Measure EFHW SWR and Resonance at the Right Reference Plane

I do not want an EFHW analyser trace without a circuit drawing. The same wire can produce a different curve when the counterpoise changes, the coax moves, the choke moves or the instrument body becomes part of the return path. That does not make SWR useless. It means we must measure the system we actually intend to operate.

State What the Instrument Is Measuring

A one-port vector network analyser measures the complex reflection coefficient at its calibration plane. From that value it can display return loss, SWR and complex impedance. Everything beyond that plane—adapter, cable, choke, transformer, counterpoise and radiator—contributes to the reported network unless it is accurately de-embedded.

Here Z is the complex impedance at the measurement plane and Z₀ is the analyser's reference impedance, commonly 50 Ω. SWR retains only the magnitude of Γ. Saving resistance and reactance—or complex S₁₁—preserves the information needed to distinguish several possible causes of the same SWR value.

Resonance and Minimum SWR Need Not Coincide

At a named input plane, a common definition of resonance is the frequency where the input reactance crosses zero:

The resistance can still be far from 50 Ω, so resonance need not produce a low SWR. Conversely, a transformer or matching network can move a reactive antenna load closer to 50 Ω at its input and create a low-SWR point for the combined network. That does not prove that the bare wire is resonant at the same frequency.

For an EFHW, the coax-side trace can include the high-impedance wire mode, the chosen return branch, transformer magnetising impedance, leakage, winding capacitance, any compensation capacitor, fixture parasitics and common-mode conversion. Call the result the input resonance or match of that declared assembly—not automatically the resonance of an isolated half-wave wire.

Do Not Assume One Feedpoint Impedance or Ratio

An end-fed half-wave-like mode often presents a high terminal impedance, but it is not universally 2–5 kΩ. The value changes with conductor geometry, height, ground, frequency, feed arrangement, loading, loss, nearby objects and the return system. The transformer ratio therefore cannot be selected from the EFHW name alone.

A nominal 49:1 or another ratio states an impedance transformation under ideal assumptions. It does not guarantee 50 Ω at the coax connector, low loss, correct compensation, low common-mode conversion or safe voltage and temperature across the operating bands. Measure the installed complex load and qualify the transformer into representative loads.

The small signal from a VNA is appropriate for impedance characterisation. It does not reproduce operating-power flux, voltage, dielectric stress, connector heating or ferrite temperature. Perform separate controlled power and thermal tests after the analyser has been removed and protected.

Define the EFHW Return Path

Current cannot leave one transformer terminal without a second-terminal structure. An EFHW return can include a dedicated counterpoise, radial or conducting support, a deliberate section of coax exterior, station wiring and distributed capacitance to the environment. Several branches can participate at once.

Inside coax, the intended transmission-line mode uses the centre conductor and the inner surface of the shield. Additional net current can flow on the shield exterior. If a declared exterior section is meant to be part of the antenna, include its length and route in the device under test and place the common-mode boundary at its intended end. If a separate counterpoise is meant to close the current locally, control exterior current near that boundary instead.

The counterpoise and choke are not accessories added after the measurement. They determine the circuit being measured. Disconnecting the intended return structure to “see the wire alone” forces the analyser, test lead and stray capacitance to provide another return and creates a different antenna.

A Fixed Coax Fraction Is Not a Reference-Plane Rule

A 0.05λ coax jumper does not have a universal measurement property. In the intended internal coaxial mode, its electrical length depends on cable propagation and frequency, and any mismatched section transforms complex impedance. On an ideal lossless line, SWR magnitude remains constant while reflection phase and R + jX rotate. Real cable adds attenuation, so a longer path reduces the reflected wave before it returns to the instrument and can make source-end SWR appear closer to 1:1.

The exterior-current path is a different circuit. Its electrical behaviour depends on cable route, jacket, soil, mast, station, nearby conductors and termination. The manufacturer's internal-mode velocity factor does not locate a universal exterior-current maximum, minimum or choke position.

Use a jumper because it is characterised, mechanically repeatable and long enough to place the instrument safely—not because one wavelength fraction magically removes transformation. Calibrate at its far end or de-embed its measured S-parameters when the question requires the antenna-side plane.

Place the Choke at the Intended Current Boundary

A common-mode choke adds complex impedance to the exterior-current path while ideally disturbing the intended differential coaxial mode very little. It does not define the VNA calibration plane; calibration and de-embedding do that.

Transformation and common-mode suppression are separate jobs. Where the measured load is intentionally unbalanced and calls for the ratio, an UNUN can perform the impedance transformation while a separately specified choke defines the exterior-current boundary. That is a practical arrangement, not a universal claim for one ratio or topology. A suitable current balun remains valid where the installed port is genuinely balanced and its loss, stress and common-mode behaviour support the use.

Choose choke placement from the intended return geometry and measured exterior current. Verify complex choke impedance, differential insertion loss, voltage, current and temperature over every required band. A changed SWR after adding the choke proves that the network changed; it does not by itself prove improvement or identify the previous radiation loss.

Use Measurement Configurations Deliberately

A station-end result is not meaningless. It answers the station-end question. It simply cannot be relabelled as wire-terminal impedance without a line model or de-embedding. Likewise, bypassing a tuner is useful when characterising the downstream antenna system, but the tuned result remains valid evidence of the load presented to the transmitter at the tuner's input.

Read the Shape of the Curve Carefully

A narrow dip can come from a high-Q resonance or matching network, but it does not prove high radiation efficiency. A broad dip can result from genuine bandwidth, multiple coupled responses, transformer or conductor loss, ground loss, common-mode participation or insufficient sweep resolution. Broad is not automatically bad, and sharp is not automatically good.

Multiband EFHW minima are not exact integer harmonics in every installation. End effects, height, bends, ground, loading, transformer reactance, compensation and coupling shift the modes differently. Record the frequencies where X crosses zero and where SWR is minimum, then inspect R + jX around both points.

If the curve moves when the coax, analyser or nearby conductor moves, common-mode participation is a strong hypothesis—not the only possible cause. Connector faults, calibration-cable flex, poor repeatability and changing environmental coupling can also move the trace. Test each mechanism rather than diagnosing from one symptom.

A Repeatable EFHW Measurement Sequence

1. Draw the complete system. Mark transformer ports, radiator, intentional return, coax exterior, choke, feedline route, mast, station bonds and nearby conductors.
2. Name the result needed. Decide whether the question is wire-side resonance, transformer-input match, station-end SWR, component loss, exterior current or radiation.
3. Fix the installed geometry. Record wire length and shape, height, ground condition, return conductor, cable route and choke position.
4. Choose and calibrate the plane. Keep the adapters and test cable in the same configuration after calibration; use accurate de-embedding only for a characterised network.
5. Save the vector sweep. Record frequency, complex S₁₁ or R + jX, reference impedance, sweep power, points and bandwidth—not only the minimum SWR.
6. Map exterior current separately. Use a calibrated RF current probe around the complete coax at marked positions on both sides of the intended choke boundary.
7. Change one variable. Move one choke, alter one return branch or use one known cable while holding the remaining geometry fixed.
8. Restore the baseline. Repeat A/B/A to expose connector repeatability, instrument drift and accidental cable movement.
9. Repeat band by band. Wire modes, transformer behaviour, line transformation and common-mode current all change with frequency.
10. Qualify power separately. Remove the analyser, reconnect protection and check loss, voltage, current and temperature under the intended load, waveform and duty cycle.

What the Final Record Should Contain

A useful EFHW record includes the circuit drawing, photographs or coordinates of the installed geometry, conductor dimensions, return structure, transformer and compensation state, coax type and length, choke position, analyser model and calibration plane, complex sweeps, exterior-current map and environmental conditions.

For efficiency or pattern claims, add measured component/feedline losses, accepted power at a common plane, thermal results, a validated full-geometry model and calibrated field or controlled A/B/A path measurements. SWR alone cannot separate radiation from transformer, feedline, conductor, ground or common-mode loss.

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