loss, choke boundary and installed pattern suit the site. Only same-boundary evidence can rank them.