SWR Cannot Validate Gain, F/B or Radiation Efficiency
SWR Cannot Validate Gain, F/B or Radiation Efficiency
Why the field screenshots in Greg Mihran's antenna primer do not confirm the model's principal radiation-performance claims—and why that does not make the screenshots worthless.
RF.Guru working definition: Common-mode current is the non-cancelling phasor-sum current in a specified set of conductors, evaluated at a defined cross-section and using a declared current-direction convention. In the intended differential transmission-line mode, the outgoing and return currents are equal and opposite, so their phasor sum is zero. When they do not cancel, the remaining current must close through another reference or return path—such as the outside of a coax shield, a mast, equipment chassis, station wiring, nearby structures, earth, the operator, or distributed coupling through the environment.
This broader working definition is especially useful in practical antenna systems. On transmit, non-cancelling current on the outside of the coax can make the feedline and connected structures part of the radiating antenna system unless that path is intentional, clearly defined and properly controlled—for example by providing the required return path and placing a suitable common-mode choke at the correct boundary.
Greg Mihran's July 2026 antenna primer repeatedly places precise 4NEC2 predictions beside photographs of low-SWR RigExpert sweeps. Gain, efficiency, radiation angle, beamwidth and front-to-back ratio appear alongside the field results. I want to separate the two things that the presentation brings together: a measured input match and a predicted radiation performance.
The screenshots need not be false for the inference to be false. They answer a different question. My objection is not that Mihran measured SWR, nor that antenna modelling is useless. It is treating an attractive port measurement as experimental confirmation of quantities that the port measurement cannot see.
Measurement boundary: SWR describes mismatch magnitude at one electrical port. Gain and front-to-back ratio require calibrated field data with direction. Radiation efficiency requires radiated power relative to accepted power.
Six Photographs Do Not Confirm Every Column Beside Them
Slide 58 makes the distinction especially clear. Six real analyser photographs sit beside a table explicitly headed Computer Model Specifications. The table predicts gains from +0.67 to +4.15 dBi, a peak radiation angle of 86°, efficiencies of 85.5–87.7%, and beamwidths of 75–131°. Those are model outputs, not quantities read from the six SWR screens.
The slide's headline combines broad matching with gain near the zenith. Its photographs do support useful practical information about reflection across the displayed bands. They do not supply an elevation scan, calibrated field comparison or radiated-power measurement. Six bands of the same observable are still only one kind of observable.
| What is presented | What the evidence actually addresses | What it does not establish |
|---|---|---|
| PERformer: model results in slides 42–50, field sweeps in slide 51 | Modelled radiation behaviour and measured input-match curves are separate results. | The photographs do not measure the model's gain, 24° peak angle or efficiency. |
| PERformer NVIS: slide 58 | Six SWR photographs and a separately labelled model table. | The photographs do not confirm the 86° peak, gain, beamwidth or efficiency columns. |
| Parasitic PERformerArray: slides 60–69 | Deployment and input-match evidence accompany the coupled-array model. | A driven-port match does not reveal the induced element current or installed directional pattern. |
| Phased PERformerArray: slides 75–79 | For example, slide 75 labels its calculated traces as 50 Ω, source 1. | That is not a measurement of the completed physical feed network, element-current phase or front-to-back ratio. |
| Challenger and Dominator: slides 95–109 | Radiator models and field sweeps through practical matching arrangements. | A low transformer-input SWR does not determine radiation efficiency or transformer loss. |
| ChallengerArray: slide 122 | A RigExpert screenshot overlays a calculated 200 Ω SWR plot next to modelled patterns. | Visual overlap does not align the physical reference planes or turn the patterns into measurements. |
| DominatorArray: slides 127–128; later designs in slides 131–144 | More model predictions and practical low-reflection displays. | More SWR screenshots do not become a gain or radiation-efficiency survey. |
I am not concluding that these radiation predictions must be wrong. I am concluding that the screenshots do not validate them. The primer does label model results; that distinction needs to survive when their performance is described or compared.
What a Calibrated One-Port Measurement Establishes
A vector network analyser measures complex reflection coefficient S11 at its calibration plane. For a real reference impedance Z0 under the usual power-wave convention:
Γ = S11 = (Zin − Z0)/(Zin + Z0)
SWR = (1 + |Γ|)/(1 − |Γ|)
Preflected/Pincident = |Γ|2
Paccepted/Pincident = 1 − |Γ|2
Keysight's reflection-measurement tutorial distinguishes the reflected-to-incident wave ratio from what passes into a load. An SWR display retains only |Γ|; complex S11 also retains phase and can be converted to input impedance when its assumptions are met.
A useful report includes frequency span, reference impedance, calibration method and plane, fixture or feedline, instrument uncertainty, antenna configuration and nearby environment. If complex measured and modelled S11 agree after those boundaries are aligned, the model has passed an input-impedance test.
Nothing in SWR identifies azimuth, elevation, polarization, beamwidth, sidelobes, a forward direction or a rear direction. It also does not divide accepted power between radiation and loss.
Accepted Power Is Not Radiated Power
The primer's repeated SWR-matching box, including slide 58, gives 99.8% for 1.10:1 and 96% for 1.50:1. The arithmetic is correct as accepted incident-power fractions at the stated port. It does not calculate radiation efficiency. Slide 155 is more careful in referring to power delivered to the antenna structure; the next question is what that structure does with it.
What happens after power crosses the reference plane requires another power balance:
Paccepted = Pradiated + Ploss
ηrad = Pradiated/Paccepted
ηtotal = (1 − |Γ|2)ηrad
The last relation includes mismatch and is commonly called total efficiency; the terminology and system boundary should be stated. A feedline, transformer, matching network or choke belongs inside the efficiency boundary if it lies between the declared input plane and the radiator.
| SWR | |Γ| | Reflected incident power | Accepted incident power |
|---|---|---|---|
| 1.10:1 | 0.0476 | 0.23% | 99.77% |
| 1.20:1 | 0.0909 | 0.83% | 99.17% |
| 1.50:1 | 0.2000 | 4.00% | 96.00% |
| 2.00:1 | 0.3333 | 11.11% | 88.89% |
At the same 1.50:1 SWR, a 1%-radiation-efficient load radiates 0.96% of the incident power, while a 95%-efficient antenna radiates 91.2%. Those are illustrative power balances, not results for Mihran's antennas. The match is identical; the radiation is not.
The NIST method for determining a lower bound on antenna efficiency uses reverberation-chamber observations and stated assumptions to obtain additional information about efficiency. That is a much richer evidence set than an ordinary S11 sweep.
Two Counterexamples Keep the Terms Straight
A 50 Ω dummy load can show almost 1:1 SWR while converting nearly all accepted RF power to heat. An efficient antenna can show the same port match while radiating most of its accepted power. Identical SWR does not imply identical radiation efficiency.
Conversely, an ideal lossless antenna at 2:1 SWR accepts 88.89% of the incident travelling-wave power and radiates all of that accepted power. Its radiation efficiency is 100%, while its total efficiency and realized gain include the 0.51 dB mismatch penalty.
Keep the boundary explicit. Multiple reflections, transmitter behaviour and feedline loss can change delivered system power. They still do not let a single port-reflection result separate radiation from dissipation.
Gain, Directivity and Realized Gain
IEEE 145-2025 is the active IEEE standard for antenna terminology. For radiation intensity U(θ,φ):
Prad = ∫4π U(θ,φ)dΩ
D(θ,φ) = 4πU(θ,φ)/Prad
G(θ,φ) = ηradD(θ,φ)
Grealized(θ,φ) = (1 − |Γ|2)G(θ,φ)
Directivity describes how radiated power is distributed with angle. Gain also includes radiation efficiency. Realized gain additionally includes mismatch at the stated port. SWR supplies only that last mismatch factor; it does not supply radiation intensity, total radiated power or loss.
NIST Technical Note 1551 provides useful antenna definitions in a radiometry context. For general antenna terms, IEEE 145 remains the direct definitions reference.
Front-to-Back Ratio Requires Directional Data
This is why a claimed deep rear null in the PERformerArray discussion cannot be confirmed by another low-SWR trace. Front and back are directions in the field, not coordinates on the analyser's one-port display.
Antenna orientation is not an S11 coordinate. Rotate a directional antenna by 180° in a uniform environment without changing its feed geometry and its input match may remain essentially unchanged even though its main lobe now points the other way.
F/B = 10 log10(Ufront/Uback)
F/B = 20 log10|Efront/Eback| for the same field component, distance and wave impedance
A valid F/B result defines the forward direction and whether “back” means exactly 180° from it, the strongest rear lobe or a rear-sector value. It also states frequency, elevation, polarization, angular resolution, range geometry, multipath control and uncertainty. A deep rear null is especially sensitive to element-current error, feedline current, ground reflection and nearby objects.
Reference Plane, Feed Network and Common Mode
The overlay on slide 122 is a concrete example. Its calculated ChallengerArray trace is explicitly referenced to 200 Ω. The nominally 50 Ω RigExpert sweep is shown alongside the practical matching arrangement discussed in the primer. An ideal 4:1 transformation can relate suitable impedances; the real network, intervening cable and measurement plane still have to be accounted for. Matching two plotted minima does not account for their losses.
The same distinction matters for Dominator's kilohm-class end feed and high-ratio transformer. Slide 91 itself lists approximately 0.51 or 1.08 dB of combined transformer/choke loss for the alternatives it discusses. Those are the primer's stated values, not fresh measurements of every installation. They nevertheless illustrate why the radiator's modelled efficiency and the complete system's efficiency are different quantities even when the input SWR is low.
An analyser at the station end of a feedline does not directly observe the radiator terminals. A matching network can transform impedance, while loss in the network and line attenuates both outward and returning waves. For a uniform line:
Γin = ΓLe−2(α+jβ)l
|Γin| = |ΓL|e−2αl
A good instrument-end SWR may therefore coexist with feed-network loss. Compare at the same plane by including the real network in the model, moving the calibration plane or characterizing and de-embedding the intervening network.
Common-mode current adds another path. Current on the coax exterior, mast or nearby conductors can change loss and radiation pattern while the one-port match still looks good. NIST’s discussion of antenna-measurement challenges treats unwanted common-mode current as a measurement concern. Document choke impedance and cable routing, measure exterior current where practical and repeat after a controlled routing change.
Impedance Agreement Is Partial Model Validation
Terminal resistance can contain radiation resistance, conductor loss, ground loss, matching loss and unintended feedline participation. Different current distributions can produce similar terminal impedance yet different gain and pattern. That non-uniqueness matters for parasitic and phased arrays because their pattern depends on complex element currents.
Report impedance agreement narrowly: measured complex input impedance agrees with the model within the declared frequency, reference-plane, configuration and uncertainty limits. Validate current distribution, gain, efficiency and pattern with independent observables.
Two identical SWR readings are a narrower agreement still: they match |Γ|, not necessarily its phase or the complex impedance. For Mihran's arrays, the meaningful next information is how the element currents and radiated field compare with the model—not another photograph of the same scalar.
Measurements That Address the Actual Claim
IEEE 149-2021 is the active recommended practice for measuring transmitting and receiving antenna properties. It covers radiation-pattern measurement and antenna test facilities.
- Gain: use calibrated substitution, two-antenna or three-antenna methods, or a suitable near-field measurement transformed to the far field. NIST’s Technical Note 1311 describes an extrapolation-range gain and polarization method.
- Pattern and F/B: measure calibrated field over the required azimuth and elevation angles with polarization, distance, reflections and uncertainty controlled.
- Radiation efficiency: measure radiated versus accepted power, derive ηrad from gain and directivity, integrate a calibrated full-sphere pattern, or use an applicable calibrated chamber or Wheeler-cap method.
- Array operation: measure complex element currents or port waves where accessible, characterize the feed network and mutual coupling, and verify the installed pattern.
NIST’s reverberation-chamber work on radiation and total efficiency measures the missing power relationship. A one-direction field comparison can demonstrate a directional level difference, but without sufficient pattern data it cannot distinguish higher total radiation from redistribution toward that direction.
A controlled A/B field comparison can therefore answer a useful, specific question: which installation places more signal in the tested direction? With equal accepted antenna-port power it compares gain; with equal incident power at the declared system input it also includes mismatch and intervening losses. State which comparison was made. Neither result should be renamed total radiation efficiency without the additional power-distribution evidence.
Impedance Bandwidth Is Only One Bandwidth
The broad matching spans in the primer are useful to an operator who wants to know where retuning may be needed. That is impedance bandwidth under the tested conditions. Gain, efficiency, beam direction, F/B, null depth, matching-network loss and common-mode current can change across the same span.
The NTIA Antenna System Guide distinguishes terminal-impedance bandwidth from bandwidth defined by acceptable gain, pattern or another performance criterion. Loss can also broaden a match, so a wide, smooth SWR curve is not automatically a wide radiation-performance result.
Match Every Claim to an Observable
| Claim | Suitable evidence | Report at least |
|---|---|---|
| Input match and impedance bandwidth | Calibrated complex S11 | Touchstone data, reference plane, fixture, feedline and environment |
| Relative or absolute gain | Calibrated substitution, antenna method or controlled field comparison | Power convention, reference antenna, direction, polarization and uncertainty |
| F/B, beamwidth and lobe direction | Azimuth/elevation pattern or suitable near-field scan | Angular convention, resolution, multipath control and field levels |
| Radiation efficiency | Radiated-versus-accepted power, G/D, chamber or applicable Wheeler method | System boundary, mismatch treatment, feed-network loss and uncertainty |
| Array pattern | Complex element data plus measured pattern | Amplitude, phase, coupling, network loss and common-mode current |
| Model validation | Agreement across independent observables | Geometry, material losses, ground, convergence, currents, power budget and measurements |
What I Would Take From the Primer—and What I Would Not
I would use the RigExpert sweeps as practical matching evidence and the models as predictions worth examining. I would not combine them into a claim that the published field work has measured every gain, angle or efficiency number in the adjoining tables.
That is not a request to abandon either instrument or modelling. It is a way to make the engineering claim stronger: say what was predicted, say what was measured, and keep the power and reference-plane definitions consistent. A model-supported design advantage is a prediction with stated assumptions; a measured directional advantage needs field evidence. Neither becomes stronger by giving an SWR photograph a job it cannot do.
The screenshots can be genuine and the inference still be wrong. A good match is useful. It is not a measurement of where the accepted watts went.
Primary technical references
- IEEE 145-2025 — Standard for Definitions of Terms for Antennas
- IEEE 149-2021 — Recommended Practice for Antenna Measurements
- NIST — Antenna gain and polarization calibration services
- NIST TN 1311 — Extrapolation-range gain and polarization measurements
- NIST — Radiation and total efficiency in a reverberation chamber
Mini-FAQ
- What does SWR prove about an antenna? SWR quantifies mismatch magnitude at a stated reference plane. Calibrated complex S11 can also support input-impedance validation, but neither measurement establishes gain, pattern or radiation efficiency.
- Can an inefficient load have a one-to-one SWR? Yes. A dummy load can accept almost all incident power and convert it to heat while presenting an excellent match.
- Can a high-SWR antenna still be radiation-efficient? Yes. Radiation efficiency is radiated power divided by accepted power. Mismatch is a separate factor in total efficiency and realized gain.
- Can SWR validate an antenna model? Agreement in SWR corroborates reflection magnitude only. Complex S11, including phase at an aligned reference plane, can corroborate modelled port impedance. Neither independently validates current distribution, gain, pattern or efficiency.
- What measurement establishes front-to-back ratio? A controlled angular field or pattern measurement that defines the forward and rear directions, elevation, polarization, resolution and uncertainty.
- What measurement establishes radiation efficiency? A method that relates radiated power to accepted power, directly or through a recognized method such as gain-to-directivity comparison, full-sphere integration, a calibrated chamber or an applicable Wheeler cap.