When Open-Wire Feed Line Becomes Part of the Antenna

The myth says coax radiates because it is “unbalanced,” while open-wire line cannot radiate because it is “balanced.” Neither label settles the installed result. Coax can carry exterior-shield current. Open-wire line can carry a common-mode component, and even its intended differential mode has finite external field because the conductors have finite spacing and finite length. The right question is not which cable has the better label. It is which modes exist, where their currents flow and what those currents do to the measured field.

Balanced Means a Mode, Not a Guarantee

Define both conductor-current arrows in the same physical direction along the line. The two measured currents can then be decomposed into differential and common components:

Pure differential mode has equal-magnitude, opposite-direction currents: I_c = 0. Their external fields cancel strongly when the conductors remain close compared with a wavelength and follow consistent geometry. Pure common mode has in-phase current under this arrow convention: I_d = 0. Its return is distributed through antenna structure, equipment, nearby conductors and electric-field coupling to the environment.

Most real installations carry both modes. “Balanced line” describes a geometry intended to support the differential mode with symmetric impedance to its surroundings. It does not certify the antenna arms, tuner, supports, building and station cables as one balanced electromagnetic system.

Bockelman and Eisenstadt’s mixed-mode scattering theory formalises differential, common and conversion terms for coupled conductors. Keysight’s balanced-measurement documentation shows how S_dd, S_cc, S_cd and S_dc keep wanted transmission and mode conversion separate.

Geometry Creates Cancellation—and Its Limits

The statement “radiation follows imbalance, not geometry” goes too far. Geometry sets conductor spacing, characteristic impedance, mutual coupling and the phase relationship between fields observed away from the line. Reduce spacing relative to wavelength and preserve equal-and-opposite current, and far-field cancellation generally improves. Increase spacing, distort one conductor differently, add bends or change the terminal leads, and the radiated field can change even before a simple current-ratio label tells the whole story.

A finite, perfectly differential two-wire line is not mathematically invisible. Source and load transitions, finite conductor separation, ends and bends all contribute to the complete field. The 2015 primary study Radiation Physics from Two-Wire Transmission Lines calculates total radiation from straight and bent two-wire geometries and shows that current distribution and the locations of conductor segments matter. Its gigahertz examples are not HF threshold tables; the transferable lesson is that geometry and current must be evaluated together.

Open-wire line usually radiates little in its intended mode because the opposing conductors are close compared with a wavelength and carry nearly equal currents. “Little” is an engineering result to verify, not a synonym for zero.

Coax Is Not Unbalanced Merely Because It Is Grounded

Coax is an asymmetric two-conductor structure: one conductor surrounds the other, and the outer conductor is normally the equipment-reference terminal. That port geometry—not the optional presence of a DC earth bond—is why it is conventionally called unbalanced. A floating length of coax does not become a balanced pair merely because neither end is grounded.

In the intended coax TEM mode, current flows on the centre conductor and the inner surface of the shield, with the fields concentrated between them. A separate current can flow on the shield’s outer surface relative to the installation and environment. At HF, skin effect makes the inner- and outer-surface current paths usefully distinct, while finite shield transfer impedance, apertures and connectors can still couple them.

The open-wire comparison is therefore not “shielded cable bad, exposed cable good.” The coax geometry gives strong confinement for its differential mode; exterior-shield current belongs to another installed mode. Open wire exposes both conductors to the environment, so unequal proximity more readily changes their capacitance and coupling. The NBS precision-coax monograph ties coax characteristic impedance directly to its concentric dimensions, while current on the cable exterior is treated as a separate circuit in the IEEE EMC Society current-probe paper.

Where Differential Energy Converts to Common Mode

Mode conversion appears where the two conductor paths stop seeing equivalent boundary conditions. The important asymmetry may be at the line, but it may just as easily be at the antenna or equipment.

A line leaving a symmetric dipole approximately at right angles is often a sensible starting geometry because it avoids a long parallel run beside one antenna arm. It is not a law that the line must remain straight for exactly 0.1λ. The useful distance depends on the antenna current distribution, line spacing, height, supports, nearby conductors and frequency. Preserve symmetry where practical, then measure.

Transmit and receive do not acquire different return-path physics. In a passive linear installation, the same asymmetry that lets differential transmitter energy excite common mode can let environmental fields or conducted noise convert into receiver differential voltage. In either direction, conduction and displacement current close the complete circuit; the absence of a drawn “third wire” does not remove the environmental return.

There Is No Universal −20 dB or −10 dB Boundary

A current-imbalance ratio can be useful when its definition and measurement are clear. For example, an amplitude ratio may be written as 20 log₁₀(|I_c|/|I_d|). That number is not a universal radiation threshold.

The field produced by the line depends on the spatial current distribution, phase, physical length, spacing in wavelengths, bends, height, termination, nearby conductors and observation direction. A −20 dB current ratio on a long resonant route may matter more than −10 dB on a short, tightly coupled section. The instrument noise floor can also decide when a current becomes detectable; it cannot decide when physics “starts.”

Replace the threshold with an engineering limit tied to the actual objective:

• maximum allowed pattern change or cross-polarised field;
• maximum accessible common-mode current or RF voltage;
• maximum receive-noise increase at the receiver input;
• maximum conducted or radiated disturbance at a specified measurement point; or
• repeatable improvement relative to measurement uncertainty.

A regulation, equipment specification or controlled design can assign a numeric limit for one setup. It should not be recycled as a universal law of open-wire antennas.

A Tuner or Balun Label Does Not Finish the Job

A genuinely balanced tuner can reduce one transition asymmetry, but its output balance still depends on topology, component tolerances, enclosure coupling and the connected load. An unbalanced tuner followed by a current balun can also work well when the balun presents sufficient complex common-mode impedance and survives the actual differential voltage, current, mismatch, duty cycle and temperature.

A current balun impedes one unwanted current path; it does not force every installed current to be equal by name. Its winding capacitance and surrounding metal can create a parallel path, and a high impedance at one frequency is not a broadband power rating. A voltage balun that produces equal terminal voltages does not guarantee equal-and-opposite currents into an asymmetric load.

At each transition, keep four questions separate:

• Does the network provide the required differential impedance transformation?
• How much differential insertion loss does it add under the actual complex load?
• What common-mode impedance and mode-conversion performance does it provide over frequency?
• Can its conductors, insulation and magnetic material survive the installed voltage, current, waveform and temperature?

One-port SWR at the transmitter answers none of those questions completely. The tuner can present a good match at its input while standing waves and common-mode current remain downstream.

Measure Current, Field and Pattern as Different Quantities

A calibrated RF current probe around both open-wire conductors responds to their algebraic sum. Under the convention above that is 2I_c. It does not directly report radiated power, differential current or pattern. Probe transfer impedance, conductor position in the aperture, orientation, frequency response, field pickup and measurement uncertainty belong in the record. CISPR 16-1-2 treats current probes as calibrated radio-disturbance measuring apparatus rather than qualitative clamps.

To estimate I_d, use phase-aware, matched measurements of the individual conductor currents or a characterised balanced fixture. A one-port VNA connected at the tuner plane cannot separate the installed differential and common modes by itself. A multiport mixed-mode measurement can quantify conversion through a device or fixture, but the test port definitions, common- and differential-mode reference impedances, calibration and de-embedding must be declared.

IEEE 149-2021 documents the range, instrumentation and uncertainty discipline needed for defensible antenna-pattern measurements. In a home station, a fixed remote receiver and several stable bearings can provide comparative evidence, but they are not automatically a calibrated antenna range.

Use an A/B/A Test That Can Falsify the Story

Changing the line and seeing the SWR move proves only that the system changed. It does not prove radiation increased or decreased. Use an A/B/A sequence so drift, propagation and wishful interpretation have less room to hide.

Measure common-mode current at several positions because it can form a standing-wave distribution. A convenient current null does not prove the whole line is quiet. Repeat on every intended band. If pattern predictability matters, include the feed line, tuner enclosure, mast, earth coupling and nearby conductors in a full-geometry model such as Lawrence Livermore’s NEC-5, then compare the model with installed measurements.

Design for Symmetry, Then Verify the Installed System

• Keep conductor spacing consistent and give both conductors similar dielectric and conductive surroundings.
• Avoid unequal sharp bends and long runs with only one conductor close to a mast, gutter, frame, wall reinforcement or wiring.
• Use symmetric feed-throughs and supports that remain suitable when wet or contaminated.
• Treat a near-right-angle departure from a symmetric antenna as a starting geometry, not a guaranteed 0.1λ prescription.
• Choose tuner and balun topology from measured differential load, common-mode path, loss and voltage/current stress.
• Measure at marked positions, change one variable and restore the baseline before declaring success.

Coax, twin-lead and open-wire line do not share identical field confinement, but they do share Maxwell’s equations. The useful engineering comparison keeps their geometries and modes explicit. Ideal coax confines its differential field inside the shield; open wire obtains strong but finite differential cancellation from opposing currents; either installation can develop an external common-mode path.

Practical Conclusion

Open-wire line does not suddenly cross a universal dB threshold and become an antenna. It always has a finite field, and its contribution to the installed radiation changes continuously with current distribution and geometry. Disturbed balance can make that contribution large enough to alter pattern, noise pickup, RFI and tuning.

So keep the myth-busting conclusion: radiation is not a coax-only problem. Then finish the sentence properly. Geometry controls cancellation, asymmetry drives mode conversion, transitions set additional current paths, and calibrated A/B/A measurements decide whether the open-wire line is materially participating in this antenna system.

Questions, antenna-factor records or height trials to share? Contact RF.Guru.