Why S11 Alone Cannot Measure Antenna Efficiency

A beautiful 1.1:1 SWR proves that little power is reflected at the chosen reference plane. It does not prove where the accepted power goes. A 50 Ω dummy load and a well-matched antenna can present almost the same S11 while producing radically different radiation.

Start With the Efficiency Definition

At a declared antenna port, let P_inc be incident power, P_ref reflected power, P_acc accepted power and P_rad radiated power. For a linear single-port antenna under the stated conditions:

The first equation is what a calibrated S11 measurement can support: it measures Γ at its reference plane and therefore the mismatch factor. The second needs a radiated-power result. Total efficiency combines radiation efficiency and mismatch, so every report must say which efficiency it means.

The language also has to match the measurement boundary. Does “antenna” include the matching network, loading coil, transformer, radials, feedline section or enclosure? Moving the reference plane can move a real loss from “feed system” into “antenna” without changing the station's radiated watts.

Why One Complex Number Is Not Enough

At resonance, a simplified antenna terminal model is often written:

This is useful only when both resistances are referred to the same current and port, and when the chosen model contains the relevant modes and losses. The VNA measures their sum as part of the input impedance; it does not label one part “radiation” and the other “heat.”

Matching makes the ambiguity even easier to hide. A network can transform 5 Ω, 500 Ω or a complex load to 50 Ω. The instrument then reports a good match at its port while conductor, dielectric, ground, core, coil and common-mode losses remain inside the accepted-power budget.

What a One-Port VNA Actually Establishes

With a suitable calibration and stable test environment, a VNA can measure:

• complex reflection coefficient, return loss and SWR;
• complex input impedance versus frequency;
• resonance and anti-resonance behaviour;
• the mismatch factor at the calibrated reference plane;
• changes caused by tuning, loading, geometry or nearby objects; and
• network quantities such as cable or matching-network transmission when a valid two-port fixture is used.

Those are important measurements. They tell you whether the source can deliver power into the declared port and whether the electrical model agrees with reality. They do not, by themselves, say how much accepted power was radiated.

Bandwidth is not a shortcut. Loss can broaden a resonance, while coupling and matching can alter the loaded response. An efficiency estimate derived from Q needs a validated equivalent circuit, appropriate unloaded-Q method and a defensible separation of radiation and non-radiation losses.

S11 Needs a Second Observation

A VNA can supply the coherent amplitude-and-phase measurements used by several accepted antenna-efficiency techniques. The limitation applies to the one-port reflection experiment, not to the instrument:

Change the experiment and the same instrument can help reveal the missing radiation or loss term.

Methods That Add the Missing Information

1. Gain and directivity

A calibrated antenna range can measure gain. A sufficiently complete, polarization-aware radiation pattern can be integrated to obtain directivity. With consistent accepted-power and mismatch conventions:

This is not a one-point field-strength test. Gain calibration, pattern coverage, polarization, alignment, range reflections, cable movement, ground, reference antennas and uncertainty all matter. At HF, the required range and site control can be demanding.

2. Wheeler-type radiation suppression

For suitable electrically small antennas, a conducting enclosure can suppress radiation while leaving loss mechanisms to be inferred from a second impedance or reflection measurement. The method uses VNA data, but the result depends on enclosure size, cavity modes, detuning, fixture currents and the antenna model. It is not a universal “put a box over it” recipe, especially for large or multiresonant HF structures.

3. Reverberation-chamber methods

A stirred reverberation chamber creates a statistical multipath environment in which transfer measurements, chamber decay or reference methods can be used to estimate radiated power and efficiency. NIST has published one-, two- and three-antenna approaches and their uncertainty considerations. S11 is still used to correct mismatch; transmission or decay statistics provide information that one-port reflection lacks.

4. Calorimetric and loss-accounting methods

If every significant dissipative loss can be measured, efficiency can be built from accepted power minus loss power. This is straightforward in principle and difficult in practice. RF current is not necessarily uniform, loss resistance is frequency- and temperature-dependent, and heat can spread into structures that are hard to instrument.

The familiar I²R calculation is valid for a defined current and RF resistance. Feedpoint current multiplied by a guessed “ground resistance” does not automatically describe distributed soil, conductor, coil, joint and dielectric loss.

Why Subtracting a Model From R_in Is Only an Estimate

Measuring input resistance at resonance, calculating radiation resistance and treating the remainder as loss can be a useful engineering estimate for a well-defined geometry, but it is not a general measurement of efficiency.

The uncertainty can be large because:

• the modeled current distribution may differ from the installed antenna;
• nearby conductors, earth and the feedline can alter both radiation and loss;
• a matching network transforms resistance between reference planes;
• common-mode current can create an additional radiating path;
• the resonance condition X = 0 does not prove a simple series equivalent circuit; and
• subtracting two uncertain numbers can produce a very uncertain remainder.

If you use this method, publish the geometry, soil model, conductor and joint assumptions, current reference, calibration plane, modeled R_rad, measured R_in, sensitivity analysis and uncertainty. Call the result an estimate unless it has been validated independently.

A Far-Field A/B Test Measures Realized Gain Difference

Swapping a reference antenna and an antenna under test while recording received power can be extremely useful. What it most directly compares in one direction is realized gain under the two test conditions—not radiation efficiency alone.

The observed difference can include:

• radiation efficiency;
• directivity and elevation-pattern changes;
• mismatch and feed-system loss;
• polarization mismatch;
• ground reflection and multipath;
• propagation variation; and
• changes in common-mode current or surrounding objects during the swap.

For station decisions, realized gain in the desired direction may be the metric you actually care about. Just do not rename it “efficiency” without also establishing directivity and the relevant mismatch boundary.

HF VNA Measurements Fail in Predictable Ways

Reference-plane and feedline errors

Calibrate at the antenna connector when practical, or use a characterized cable and a valid de-embedding method. A calibration at the instrument followed by an uncharacterized feedline reports the impedance at the wrong plane. Port extension corrects electrical delay; it does not automatically remove cable loss, connector errors or radiation from the feedline.

Common-mode current changes the DUT

A VNA, laptop, USB cable, operator and coax can become part of the antenna. Use an appropriate choke or isolation arrangement, check outside-shield current at several positions and repeat the measurement after changing cable routing. If the trace moves, the “antenna impedance” was partly an installation-mode result.

External signals and receiver overload

HF antennas collect broadcast and local transmitter energy. An interferer can compress the receiver or corrupt the coherent measurement, producing ripple, drift or impossible impedance. Repeat with different stimulus power and IF bandwidth, inspect receiver levels and measure at a quiet time. If attenuation or filtering is added, include it in the calibration and uncertainty budget; a generic 10 or 20 dB pad is not automatically the right cure.

Environment and heating

People, wet soil, foliage, buildings and cables alter the antenna. Small-signal VNA data also do not prove that a loading coil, ferrite or joint remains linear and cool at transmitter power. Pair low-level impedance work with controlled stepped-power thermal and current tests when the antenna will transmit.

A Defensible Amateur-Radio Workflow

1. Define the metric. State radiation efficiency, total efficiency, gain, realized gain or link performance—do not use “efficient” for all of them.
2. Define the boundary. Identify the antenna port and whether feedline, matching unit, transformer, radials and common-mode path are inside the result.
3. Measure S11 correctly. Calibrate at the declared plane, save complex data and verify stability against stimulus power, cable position and time.
4. Measure feed-system loss separately. Characterize cables and matching networks with appropriate fixtures instead of assigning all accepted loss to the radiator.
5. Control common mode. Measure outside-shield current and repeat after changing choking and routing.
6. Add a radiation or loss method. Use calibrated gain and pattern, a validated Wheeler fixture, a reverberation chamber, calorimetry or a carefully bounded model-plus-loss estimate.
7. State uncertainty. Include calibration, drift, repeatability, reference antenna, range, alignment, environment and model sensitivity.
8. Validate at operating conditions. Check temperature, power, duty cycle and installation changes before turning a small-signal result into a station claim.

The Defensible Verdict

A one-port S11 sweep tells you how much incident power is reflected and what complex impedance appears at the calibration plane. It does not separately reveal radiation and dissipation, so it cannot by itself yield radiation efficiency.

A VNA can nevertheless be central to a valid antenna-efficiency measurement. Pair it with an experiment that observes radiated power or suppresses/quantifies radiation, keep mismatch conventions explicit and publish enough uncertainty information for someone else to understand what was actually proved.

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