Two Radials and 4 dB of “Gain”: What Actually Changed?
Two Radials and 4 dB of “Gain”: What Actually Changed?
An asymmetric radial system can alter efficiency, feedpoint impedance, common-mode current and the radiation pattern at the same time. A stronger signal in one direction may be real—but the word gain needs a reference plane and a controlled comparison.
Two radials do not manufacture energy, but neither are they electromagnetically invisible “ground wires.” Their currents are part of the antenna. Changing their number, length, height, direction or balance can change loss and redistribute radiation. The only defensible 4 dB claim is one tied to a stated direction, reference, input-power convention and repeatable measurement or model.
Evidence boundary: a 4 dB result is incomplete without antenna dimensions, radial geometry, frequency, soil parameters, feedline route, choke impedance, calibrated field-strength or pattern data and an uncertainty budget. Without those conditions, the number cannot distinguish gain, realized gain, front-to-back ratio, RDF, mismatch improvement, feedline radiation or propagation variation.
Five Different Numbers Can All Be Reported as “4 dB”
A plot or signal report must say what its decibels compare. These are not interchangeable:
| Claim | What it compares | What 4 dB means |
|---|---|---|
| Directivity | Radiation intensity in a stated direction relative to the radiation intensity averaged over all directions. | The pattern concentrates about 2.51 times the average radiation intensity in that direction; dissipative loss and mismatch are excluded. |
| Gain | Directivity with radiation efficiency included, based on power accepted by the antenna. Absolute gain uses an isotropic reference; relative gain must name its reference, such as a dipole. | About 2.51 times the power density of the stated reference in that direction for the stated accepted-power convention and polarization. |
| Realized gain | Directional radiation referenced to incident power at the antenna port, so input mismatch is included. | The combined result of pattern, dissipative efficiency and mismatch at that reference plane. |
| Front-to-back or pattern difference | Two directions on the same pattern, often at one elevation angle. | One direction is 4 dB stronger than the other; neither direction is necessarily stronger than a reference antenna. |
| RDF | Forward gain minus the antenna's average gain under the stated integration convention. | A receive-pattern figure of merit, not automatically 4 dB more forward transmit gain or EIRP. |
IEEE 145-2025 is the active antenna-terminology standard. ITU-R BS.705-2 likewise separates directivity, which depends on pattern shape, from gain, which also accounts for efficiency. The compact relationship is:
G(θ,φ) = ηrad · D(θ,φ)
GdB = 10 log10(ηrad) + DdB
Grealized(θ,φ) = (1 − |Γ|²) · G(θ,φ)
Here ηrad is radiation efficiency, D is directivity and Γ is the reflection coefficient at the stated antenna port. Realized gain includes the mismatch factor at that port. A normalized pattern has deliberately thrown away absolute level, so it can show a 4 dB directional difference while hiding that the entire antenna became less efficient.
Radials Are Part of the RF Structure
A radial system provides the return path for a monopole, but “return path” does not mean “non-radiating.” Each radial carries RF current and interacts with the vertical conductor, soil, other radials, feedline, mast and nearby objects. With equal currents in a symmetric radial fan, horizontal field components largely cancel by symmetry. With only a few radials—or with unequal electrical lengths—that cancellation may be incomplete.
Changing two radials can therefore change several quantities at once:
- Ground and conductor loss: a denser or better-positioned return system can reduce power dissipated near the base. The gain can rise even if directivity barely changes.
- Current division: unequal radial currents can create an azimuth pattern and change the elevation pattern.
- Feedpoint impedance: the measured resistance contains radiation and loss contributions; a convenient 50 Ω reading does not by itself prove high efficiency.
- Common-mode current: if the feedline is not isolated, its outer surface and connected equipment can become another antenna element.
- Input mismatch: a change in SWR changes realized gain when incident transmitter power is the reference, even if gain based on accepted power is unchanged.
Rudy Severns, N6LF, measured HF vertical ground systems with the feedline isolated by a common-mode choke and found that radial geometry, radial count and missing sectors can affect signal level, impedance and pattern. His elevated-radial work also shows that few-radial systems can have unequal currents, pattern distortion and additional ground loss. Those results do not yield a universal correction factor; they show why the complete geometry and measurement boundary matter.
Can Two Radials Produce a Real 4 dB Improvement?
Yes, in principle—but not as a universal property of “two radials.” There are at least three physically different routes to a 4 dB result.
1. Lower loss
If directivity is nearly unchanged, a 4 dB gain improvement requires efficiency to increase by a factor of about 2.51. An antenna moving from 20% efficiency to roughly 50% efficiency would do that. A better ground system can produce a large improvement when the starting loss is severe, especially for an electrically short or heavily loaded vertical. It cannot raise efficiency above 100%, so the available improvement depends on the starting point.
2. Pattern redistribution
Asymmetric currents can increase radiation in one direction while reducing it elsewhere. That is directivity, and when efficiency is included it can be genuine directional gain. Calling every asymmetric result “only pattern skew” is too dismissive: pattern redistribution is precisely how directivity changes. The proper question is whether the absolute level in the favored direction increased at equal accepted power, and whether that direction and elevation are useful.
3. Mismatch or an uncontrolled feedline
If the test holds transmitter setting constant rather than accepted antenna power, an improved match can deliver more power to the structure and raise the received signal. That is a realized-gain or system-level change, not proof of increased radiation efficiency. If the feedline current also changes, the before and after tests are different multi-conductor antennas. The extra field is real, but it cannot be credited cleanly to the intended vertical and radial pair.
Do not infer efficiency from feedpoint resistance alone. At resonance, Rin includes radiation resistance transformed by the actual geometry plus conductor, loading-coil, ground, connection and other losses. A resistance moving toward 50 Ω may represent less loss, more loss, a changed current distribution or several effects together.
What N6LF’s Experiments Actually Support
N6LF’s controlled work is valuable because it separates claims that casual on-air comparisons often mix:
- In his experiments, a feedline choke was used to keep the test cable from silently becoming part of the radial system.
- Four elevated radials with well-controlled current division could be close to symmetric, while a one-radial worst-case model showed several decibels of front-to-back difference and lower average gain.
- With asymmetric elevated-radial lengths, both pattern distortion and lower gain could occur; the favored direction did not erase the additional loss.
- Ground-mounted systems with missing sectors were compared against a 360° radial fan, showing that placement and total system design matter more than a slogan about radial count.
These findings support neither “two radials always create gain” nor “two radials can never create gain.” They support a conditional engineering statement: few or asymmetric radials can alter current division, loss and pattern, so absolute gain has to be modeled or measured with the feedline and ground boundary controlled.
RDF Is Useful—Within Its Noise Model
Receiving Directivity Factor is commonly expressed in decibels as forward or peak gain minus average gain. Low-band references often average over the upper hemisphere, while other antenna texts use a full-sphere average, so the integration convention must be stated. RDF is useful when unwanted noise arrives over a broad set of directions. A deeper null or lower response away from the desired direction can improve RDF even when absolute forward gain is low.
RDF is not a universal prediction of signal-to-noise ratio. The result depends on how the average is integrated and on the real angular, polarization and spatial distribution of noise. One local noise source in a sidelobe can dominate; a diffuse-noise assumption may then be a poor model. Always state the pattern convention and examine the full three-dimensional pattern, not only one normalized azimuth cut.
ERP and EIRP Need the Correct Reference
The ITU Radio Regulations define EIRP with gain relative to isotropic and ERP with gain relative to a half-wave dipole:
EIRP(dBW) = Pantenna input(dBW) + G(dBi)
ERP(dBW) = Pantenna input(dBW) + G(dBd)
G(dBi) ≈ G(dBd) + 2.15 dB
Keep feedline loss and mismatch accounting consistent. Use gain with power accepted at the antenna input, or use realized gain with incident power at that same plane. Do not count mismatch twice. A bare “+4 dB” without dBi, dBd, direction, frequency and reference plane is not enough for an ERP or EIRP calculation.
A Test That Can Defend the Claim
- Define both antennas completely. Record frequency, radiator and radial dimensions, radial height and azimuth, wire size, soil conductivity and permittivity, feedline route, mast, nearby objects and weather.
- Choose the power convention. Compare at equal accepted power if the question is antenna gain. Compare at equal incident power if the question is realized gain. Measure at the antenna reference plane or correct calibrated line loss and mismatch.
- Control the external mode. Install a characterized choke at the intended current boundary and measure common-mode current at several feedline points. “Same choke” is not enough if the antenna impedance changes its stress or the cable route moves.
- Measure more than SWR. Record complex feedpoint impedance, accepted power, current distribution where practical and temperature of lossy loading or matching components.
- Measure an absolute pattern. Use a calibrated far-field range, a drone-based method with known uncertainty, or reciprocal links over multiple azimuth and elevation angles. A normalized plot alone cannot establish gain.
- Repeat and reverse. Alternate configurations several times to expose propagation drift, soil change and connector repeatability. Include a stable reference antenna or beacon path.
- Model the complete structure. Include finite-conductivity ground and the feedline/mast when they carry current. Compare modeled and measured impedance and pattern before trusting a gain number.
- Report uncertainty. A claimed 1 dB improvement is not persuasive if path, power, calibration and repeatability uncertainty total more than 1 dB.
How to Write the Result Without Ambiguity
| Weak statement | Engineering statement |
|---|---|
| “Two radials give 4 dB gain.” | “At 14.2 MHz, the two-radial configuration measured 4.0 dB higher realized gain at 25° elevation and 210° azimuth, relative to the stated four-radial reference, at equal incident power.” |
| “The pattern has 4 dB gain.” | “The normalized azimuth cut has 4 dB front-to-back ratio at 20° elevation; absolute gain was not measured.” |
| “RDF improved 4 dB.” | “RDF rose 4 dB under the stated three-dimensional averaging convention; forward absolute gain changed by X dB.” |
| “SWR proves it works better.” | “Mismatch loss fell by X dB at the antenna plane; radiation efficiency and pattern were evaluated separately.” |
Bottom Line
A radial change can produce a stronger field in one direction. It can do so through reduced loss, increased directivity, improved mismatch, changed feedline radiation—or a combination. That does not make the result imaginary; it makes the word “gain” conditional.
Measure the full system at a defined power plane, quantify uncertainty and name the metric actually obtained. A 4 dB change can be meaningful, but it is not a portable property of “two radials” unless the direction, reference, efficiency, mismatch, external feedline current, site and repeatability remain under control.
Primary and Experimental Sources
- IEEE 145-2025, Standard for Definitions of Terms for Antennas, active edition.
- Recommendation ITU-R BS.705-2, current HF antenna characteristics and gain/directivity treatment.
- ITU-R Radio Regulations reference formulas for EIRP, ERP and the dBi/dBd relationship.
- N6LF, Experimental Determination of Ground System Performance for HF Verticals, Part 1, including the controlled range and feedline-isolation method.
- N6LF, Part 3: Ground-Surface and Elevated Radials, including current-division and pattern comparisons.
- N6LF, Part 7: Ground Systems With Missing Sectors.
- N6LF, A Closer Look at Vertical Antennas With Elevated Ground Systems, Part 2, including asymmetric radial-current and gain results.
Mini-FAQ
- Can two radials really create 4 dB of gain? They can change real directional gain, but 4 dB is not a universal two-radial value. It must be tied to frequency, direction, geometry, ground, reference antenna, input-power convention and uncertainty.
- Is a 4 dB front-to-back ratio the same as 4 dB gain? No. Front-to-back compares two directions on one pattern. Gain compares absolute radiation in a stated direction with a defined reference for a defined input power.
- Can a better radial system increase gain without changing directivity? Yes. Lower ground or conductor loss raises radiation efficiency, and gain rises when efficiency rises even if pattern shape and directivity change very little.
- Does a 50 Ω feedpoint prove good efficiency? No. Feedpoint resistance includes radiation and multiple loss contributions, and it can also be affected by common-mode current. Match and efficiency must be evaluated separately.
- Does higher RDF mean higher transmitted field strength? Not necessarily. RDF compares forward response with average response under a stated convention. It can improve through rejection elsewhere without an increase in absolute forward gain.
- What is the minimum useful verification? Compare repeatable absolute field or gain at a defined power reference plane, characterize feedline common-mode current, record the complete geometry and ground, and report measurement uncertainty.