Non-Resonant HF Traps: When Broadband Current Shaping Helps

Traditional parallel-LC traps are familiar: put a high impedance in the radiator near one frequency and use the conductor beyond it differently on another band. I also like a broader design option—a section that shapes current across a declared range rather than relying on one sharp parallel resonance. But I will not call that option universally superior. First define the network, then show what it does in the complete antenna.

“Trap” Describes a Job, Not One Circuit

An antenna trap is a two-terminal section used to alter current in a radiator. Several very different networks can do that job:

Even a section designed without a deliberate LC resonance can resonate through winding capacitance, distributed line length or interaction with the radiator. “Non-resonant” should therefore mean that the intended operating mechanism does not depend on one sharp resonance inside the declared range—not that resonance is absent from all frequencies.

The Installed Current Ratio Is the Design Variable

A high standalone impedance does not automatically create the desired antenna current distribution. The result depends on the impedance looking into the conductor on both sides, the section position, electrical length, feed arrangement, ground and return paths, nearby objects and the other loading structures.

At each operating frequency, measure or model current magnitude and phase at marked conductor positions before and after the section. The useful quantity may be a current ratio such as |Iouter/Iinner|, but that ratio needs a defined position and phase reference. A low outer current may be correct for one band and harmful on another where the outer conductor is meant to radiate.

Radiation pattern is the vector result of the full current distribution in its installed geometry. A smoother current curve does not guarantee smoother azimuth response, lower elevation angle, more gain or higher efficiency. Those are separate model or measurement results.

Broadband Must Name the Metric

A current-shaping section can work over a wider frequency interval than a high-Q parallel trap. That is a valid design possibility, not a guaranteed outcome. “Broadband” must say which requirement remains inside its limit:

• minimum section impedance or a target resistance/reactance region;
• maximum outer-section current relative to inner-section current;
• feed-point impedance or SWR at a declared reference plane;
• accepted power, radiation efficiency or realized gain;
• pattern, polarization or null depth;
• temperature rise at stated power, mode, duty cycle and ambient conditions; or
• repeatability across built samples and installed environments.

Low SWR over a broad range is particularly easy to misread. A resistive ferrite section or another loss can damp a resonance and widen the match while reducing radiated power. SWR shows reflection at its reference plane; it does not separate radiation from heat.

Reflection and Dissipation Are Different Mechanisms

In a calibrated two-port fixture with matched reference impedances, incident power at port 1 is divided among reflection, transmission and dissipation. For a passive section under those fixture conditions, the unaccounted fraction is:

A low S21 alone is ambiguous: the section may be reflecting energy reactively, dissipating it, or doing both. Record complex S11 and S21, fixture loss and uncertainty. For a reciprocal device, reverse measurements are a useful consistency check.

The 50-ohm fixture is not the antenna. The conductors on either side of a series trap are generally not matched 50-ohm ports, so fixture insertion loss cannot be copied directly into an antenna-efficiency claim. Use the measured network as a component model, embed it between the actual radiator impedances, and verify accepted power, dissipative heating and radiated performance on the complete structure.

Resonant LC Traps Have Strengths and Costs

Near parallel resonance, a finite-Q LC trap can create a large impedance and a sharp current boundary with relatively little series dissipation when the components are good. Off resonance, the same network supplies inductive or capacitive loading; that behaviour can be part of a deliberate multiband design rather than an unwanted side effect.

The costs are design-specific. Voltage across the trap and circulating current within the inductor and capacitor can be much larger than the current entering the branch. Component Q, self-resonance, capacitor voltage and current ratings, coil spacing, enclosure capacitance, corona margin, conductor loss and thermal rise all matter. A high-Q trap can be narrow and sensitive, but a well-built, sealed trap with adequate component margin can also be stable and repeatable.

Do not penalize every LC trap for the weakest garden example. Measure the actual trap and antenna.

Broadband Sections Move Stress Rather Than Abolish It

A reactive loading or transmission-line section may distribute impedance over more length and avoid one intentional high-Q tank. It can still create local voltage maxima, current maxima, conductor loss and parasitic resonances. A shorted or open line has its own frequency-dependent transformation; velocity factor, spacing, bends, support dielectric and termination matter.

A ferrite section adds another boundary. Fair-Rite’s technical data represent ferrite permeability as a complex quantity: the real part contributes inductive reactance and the imaginary part contributes an equivalent resistance. Both change with frequency, material and geometry. Material curves and small-signal impedance are starting data, not a transmitter power rating.

When ferrite resistance is placed directly in wanted antenna current, heat is evidence of dissipative loss. That loss may be an intentional compromise to obtain current control or bandwidth, but it belongs in the efficiency budget. Core volume, turns, flux density, electric-field stress between turns, winding capacitance, ambient temperature, airflow, enclosure and duty cycle determine whether the structure remains linear and thermally stable.

Weather and Production Need Their Own Tests

No topology owns environmental stability. Moisture can change coil and enclosure capacitance, wet conductor insulation, leakage and nearby ground coupling. Temperature changes metal dimensions, capacitor values, dielectric properties and ferrite permeability. Wind changes wire angle and coupling. Corrosion changes contact resistance.

Likewise, “simpler to manufacture” is a measured production claim. LC traps have inductor, capacitor, enclosure and assembly tolerances. Ferrite or line sections have material, dimension, turn count, winding placement, conductor length, joint, adhesive and parasitic tolerances. Compare enough samples to report the distribution of impedance, resonant features, current ratio and temperature—not only the best prototype.

Compare Complete Antennas, Not Labels

The fair comparison gives both complete antennas the same operating bands, geometry envelope, conductor material, feed/return-path boundary, test site and declared tuning allowance. If each design is optimized separately, say so and record every changed dimension. Otherwise a radiator-length change can masquerade as a trap result.

A Measurement Programme That Can Decide

Add wet/dry and temperature cycles when weather claims matter. Run an A/B/A sequence or matched builds so drift is visible. Report VNA calibration, fixture removal, current-probe transfer impedance, power-meter planes, site uncertainty, thermal-sensor emissivity/contact and repeatability. Low-power data cannot certify high-power behaviour, and a warm ferrite core cannot by itself quantify total antenna efficiency.

The Defensible Case for Broadband Current Shaping

Broadband current-shaping sections deserve a place in HF antenna design. They can provide a gradual, useful impedance profile where one sharp LC boundary is not the objective. They may reduce tuning sensitivity or distribute stress in a particular design. They can also waste wanted power, resonate parasitically, heat, shift with temperature or produce the wrong installed pattern.

So my case is deliberately conditional: choose the broadband section when its measured current distribution, loss, stress, thermal margin, pattern and repeatability beat the resonant alternative for the declared installation. Choose the LC trap when its sharper current boundary and measured efficiency fit the objective better. Engineering is not picking the modern-sounding label. It is showing where the current and power went.

Primary and authoritative technical sources

• ARRL: HF Trap Antennas—parallel-tuned trap practice, between-band trapping and practical component/power boundaries.
• David Birnbaum, K2LYV: “Design of a Two-band Loaded Dipole Antenna”—parallel-LC trap reactance below, at and above resonance within a complete radiator design.
• ARRL QST: “A Portable Dual-Band End-Fed Half-Wave Antenna”—practical trap voltage/current and measured thermal power-handling boundaries.
• Fair-Rite 17th Edition Catalogue, Technical Information—complex permeability, equivalent reactive/resistive impedance, frequency, geometry and temperature data.
• Keysight: Impedance Measurement Handbook—complex impedance, real component parasitics, test fixtures, calibration and compensation.
• IEEE 149-2021—antenna impedance, pattern, gain, efficiency, site and uncertainty measurement practice.

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