Yagi Antennas: Common-Mode Listening Machines?

Here is my deliberately sharp warning: a Yagi can become an excellent common-mode listening machine. If current is allowed onto the coax exterior, or if the boom, mast, rotor cable and station bonding become part of an unintended RF path, the receiver may be listening with a much larger structure than the driven and parasitic elements on the drawing.

But the word can matters. A Yagi-Uda array is not inherently defective, and common-mode current is not guaranteed merely because coax and metal supports exist. The result depends on feed geometry, symmetry, cable route, bonding, electrical lengths, nearby conductors and the impedances that close each return path. The right response is to expose those paths and measure them—not to prescribe a magic choke value.

The Intended Coax Mode Is Only One Current Path

In the intended coaxial TEM mode, current on the centre conductor returns on the inner surface of the shield. Skin effect lets a different current exist on the shield exterior. That exterior current closes through whatever the installation offers: the antenna, mast or tower, station chassis, bonding network, other cables, earth and distributed capacitance.

Roy Lewallen, W7EL, draws this boundary clearly in “Baluns: What They Do and How They Do It”. The internal coaxial currents remain equal and opposite, while a separate outside-shield current can make the feed line radiate or receive. If that imbalance current is already small, adding a balun or choke may produce little useful change.

A clamp around the complete coax measures the signed sum of current through its aperture at that cross-section. With the internal coaxial mode well confined, that is normally a useful estimate of exterior-shield current. It is not a universal proof that every measured ampere belongs to one surface or one physical cause; connector leakage, parallel conductors, probe placement and the conductor set inside the aperture still define the measurement.

How an Installed Yagi Converts Modes

A nominally balanced driven element can still launch exterior current when the two terminal environments differ. The coax shield is electrically connected to one side of many feed arrangements, and the cable route changes the impedance of that side relative to the other. A gamma match is intentionally asymmetrical; a split dipole with a symmetrical feed arrangement is different. Neither label alone predicts the installed current.

The parasitic elements, boom and mast also carry induced current as part of the complete electromagnetic structure. That current is not automatically “coax common mode.” The diagnostic question is whether the feed and support system creates an additional path that changes the currents and fields from the intended array.

• Feed-point asymmetry: terminal hardware, matching components, connector placement and nearby metal can make the two sides see different impedances.
• Cable routing: coax that hugs the boom, passes close to an element or descends asymmetrically can couple to the array and alter the exterior-current path.
• Support structure: boom-to-mast connection, tower geometry, rotator wiring and bonding determine which conductors participate and where current can return.
• Electrical length: the outside path can have maxima and minima that move with frequency, routing, termination and nearby conductors.
• Station boundary: entry bonding, equipment interconnections, control cables and protective earth can change the path without repairing the imbalance that launched it.

This is why “HF is always worst” and “VHF is too short to matter” are both poor rules. Physical length must be expressed electrically, and the complete route and terminations matter. A short-looking section can be a substantial fraction of a wavelength at VHF; a long HF run can land near a current minimum at one band and a maximum at another.

What Common-Mode Reception Can Change

Exterior current makes the feed and support system another receiving and radiating structure. Its field combines vectorially with the field of the driven and parasitic elements. The resulting change may alter azimuth or elevation pattern, front-to-back ratio, polarization, input impedance and response toward local noise sources.

Those are possibilities, not guaranteed outcomes. The added field can reinforce one direction and cancel another. A null can fill while a different bearing improves. The wanted signal and background can rise together, fall together or move by different amounts. A single station, a single noise-floor screenshot or an S-meter change cannot establish gain, efficiency or the complete pattern.

ITU-R P.372-17 is careful about the same boundary: its environmental-noise data cover noise received through the reference antenna and feeder, while excluding other conducting cables, inadequate screening and feeder imbalance. If your test lets those paths change, you are no longer comparing only the Yagi elements.

A Choke Is a Measured Network Component

A common-mode choke inserts a frequency-dependent complex impedance into the chosen exterior-current path. Its useful value is not one headline number. Record R + jX over the operating range, because a mainly reactive impedance, a deliberately lossy impedance and a self-resonant structure behave differently. Also measure wanted-mode insertion loss, return loss and heating under the real mismatch and power conditions.

There is no universal 5 kΩ or 10 kΩ target. Required impedance depends on the driving imbalance and the rest of the common-mode loop. The current reduction that matters is the installed before/after result at the declared plane, with an acceptable change in wanted-mode loss and stress. A higher impedance in a fixture is useful evidence; it is not a guarantee that another mast, cable route or band will deliver the same SNR or pattern.

Feed-point placement is often a sensible candidate because it can define the boundary before the coax exterior becomes part of the structure. It is not automatic. A conductive mast or another cable may bypass that boundary; a choke may be electrically too far from the conversion point; or a different location may sit where the path is easier to control. Likewise, “half a wavelength down the coax” is not a placement rule. The common-mode path does not necessarily share the cable’s internal velocity factor, and changing the choke changes the standing wave it is supposed to control.

Ferrite around the complete coax acts on net current through the core. Double braid or foil does not make that principle disappear. Core material, turns, winding capacitance, cable construction, frequency, voltage, current and temperature determine performance. ARRL’s common-mode choke measurement article shows why self-resonance and frequency sweep matter and why current should be checked along the installed cable rather than at one convenient spot.

Route, Mast and Bonding Belong in the Same Model

Routing the coax away from the elements and avoiding long parallel runs near the boom are useful starting points because they reduce one obvious coupling opportunity. A right-angle departure is not a law: mechanical layout, polarization, mast geometry and wavelength decide whether it helps. Photograph the actual route and include it in the model or test record.

Bonding has multiple jobs. Protective and lightning bonding exists for safety and must not be defeated to improve an RF experiment. RF bonding can also alter the common-mode network, sometimes beneficially and sometimes by creating another current path. ITU-T K.37 treats screening, cable installation, separation, filtering, earthing and bonding as coordinated EMC measures, not interchangeable slogans.

A shack-end ground does not by itself prove that the feed-point conversion has been controlled. Nor does a feed-point choke prove that mast, rotator and station-entry paths are irrelevant. Declare the boundary you are trying to enforce, then measure on both sides of it.

Map Exterior Current Before Choosing Hardware

Use a calibrated clamp or current transformer around the complete coax. Record the frequency-specific transfer impedance or transfer factor, termination, detector or receiver input, bandwidth, cable correction, probe loading, noise floor and uncertainty. Keep every other conductor outside the aperture unless the intended measurand is the net current of a larger cable bundle.

Mark positions along the feed line and map current at each one. Include the feed point, mast transition, rotator loop, tower base or station entry where accessible and safe. A single low reading may be a standing-wave minimum. Repeat after changing the route or choke, because the current distribution can move.

For receive diagnosis, add a second map of the local-noise field. Use a small probe or portable receiver to locate candidate sources, then switch them or their coupling path under controlled conditions where safe. Do not call the problem “common mode” merely because the noise sounds impulsive, or because a choke happened to change it.

Use A/B/A for Signal, Noise and Pattern

When a wanted carrier occupies the measurement bin, the indicated power contains signal plus noise. Estimate noise in the same reference bandwidth and subtract in linear power before calculating SNR. Repeat the cycle, use a stable source where possible and record propagation or source drift. A large dB change is publishable evidence only when the reference plane, bandwidth, receiver state, source stability and uncertainty are also stated.

Pattern claims need more than one bearing. Rotate the Yagi against several stable sources, use a controlled range, or compare a full-wave model with measured installed currents and geometry. IEEE 149-2021 treats pattern, gain, impedance, instrumentation, test site and measurement uncertainty as parts of one antenna-measurement problem. Use JCGM 100 to state the measurand and uncertainty instead of presenting display resolution as accuracy.

Keep the Warning, Earn the Verdict

A Yagi can indeed let the feed line, boom, mast and station join the listening structure. That possibility deserves sharper attention than “the SWR is fine.” But the engineering verdict belongs to the installed current map and controlled performance record.

Choose a current boundary. Measure the exterior-current distribution. Characterize the choke as a complex two-mode component. Freeze the receiver state. Compare signal, noise, SNR and pattern in A/B/A cycles. Then you can say whether this Yagi was listening through its intended array, through an accidental metalwork network—or through both.

Primary and authoritative technical sources

• Roy Lewallen, W7EL: “Baluns: What They Do and How They Do It”—coax inner/outer currents, imbalance paths and pattern effects.
• ARRL QST: “Common-Mode Current and Common-Mode Chokes”—current probing, impedance sweep, self-resonance and installed verification.
• ITU-R P.372-17: Radio noise—external radio-noise categories and the antenna/feeder versus other-conductor boundary.
• ITU-T K.37 (01/2024)—current EMC guidance for cabling, screening, filtering, earthing and bonding.
• IEEE 149-2021—antenna pattern, gain, impedance, instrumentation, site and uncertainty measurement practice.
• BIPM/JCGM 100—measurement models and expression of uncertainty.

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