The EH Antenna: What Physics Predicts and Measurement Must Prove
The EH Antenna: What Physics Predicts and Measurement Must Prove
The EH antenna was presented as a compact radiator that creates electric and magnetic fields in the right local relationship. That is a testable proposition—not an exemption from small-antenna physics.
I have no objection to a small antenna radiating. Of course it can. My objection begins when a good match, a narrow box at the feedpoint or a successful QSO is presented as proof that a compact structure performs like a much larger reference antenna. The EH claim deserves the same treatment as every other antenna claim: identify all current-carrying conductors, define the reference plane, measure accepted power, and measure what reaches the far field.
The Claim That Started the Argument
Ted Hart, W5QJR, described the EH antenna as a new concept in which an external network develops electric and magnetic fields simultaneously and in the physical relationship required for radiation. His published introduction attached strong benefits to that idea: smaller size, increased efficiency and bandwidth, reduced received noise and reduced interference.
That originating claim matters. If the EH label meant only “a compact loaded antenna that can be matched,” there would be little controversy. Compact loaded antennas have been understood for a long time. The extraordinary part is the suggestion that arranging local E and H fields through phasing creates a separate route around the familiar size, stored-energy and loss constraints.
Maxwell’s equations give us a direct way to examine that proposition. They do not prohibit clever geometry or a well-designed matching network. They do require the radiated field to follow the complete time-varying charge and current distribution of the installed structure.
Local E and H Are Not Two Independent Ingredients
Close to an antenna, electric and magnetic fields contain reactive terms associated with energy stored around the conductors. Their magnitudes, phases and spatial variation depend on the complete source current, charge, geometry and surroundings. In the radiation zone of a localized source in ordinary free space, the outward-propagating electric and magnetic fields are transverse, mutually related and carry power according to the Poynting vector.
That far-field relationship is a consequence of the electromagnetic solution. It is not an extra switch that can be set inside a small box to make stored near-field energy become loss-free radiation. A network can change terminal current, conductor current and resonance. Those are legitimate design actions. The resulting current distribution still determines the field.
The useful engineering translation: do not ask whether the network has “put E and H in phase” at one chosen point. Ask where current flows, where charge accumulates, how much net power the antenna accepts, how much becomes heat, and what gain and pattern the complete installed system produces.
Small Relative to a Wavelength
“Small” should be an electrical statement, not a photograph. Antenna theory commonly uses ka, where k = 2π/λ and a is the radius of the smallest sphere enclosing the radiating structure. When ka is below roughly one, the antenna is electrically small and stored energy becomes an increasingly important design constraint.
Chu’s spherical-mode analysis established a lower bound on the radiation quality factor of an ideal antenna confined to a sphere. Later work refined how antenna Q and impedance bandwidth should be defined for real tuned one-port antennas. These are bounds and relationships, not a claim that every small antenna must be inefficient. An electrically small antenna can be efficient when conductor, dielectric, matching and environmental losses are kept sufficiently low. The practical difficulty is that small radiation resistance and large reactive energy make those losses and tolerances increasingly important.
I therefore would not assign one radiation resistance to every EH build from height alone. Cylinder dimensions, spacing, current distribution, the phasing or matching network, support material, enclosure, feed and nearby conductors all matter. A short-dipole or short-monopole equation is useful only when its current-distribution and ground assumptions actually match the structure being analysed.
Matching Does Not Establish Radiation Efficiency
A network can transform the input impedance of a compact antenna to something close to 50 Ω. That helps a transmitter deliver power at the chosen reference plane. It does not tell us how the accepted power divides between radiation and dissipation.
Radiation efficiency is the ratio of radiated power to net power accepted at the antenna terminals. In a simple series equivalent at a declared plane, it can be written as:
ηrad = Prad / Paccepted
ηrad = Rrad / (Rrad + Rloss)
The second expression is useful only when the equivalent resistances refer to the same current and reference plane. In a distributed antenna with a multi-component network, it is safer to account for conductor, capacitor, inductor, dielectric, connection, ground and other coupled losses explicitly.
An SWR trace gives reflection at its calibration plane. It cannot separate radiation resistance from loss resistance. A lossy network can produce a broad, attractive match because power is being dissipated. A low-loss small resonator can show a much narrower match. Neither curve alone proves the efficiency claim.
Q and Bandwidth Are Evidence, Not a Verdict
For a simple, isolated resonance with declared matching conditions, impedance bandwidth and Q are closely related. As electrical size falls, the best achievable lossless radiation Q rises. That is why compact antennas often become tuning-sensitive and narrowband.
But bandwidth is not an efficiency meter. Additional resistance can lower the measured Q and widen the SWR bandwidth while reducing radiation efficiency. A matching network can also create more than one resonance or alter the shape of the impedance curve. Conversely, a narrow response is consistent with high stored energy but does not quantify how much accepted power is radiated.
When someone claims that an EH build is simultaneously tiny, broadband and highly efficient, the three quantities must be measured separately:
- Electrical size: state the enclosing dimensions and wavelength at the operating frequency.
- Bandwidth: state the criterion, reference impedance, calibration plane and matching network.
- Efficiency or gain: use a method that measures radiated performance and reports uncertainty.
The Feedline May Be Part of the Antenna
A compact two-terminal structure still needs a complete current path. If the intended conductors and network do not provide that path cleanly, current can flow on the outside of the coax, through control wiring, mounting hardware, an earth connection or nearby structures. Those conductors can radiate and reshape the pattern.
This is not merely a purity argument. The extra conductor may be contributing useful radiation, destructive cancellation or loss. It also makes the result dependent on cable length, routing, station bonding and the measurement setup. NIST’s antenna-measurement roadmap specifically notes that unwanted common-mode current on a feed cable can alter measured patterns.
A feedline-current measurement is therefore valuable, but it must be interpreted carefully. Adding a choke is not a neutral observation: it changes the boundary conditions. If the match or field changes when a characterised choke is introduced, that is evidence that the previous feedline path mattered. It does not by itself reveal how much total efficiency changed or prove that the remaining compact radiator is poor.
There is no universal counterpoise length or fixed choke impedance that validates every EH installation. Map the exterior current along the line, document cable routing, and measure the choke’s complex common-mode impedance over the band. Then repeat the radiated test.
A Fair Comparison Must Measure the Right Quantity
A single signal report is not an efficiency measurement. A field-strength reading in one direction combines accepted power, antenna gain in that direction, polarization, pattern, terrain, multipath and instrument uncertainty. A QSO adds propagation and receiver variables. Those observations can be useful, but they cannot carry the whole conclusion.
For a practical outdoor A/B comparison, I would use this discipline:
- Freeze the installation. Record every conductor, support, feedline route, choke, earth connection, matching network and nearby object. The reference antenna gets the same care.
- Calibrate at the antenna plane. Measure complex impedance and account for feedline and fixture loss. Equal transmitter indication is not necessarily equal accepted power at the two antennas.
- Use the same frequency and waveform. Record duty cycle and component temperature. A network that remains cool at low duty cycle has not earned a continuous-power rating.
- Control order and time. Use rapid A/B/B/A swaps or simultaneous receivers where practical. Propagation, fading and local noise must not be mistaken for antenna performance.
- Measure more than one direction. A stronger result at one point can be a pattern difference. Pattern, gain and polarization data make the comparison transferable.
- Measure exterior current. Repeat with documented common-mode boundaries so the radiating system under test is known.
- Report uncertainty. Range reflections, alignment, polarization, calibration drift, cable movement and power measurement all belong in the result.
To determine radiation efficiency rather than directional field strength, use a recognised method such as pattern integration, a properly applied Wheeler-cap method for a suitable electrically small antenna, or a validated reverberation-chamber method. Each method has conditions and uncertainty. IEEE 149-2021 is the sensible starting point rather than inventing a backyard percentage from one distant receiver.
What a Choke and Thermal Test Can Reveal
Feedline-current and thermal checks remain useful when bounded properly. A current probe can show whether the cable exterior participates in the installed current path. A calibrated thermal test can reveal where some accepted power becomes heat. Neither measurement alone gives radiation efficiency.
Temperature rise depends on loss, thermal mass, airflow, ambient temperature, emissivity, duty cycle and test duration. Record all of them. Check capacitors, coils, conductors, connections and lossy support material. Infrared readings on shiny conductors can be misleading unless emissivity is controlled.
Likewise, do not prescribe “moderate power” without a safety and component boundary. Begin at low power, use rated components and suitable RF clearance, keep people away from the near field, and increase power only under a documented test plan.
Why Contacts Still Tell Us Something—But Not Enough
A successful contact proves that the installed system transmitted enough energy in the useful direction, at that time, for the other station to decode it. That is real and worth enjoying. It does not prove high efficiency, a particular radiation mechanism or equality with a full-size reference.
The same caution applies to “quiet on receive.” A low noise level may result from lower sensitivity, a different pattern, polarization rejection, reduced common-mode pickup, less receiver overload or lower environmental coupling. The useful quantity is wanted-signal SNR under controlled comparison, not the S-meter’s noise reading by itself.
Compact Alternatives Still Need Evidence
A small transmitting loop, loaded vertical, folded structure or capacitively top-loaded radiator can be an excellent engineering choice when its constraints fit the site. None is a universal winner. A loop can trade tuning bandwidth and high capacitor voltage for compactness. A loaded vertical can trade conductor and loading-coil loss against height, and it needs a defined return system. A longer wire may be easier to make efficient but can create a different pattern or mechanical problem.
Compare candidate antennas by the job: required bands, available volume, direction and elevation pattern, efficiency, bandwidth, retuning, feedline current, voltage and current stress, weather stability and safe installation. The name on the antenna does not answer those questions.
Primary and Authoritative Sources
- Ted Hart, W5QJR, “EH Antennas”—the originator’s own statement of the phasing concept and claimed benefits evaluated here.
- L. J. Chu, “Physical Limitations of Omni-Directional Antennas”—the foundational spherical-mode treatment of size, gain and Q.
- A. D. Yaghjian and S. R. Best, “Impedance, Bandwidth, and Q of Antennas”—Q and impedance-bandwidth definitions for tuned lossy and lossless one-port antennas.
- IEEE Std 145-2025, Standard for Definitions of Terms for Antennas—current antenna terminology.
- IEEE Std 149-2021, Recommended Practice for Antenna Measurements—gain, pattern, impedance, efficiency and uncertainty practice.
- NIST contributors, “Antenna Measurement Challenges and Opportunities”—measurement uncertainty and feed-cable common-mode effects.
- C. L. Holloway, reverberation-chamber radiation and total-efficiency methods—validated efficiency approaches and their assumptions.
- K. T. McDonald, “Crossed-Field and EH Antennas”—a source-specific Maxwell analysis including feedline radiation and earth reflection.
Joeri’s Bottom Line
The EH antenna does not need ridicule, and it does not need a new law of radiation. Treat it as a complete compact antenna system: conductors, phasing or matching network, feedline exterior, supports, ground paths and surroundings. A clever network may make a useful installation. Its value has to appear in calibrated gain, pattern, efficiency, bandwidth and stability data.
My critique remains direct: “E and H in phase” is not a performance result, a 50 Ω match is not an efficiency result, and a QSO is not a controlled comparison. Show the current paths and measure the radiated system. If an EH build performs well under that test, the measurement—not the label—deserves the credit.
Mini-FAQ
- Does an EH antenna create a new kind of electromagnetic radiation? No demonstrated EH mechanism sits outside Maxwell’s equations. Its far field follows the complete current and charge distribution of the installed conductors, network, feedline and surroundings.
- Does electrically small automatically mean inefficient? No. Small size raises stored-energy and matching challenges, while practical efficiency depends on radiation resistance and all conductor, component, dielectric and environmental losses.
- Can SWR prove that an EH antenna is efficient? No. SWR reports reflection at a reference plane. Radiation and loss both contribute to input resistance, so a good match cannot separate useful radiated power from heat.
- Does narrow bandwidth prove low efficiency? No. Narrow bandwidth can be consistent with high stored energy, but it does not quantify efficiency. Added loss can also broaden an SWR curve while reducing radiated power.
- How can feedline radiation be detected? Map current on the coax exterior and repeat impedance, pattern and field measurements with a documented common-mode boundary. A change shows that the boundary mattered, not by itself how efficient the antenna is.
- What is the fairest way to compare an EH antenna? Compare at equal net accepted power with frozen geometry and cable routing, calibrated reference planes, controlled polarization and range conditions, repeated A/B/B/A trials, pattern data and stated uncertainty.