Where Should SWR Be Measured?

There is no single best SWR-meter location for every task. A measurement is meaningful only when its port, reference impedance, intervening network, frequency range and operating state are stated. Start by defining the device under test and the boundary at which its voltage and current are to be described.

SWR Belongs to a Transmission-Line Mode

For a uniform transmission line with a real characteristic impedance Z₀, the load reflection coefficient is:

VSWR is the ratio of the maximum to minimum voltage magnitude formed by the incident and reflected waves. A directional bridge or VNA does not normally search the cable for those extrema; it measures incident and reflected travelling waves at its calibrated plane and derives reflection coefficient, return loss or SWR.

These quantities describe the differential transmission-line port. They do not directly measure radiation efficiency, gain, pattern, ground loss, matching-network loss or current on the outside of a coaxial shield.

Keysight’s reflection-measurement guide defines reflection coefficient as the reflected-to-incident wave ratio. NIST likewise states that scattering parameters are referenced to a specified idealised transmission line in its microwave S-parameter measurement service.

Three Useful Measurement Questions

All three results can be correct at the same time. A shack tuner can present a low SWR to the transmitter while a much higher SWR remains on the feedline connected to the tuner output. The tuner transforms impedance at its ports; it does not necessarily change the antenna or eliminate standing waves on the line beyond it.

What Changes Along a Uniform Feedline

On a uniform, lossless line, the complex impedance changes with position even though |Γ| and SWR remain constant. For a line of length ℓ:

A VNA at the shack can therefore show a resistance and reactance very different from those at the antenna even when the line is lossless. That is impedance transformation, not a change in SWR.

On a real line with propagation constant γ = α + jβ, the reflection coefficient at a distance ℓ from the load is:

Loss attenuates the reflected wave on the trip from the load to the instrument, so SWR normally appears closer to 1:1 toward the source. A surprisingly good shack reading can therefore coexist with substantial load mismatch and feedline dissipation. Connectors, adapters, chokes, switches and faults add further discontinuities; the measured reflection is then the vector sum of multiple contributions.

Keysight’s field cable-and-antenna measurement guide separates insertion loss, return loss and distance-to-fault measurements for this reason.

Calibration and De-Embedding

A VNA calibration establishes a reference plane at the calibration standard connection. Anything added after calibration—test cable, adapter, fixture, choke or transformer—is part of the measurement unless it is separately removed.

Use one of these approaches:

• Calibrate at the antenna port. Move the calibrated cable end to the installed feedpoint and avoid changing its shape after calibration.
• De-embed a characterised feedline or fixture. Measure or model its complex S-parameters across the complete band, then mathematically remove it.
• Use port extension only within its assumptions. A simple delay correction moves phase through a known uniform line but does not remove unknown attenuation, mismatch or connector errors.
• Use time-domain gating carefully. It can help isolate separated discontinuities, but windowing, bandwidth and overlapping reflections limit the result.

The Rohde & Schwarz de-embedding guide distinguishes the VNA calibration plane from the device plane and shows why the intervening network must be characterised. Keysight’s network de-embedding note gives the same boundary-based treatment.

Resonance, Match and Loss Are Different Results

At a named port, resonance usually means the input reactance crosses zero:

The resistance R can still differ greatly from Z₀, so a resonant antenna need not have low SWR. A matching network can also transform a non-resonant load to a low-SWR input at one frequency. That does not move the load’s physical resonance; it creates a matched input for the combined network.

For a matched source and fixed incident power at one interface, the mismatch factor is 1 − |Γ|² and the corresponding mismatch loss is:

Those values are interface calculations, not total antenna-system loss. Actual delivered power also depends on source mismatch, tuner loss, feedline attenuation, repeated reflections, transmitter foldback and load loss. Keysight’s mismatch treatment shows why complex source and load reflection coefficients are needed for a complete result.

Common-Mode Current Changes the Measurement Fixture

In the intended coaxial differential mode, current on the centre conductor is balanced by equal opposite current on the inner surface of the shield. Current on the shield’s exterior is a different path involving the antenna, feedline route, mast, station wiring, ground and surrounding structures.

A coaxial one-port measurement senses the relationship between centre conductor and shield at its port. It does not separately report exterior-shield current. If that exterior path is significant, moving the test cable, changing the analyser position or altering nearby conductors can change the electromagnetic system being measured.

A common-mode choke adds impedance to the exterior-current path. It can help keep the test cable on the instrument side from becoming a substantial part of the antenna, but it has finite, frequency-dependent impedance and loss. The choke does not define the VNA reference plane; calibration and de-embedding do that. It only changes or suppresses one current path.

Measure exterior cable current with a calibrated clamp-on RF current probe when possible. The complete cable passes through the probe so the intended internal differential currents cancel in its aperture while net cable current remains. Record frequency, power, probe position and the probe’s transfer impedance. Com-Power’s RF current-probe note describes the method and its calibration.

End-Fed, Off-Centre-Fed and Vertical Systems

These systems require particular care because the external return path can be installation dependent:

• End-fed antennas: install the intended counterpoise, feedline route, transformer and choke before characterising the system. A small analyser connected directly to an otherwise isolated matching box may supply a different return path from the final station.
• Off-centre-fed antennas: asymmetrical excitation can drive feedline exterior current. Measure it instead of assuming a current balun or choke is effective across every operating band.
• Verticals: the radial, elevated-counterpoise or ground system belongs inside the device boundary. Exterior coax current can become an unintended additional return path.
• Random and long wires: tuner, transformer, earth connection, counterpoise, wiring and capacitance to the environment can all affect the input. State which of them were present during measurement.

The connector remains a valid physical port, but a repeatable port measurement requires the rest of the current paths to be defined and controlled. The phrase “the feedpoint is at the first choke” is therefore too simple: the feedpoint, the chosen device boundary and the point where common-mode current becomes small are related but not identical concepts.

Diagnose Changes Without Guessing

A result that changes when the coax is rerouted is evidence that the setup is not invariant. Common-mode current is one strong hypothesis, but cable faults, loose connectors, nearby-object coupling and calibration-cable movement can produce changes too.

1. Inspect and verify connectors, adapters and cable continuity.
2. Repeat the calibration without changing the cable bend after calibration.
3. Measure the feedline with a known matched load at the far end.
4. Repeat the antenna sweep with a documented cable route and surrounding geometry.
5. Measure exterior current at several fixed points with a calibrated RF current probe.
6. Insert a characterised common-mode choke and repeat without changing other variables.
7. Compare complex S₁₁, not only the minimum SWR value.

A choke-induced change does not automatically mean the choke “fixed” the antenna. It proves that the common-mode network or added component influenced the measured system. Check matching loss, exterior current, pattern and installed performance before choosing the final arrangement.

A Repeatable SWR Measurement Workflow

1. State the question. Antenna impedance, feedline SWR and transmitter load are different measurements.
2. Draw the system boundary. Mark transmitter, tuner ports, feedline, choke, transformer, counterpoise, radiator and ground system.
3. Choose the reference impedance and plane. Record Z₀, connector and exact physical location.
4. Calibrate or de-embed. Remove only networks that have been characterised adequately across the band.
5. Control the fixture. Fix cable routing, analyser position, mast, counterpoise and nearby conductors.
6. Sweep complex S₁₁. Save resistance, reactance, return loss and SWR over the full operating range.
7. Check common mode separately. Use an RF current probe and controlled choke or routing comparisons.
8. Measure both tuner sides when relevant. Confirm what the transmitter sees and what remains on the feedline.
9. Account for dissipative loss. Include feedline, tuner, transformer, choke, conductor and ground losses before making an efficiency claim.
10. Report uncertainty and state. Record frequency, temperature, power, calibration method, cable configuration and repeatability.

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