Counterpoise and Return Paths with 4:1 and 9:1 UNUNs

The word counterpoise is often used as if it meant a small optional wire that steadies the match while the “real” antenna radiates alone. That picture is too tidy. At an end-fed or intentionally unbalanced feedpoint, current leaving one terminal must return through the other terminal and the conductors connected to it. The practical question is not whether a return path exists, but which conductors form it on each band.

The Ratio Does Not Build the Other Half of the Circuit

The labels 4:1 and 9:1 describe nominal impedance-transformation ratios under the transformer's defined port and mode conditions. In the ideal direction, a 4:1 device might transform a 200-ohm resistive load toward 50 ohms, while a 9:1 device might transform 450 ohms toward 50 ohms. Reverse the device and the direction of the impedance transformation reverses.

An antenna presents a frequency-dependent complex load, R + jX, rather than a fixed resistor. The nominal ratio scales the impedance according to the actual transformer topology and operating conditions; it does not cancel arbitrary reactance, guarantee 50 ohms or prove low loss across HF. Winding inductance, leakage, interwinding capacitance, transmission-line impedance, core behaviour, fixture parasitics and the connected load all limit the real result.

Most importantly, the transformer ratio does not manufacture a missing conductor. The antenna current still needs a closed electromagnetic circuit through a deliberate wire, coax exterior, mast, capacitance to the surroundings, ground-coupled conductor, bonding network, station wiring—or a combination of these.

A Counterpoise Is an Intentional Return Structure

In this context, a counterpoise is a conductor or conductor network deliberately assigned to the return side of the feedpoint. It need not be connected directly to earth. Calling it a “synthetic ground” can be misleading because an elevated counterpoise, soil connection, protective-earth bond and lightning-protection system have different electrical and safety functions.

At the feedpoint, Kirchhoff's current law still applies. Current delivered into the radiator terminal is balanced by current leaving through the return terminal. If the return terminal reaches several conductors, that current divides among their complex impedances. A short counterpoise does not automatically carry only a small current; it may carry substantial terminal current while its current varies along the wire and shares with other branches.

Every participating conductor can contribute to the radiated field. Its contribution depends on current magnitude and phase along its length, position, orientation and surroundings. The long wire may dominate a particular direction, but that is an installed pattern result—not a privilege granted by calling it the “hot” wire.

The Coax Exterior Can Be Part of the Antenna

Coax supports two different current descriptions. The wanted transmission-line mode has equal and opposite currents on the centre conductor and the inner surface of the shield. Current on the shield's exterior is a common-mode current relative to the surrounding environment. That exterior surface is available as another conductor in the antenna system.

When the coax section between transformer and choke is deliberately used as a return branch, document it as part of the radiating geometry. Its route, length, height, proximity to ground and structures, and the choke position all affect its impedance and current. It is not “hidden” electrically merely because the conductor is the outside of a cable.

Without a defined boundary, current may continue beyond the intended section onto the remaining feedline, equipment bonds and station wiring. That can change pattern, couple transmit RF into equipment, or carry local noise toward the antenna on receive. None of those outcomes follows from the word UNUN alone.

A Choke Creates a Finite Boundary, Not a Perfect Wall

A common-mode choke raises the impedance of the feedline-exterior path at the place where it is installed. Its impedance is complex and frequency-dependent. It also experiences voltage, current and heat that depend on the connected common-mode source and load.

The choke position therefore defines the end of the intended return section only approximately. A fixed fraction of a free-space wavelength or a fixed distance in metres is not universal, especially on a multiband antenna. The exterior-wave velocity and electrical length depend on the installed conductor and its surroundings, while the choke has finite impedance and parasitic coupling can bypass it.

Choose the location from the intended current geometry, then map exterior current on both sides of the choke across every required band. Reroute or alter the candidate counterpoise and coax section to see whether the result is repeatable. A low SWR cannot show that the choke established the desired boundary.

Separate Transformation from Common-Mode Control

I prefer to keep the two jobs explicit. The UNUN is selected for the measured differential load and transformation direction. The separate choke is selected for the common-mode impedance, current, voltage, frequency range and dissipation required at its installed position. That separation makes each function easier to measure and change.

This is a practical default, not a claim that an UNUN is universally superior. A genuinely balanced two-terminal antenna can be served correctly by a current balun when the installed terminal currents remain equal and opposite, common-mode conversion is acceptably low, and the network survives the measured complex load and operating stress.

Real amateur installations often disturb ideal balance through unequal wire geometry, feedline routing, support hardware, nearby conductors, capacitance to the surroundings and station bonding. For an intentionally unbalanced transformed port, an UNUN plus a separately specified choke covers those two jobs directly. The pair can also form a valid hybrid interface to a balanced load only when its arrangement, common-mode isolation, differential behaviour and thermal margin are verified as one network. Placing two boxes in series does not prove the result.

An Off-Centre Feed Is Not Automatically Unbalanced

A conventional off-centre-fed dipole still has two radiator conductors. Its feedpoint position changes the differential impedance and current distribution, but it does not by itself prove that the installed port should be treated as an unbalanced one-wire system. A current balun may be the right interface when the complete installation is genuinely balanced.

An end-fed wire with an intentionally different return branch is a different architecture. The radiator and return branch do not have the same geometry or environmental impedance, so the port is deliberately unbalanced. Use the current-path drawing and measurements rather than grouping both systems under one antenna label.

Counterpoise Length and Orientation Are Design Variables

There is no universal 2–5 m counterpoise and no universal fraction of a wavelength. Length, diameter, height, route, termination, ground proximity and nearby conductors set the branch impedance. The current distribution changes by band, and a length that is modest on a lower band can become electrically important or resonant higher in frequency.

Dropping the wire vertically, running it horizontally, placing it on the ground or routing it along a structure does not produce the same antenna. Each option changes capacitance, loss, induced current and radiated field. Soil contact can add loss and weather sensitivity. A branch near metal can couple into a larger reradiating structure. Orientation can change impedance and pattern together.

Start from a full geometry model, then test candidate arrangements at the installation. Do not choose the counterpoise solely by the SWR that looks most convenient.

Grounding and Bonding Keep Their Safety Jobs

A protective-earth conductor, lightning bond and RF counterpoise are not interchangeable labels. Safety and lightning conductors must remain installed and bonded as required by the applicable electrical code and lightning-protection design. Never disconnect a required safety bond to make an RF measurement look cleaner, and do not create an isolated earth electrode as an improvised antenna return.

Those conductors can still carry RF if the installation excites them. Include them in the current-path model and measure where accessible, but solve unwanted RF with an engineered return path, common-mode boundary and bonding layout that preserves every safety function.

Loss and Pattern Must Be Measured Separately from Match

A return-path change can move feed impedance, increase loss or reshape the pattern. These changes can happen together, but one does not prove another. A lower SWR can result from a better match, a different transformation or additional dissipation. It is not an efficiency or balance measurement.

At a minimum, compare:

• complex antenna impedance at a declared reference plane before selecting the ratio;
• transformer insertion loss with representative complex loads, not only a convenient resistor;
• differential and common-mode response, including mode conversion where a balanced port is claimed;
• current magnitude and phase on the radiator, intentional counterpoise, coax exterior, mast and accessible bonds;
• transformer and choke temperature at the stated power, waveform, mismatch, duty cycle, ambient and test duration; and
• azimuth and elevation patterns, or controlled field measurements, when a radiation claim matters.

Keysight's mixed-mode measurement framework distinguishes differential and common-mode stimulus and response. That is the right language for verifying a balanced-interface claim. LLNL's Numerical Electromagnetics Code can model currents, fields, ground, wires and transmission lines, but only when the return conductors and nearby structures are included in the geometry.

Commission the Current Path You Intended

Before normal-power operation, draw every plausible return branch. Record transformer terminals, counterpoise, coax route, choke location, mast, bonds and station wiring. Measure impedance and exterior current at low power on every operating band. Then change one branch at a time: move the choke, reroute the coax, alter the counterpoise or temporarily change a non-safety connection.

A useful design remains reasonably stable when ordinary routing and environmental changes stay inside its declared limits. If the match or current map changes dramatically when a cable is moved, that cable was already an important part of the antenna. Document it or redesign the boundary.

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