The Ugly Balun: Measure the Coax Choke Before You Mock It
The Ugly Balun: Measure the Coax Choke Before You Mock It
A coil of coax can be bulky, awkward and badly copied. It can also be a valid common-mode choke over a measured band. The difference is not beauty—it is complex impedance, installation and evidence.
The “ugly balun” name has earned its jokes. A long piece of coax wrapped around plumbing pipe can look as though the antenna bench collided with the hardware aisle. But appearance is not a transfer function. Before retiring the coil—or trusting it—measure what it does in the common-mode circuit you actually have.
Joeri’s short version: a turn-count recipe is not a choke specification. An air-core coax coil can provide useful common-mode impedance near its designed resonance, while a ferrite choke can be more compact or cover a different band. Neither topology wins without measured R + jX, cable limits and powered verification.
Start by naming the device correctly
Most “ugly baluns” are not impedance transformers. They are coax-air-core common-mode chokes: the same coax continues through the winding, so the wanted differential signal still travels between the centre conductor and the inside surface of the shield. The winding is intended to add impedance to current on the shield exterior.
Calling the coil a 1:1 current balun describes the intended system function, but it can hide two questions: does the antenna-to-line transition actually maintain the desired port balance, and how much common-mode impedance does the finished coil provide over frequency? A common-mode choke is the clearer device-level name.
| Question | Coax-air-core choke | Ferrite-loaded coax choke |
|---|---|---|
| How is common-mode impedance created? | Distributed inductance, interturn capacitance and loss produce one or more resonances. | The winding and core’s complex, frequency-dependent permeability contribute reactance and loss. |
| What sets useful bandwidth? | Cable type, turn count, diameter, spacing, form, surroundings and required minimum impedance. | Material, core geometry, turns, winding capacitance, temperature, common-mode drive and the required minimum impedance. |
| What is the main practical attraction? | No magnetic core, straightforward construction and potentially high impedance over a limited measured band. | Compact geometry and the possibility of useful impedance over a different or broader range. |
| What can go wrong? | A recipe misses the desired band, geometry shifts, the cable bend is invalid or a reactive resonance cancels the return path. | Wrong material or winding, parasitic resonance, core heating, excessive flux, insulation stress or an unverified load. |
The air-core coil is a resonant network
For common-mode current, the coiled shield exterior has inductance. Its turns also have capacitance to one another, the form, earth and nearby conductors. Cable and conductor loss add resistance. The result is a distributed network whose common-mode impedance is complex:
ZCM(f) = RCM(f) + jXCM(f)Below the first important self-resonance, a well-spaced coil often looks mainly inductive. Near a parallel resonance, its impedance can become high enough to make a useful choke. Above that region, capacitance can dominate and further resonances can appear. ARRL’s measured 2024 treatment reaches the practical conclusion cleanly: reactive coax chokes can be excellent near resonance, but a turns-and-diameter recipe does not guarantee performance on a stated frequency.
That does not mean an air-core choke works at only one infinitesimal frequency. Its usable bandwidth is the range over which the measured complex impedance satisfies the installed requirement. A reproducible solenoid winding can cover a useful band; a loose scramble winding is harder to reproduce. “Narrow” and “broad” are results to define and measure, not labels attached to a photograph.
Impedance is not a universal suppression number
The original temptation is to put one number in decibels beside each choke. That shortcut collapses the circuit. If the common-mode source and return path are represented by ZS and ZR, a first-order current ratio after inserting a series choke is:
Iafter / Ibefore ≈ (ZS + ZR) / (ZS + ZR + ZCM)The same choke produces different attenuation in different return-current circuits. Its reactance can add to—or partly cancel—the loop reactance. A measured insertion-loss result also depends on port impedances, fixture topology and reference planes. Converting the magnitude of ZCM into an “impedance-equivalent dB” value does not make it installed current suppression.
Report the full R + jX trace across the required band. Then measure exterior coax current before and after installation with the route, bonds, antenna configuration and power held constant. That evidence is more useful than a universal 5–15 dB or 30–45 dB promise—neither range belongs to a topology by itself.
Ferrite changes the design, not the laws
Ferrite is not a magic replacement for measurement. Fair-Rite’s own material data show that suppression impedance depends on complex permeability, core geometry, turns and frequency. Temperature and bias can derate performance. More turns can raise low-frequency impedance while also increasing winding capacitance and moving resonances.
A suitable ferrite construction can provide a high resistive component over a useful range, damp a common-mode resonance and fit into a compact enclosure. Another material or winding can miss the band, run hot or create an inconvenient impedance profile. Air-core and ferrite designs can both be correct; each needs a stated frequency range, complex small-signal data and large-signal thermal limits.
Resistance has two faces. A resistive component in ZCM can damp the unwanted current-path resonance, but the same term dissipates power approximately as P ≈ ICM,RMS2RCM. Small-signal impedance does not establish transmitter power handling.
The cable still has mechanical limits
Coiling coax does not automatically “torture” it. The mechanical question is whether the chosen cable stays inside the manufacturer’s minimum installation bend radius, repeated-flex limit, pulling limit, jacket temperature and environmental rating. Those limits vary with dielectric, shield and centre-conductor construction.
A too-tight winding can deform dielectric, wrinkle a foil shield, fatigue a solid conductor or load a connector. A generous, supported winding can be entirely legitimate. UV exposure and water ingress are properties of the cable, connectors, form and enclosure—not automatic failures of the air-core topology.
Moisture or nearby conductors can change interturn and environmental capacitance, particularly in an exposed resonant coil. A rigid form, controlled spacing, drainage and weather-stable materials improve repeatability. The correct test is still to compare the dry, wet, cold, hot and mechanically loaded result required by the installation.
Placement follows the current path
A choke only acts in a common-mode circuit that drives current through it. At a genuinely balanced antenna transition, a feedpoint choke can define the boundary between the radiator and feedline exterior. In an antenna that deliberately uses a section of coax exterior as a return branch, placing the choke directly at the feedpoint changes the intended architecture.
The required impedance therefore depends on common-mode drive, return-path impedance, frequency and the placement objective. A feedpoint boundary, station-entry boundary and noise-current boundary are not automatically the same location. Map current on the installed cable instead of fastening the coil wherever the PVC looks least embarrassing.
Qualify the finished choke
- Define the job. State the unwanted current path, frequency range, placement, maximum power, duty cycle and acceptable residual current.
- Preserve the geometry. Record cable type, turn count, winding diameter, spacing, form, ties, connector leads and distance to nearby conductors.
- Calibrate at the fixture plane. Use the impedance or network analyzer method appropriate to the expected impedance and remove fixture delay, stray capacitance and residual inductance within a documented uncertainty.
- Save complex data. Keep R, X and magnitude over the entire required band, not one peak or a screenshot without reference planes.
- Check wanted transmission. Verify differential return loss, insertion loss and connector integrity independently of common-mode impedance.
- Test the installed current. Compare exterior current with the same antenna, route, bonding, tuner state and accepted power.
- Increase power carefully. Monitor cable, connectors, form and any ferrite temperature through the intended duty cycle; inspect again after weather and mechanical exposure.
A pretty trace at low power is not a transmitter rating. Common-mode drive and thermal dissipation depend on the installed antenna and return path. Verify the completed choke under the intended load, power, duty cycle and environment.
Primary and authoritative references
- ARRL QST, March 2024 — Common-Mode Chokes
- Fair-Rite — 17th Edition catalogue and ferrite technical information
- Fair-Rite — suppression-material selection and environmental derating
- Keysight — Impedance Measurement Handbook, 6th Edition
- Keysight — fixture compensation for impedance measurement
- IEC 61196-1-314:2015 — coaxial-cable bending test methods
- IEC 61196-1-100:2022 — coaxial-cable electrical test requirements
Joeri’s bottom line
The ugly balun deserves neither blind worship nor automatic exile. A random cable sculpture copied from a turn-count table is still cable sculpture. A mechanically sound coax-air-core choke with measured complex impedance over the required band is an engineered component, however little glamour its PVC form brings to the mast.
Ferrite often gives us a compact and flexible route to useful common-mode impedance, but it does not repeal frequency dependence, parasitics or heating. Choose the topology that fits the current path, then make the measurement decide. The antenna will not award style points.
Mini-FAQ
- Is an ugly balun really a balun? It is usually a coax-air-core common-mode choke with no impedance transformation. “Balun” describes the intended balance function; the finished installation must demonstrate it.
- Can an air-core coax choke work? Yes. It can provide useful common-mode impedance over a measured frequency range when its geometry, cable limits and installed current path are controlled.
- Does an air-core choke work at only one frequency? Not necessarily. It is resonant and frequency-dependent, but its usable bandwidth is the range over which measured complex impedance satisfies the application.
- Is a ferrite choke always better? No. Ferrite can enable compact or broader designs, but the result depends on material, geometry, turns, parasitics, temperature, drive and the required band.
- How many decibels of suppression will a choke provide? There is no topology-only number. Installed current reduction depends on the choke’s complex impedance and the source and return impedances of the common-mode circuit.
- Can winding coax into a coil damage it? It can if the winding violates the cable manufacturer’s bend, flex, pulling, temperature or environmental limits. A correctly supported winding within those limits is valid.