End-Fed Antennas: Choke Impedance, Current and Placement
End-Fed Antennas: Choke Impedance, Current and Placement
An end-fed wire still needs a return path. The engineering job is to decide which conductors belong to that path, measure where exterior-feedline current flows, and place a qualified choke at the boundary the installed antenna actually needs.
“Use a high-CMR choke” sounds precise, but it mixes several different quantities. Common-mode rejection, common-mode impedance, fixture insertion loss and installed current reduction are not interchangeable. A useful design keeps each one attached to its circuit, reference plane, frequency and test conditions.
Joeri’s practical position: choose the impedance transformer from the measured antenna-side complex load and choose the choke from the installed exterior-current path. An UNUN plus a separately characterised choke keeps those two jobs visible when the port is intentionally unbalanced. The current map—not a label—sets the choke position.
One Wire to the Sky Is Still a Complete Circuit
An end-fed radiator presents one obvious wire at the matching network. The source still needs a second current path. Depending on the installation, that return can include a deliberate counterpoise or radial, a declared section of coax exterior, a mast, transformer and enclosure capacitance, ground coupling, bonding conductors, station wiring or several parallel branches.
If the return is not designed, the environment supplies one. Tom Rauch, W8JI, demonstrates in his measured end-fed half-wave analysis how matching-network capacitance and conductors attached to the feed system participate in current closure. That is why SWR can change with cable routing—and why an unchanged SWR does not prove the exterior current is small.
The outside of the coax is therefore not always an accidental radiator. A declared coax-exterior section can be the intentional return branch before a choke. Current that continues beyond that designed boundary is a different problem. The choke should end the intended antenna; it should not erase the only return conductor and leave the rest of the station to replace it.
CMR, CMRR, ZCM and Current Reduction Are Different
| Quantity | What it describes | What must accompany it |
|---|---|---|
| Common-mode rejection ratio, CMRR | The ratio of differential response to common-mode response in a defined differential system | Port definition, frequency, source impedances, loads and measurement method |
| Common-mode impedance, ZCM | The complex impedance inserted into a specified common-mode path | R + jX versus frequency, fixture, calibration plane and winding construction |
| Common-mode S21 | Transmission through a particular common-mode test fixture | Fixture source/load impedances, calibration and topology |
| Installed current reduction | The change in exterior current at defined positions after one installation change | Current-probe method, geometry, power, frequency and restored baseline |
Tektronix defines CMRR for a differential measurement system as the differential gain divided by common-mode gain, normally expressed as a ratio or in decibels. That is not the same physical quantity as a choke’s complex impedance in ohms.
A manufacturer or builder can define a common-mode rejection measurement for a particular choke fixture, but “20 dB CMR” by itself does not tell us the installed ZCM, current reduction, differential loss or power limit. Ask what was driven, what was terminated, where calibration stopped and over which frequencies the number applies.
The Simple Current-Divider Equation Has Conditions
If an unchanged exterior loop can be represented by a Thevenin source driving a lumped series path impedance Zpath, then adding a series choke gives:
I0 = Vth / Zpath
I1 = Vth / (Zpath + Zchoke)
I1 / I0 = Zpath / (Zpath + Zchoke)
That relation is useful only while the source, path and coupling remain the same. Moving a choke can change the effective source, redistribute current to a mast or another cable, and create a new resonance with the conductors on either side. The impedances are complex, so their magnitudes cannot simply be added as scalar “ohms of choking.”
This is also why no universal 2 kΩ, 5 kΩ or 10 kΩ target can guarantee a result. The needed impedance follows the installed source and load impedances, the allowable current and the frequencies of interest. More impedance can help, but its reactive part can participate in a resonance and its resistive part converts common-mode power into heat.
Place the First Choke at the Intended Antenna Boundary
Start by declaring the return branch:
- Separate counterpoise or radial completes the antenna at the transformer: a characterised choke at the feedpoint is a reasonable first candidate because the downstream coax exterior is not intended to radiate.
- A section of coax exterior is the deliberate return: place the candidate choke where that branch should end, then verify current on both sides and the complete installed pattern.
- The coax exterior is carrying current accidentally: first create or repair the intended return. Ferrite can otherwise move current into the mast, bonding network or equipment rather than solve the antenna.
- The installation is multiband: repeat the current map on every band. One physical branch length and one choke have different electrical behaviour across HF.
A fixed 0.05-wavelength distance can describe one proposed geometry, but it is not a general choke rule. The exterior mode does not automatically use the cable’s published internal differential-mode velocity factor, and the transformer, cable route, ground and nearby conductors change the current distribution.
Transformation and Choking Remain Separate Jobs
An impedance-transforming device is selected from the measured complex load and operating envelope. A common-mode choke is selected from the exterior-current loop. Combining those decisions into one “balun” label conceals which function was actually demonstrated.
In an intentionally unbalanced installation, Joeri’s practical default is an appropriate measured-load UNUN for the differential transformation plus a separately specified 1:1 choke at the measured common-mode boundary. This is not a universal 4:1 or 9:1 ratio, a fixed placement or proof that every end-fed geometry is unbalanced in the same way.
A suitable current balun remains valid for a genuinely balanced installed load. An integrated transformer-and-choke assembly can also work when its differential transfer, balance, common-mode impedance, voltage, current and thermal behaviour are qualified together under representative loads.
Shack and Equipment Chokes Must Earn Their Places
The building entry and the equipment end are useful measurement points, not mandatory choke positions. A second exterior-current loop can be excited after the antenna boundary through mast coupling, a parallel control cable, house wiring or the near field. A choke at that second boundary may help when a current scan demonstrates the path.
Installing chokes at the antenna, entry and transceiver by habit can also create three unexamined resonant cable sections. A choke near equipment may see the highest common-mode voltage if substantial current reaches that point. It can hide a poor antenna boundary, become hot, or shift current into USB, audio, Ethernet, power and bonding conductors.
A cable-entry choke does not replace shield bonding, protective earth, equipotential bonding, lightning protection or surge protection. IEC TR 61000-5-1:2023 treats earthing, bonding, cables, shielding, filtering, isolation and surge protection as coordinated but distinct EMC measures. Never defeat a safety conductor during an RF test.
Series Chokes Are a New Assembly
Two chokes in series can provide more impedance over part of a frequency range, but their standalone impedance curves do not guarantee the combined result. Interconnect length, mutual coupling, winding capacitance, cable bends, connector fields and the conductors on both sides can shift resonances.
Measure the finished series assembly in the intended physical arrangement. Record complex common-mode impedance and differential S-parameters at the same reference planes used to define the result. Then repeat the installed current scan. Stacking cores or products is not a substitute for that completed-assembly test.
Power Ratings Need Current, Voltage and Temperature
A transmitter-power label does not state what the choke experiences. Common-mode loss depends on the actual exterior current and ZCM. Differential cable and connector stress depend on line current, voltage, mismatch and frequency. Ferrite temperature depends on loss distribution, thermal mass, enclosure, airflow, ambient temperature, waveform and duty cycle.
Qualify the exact finished assembly under representative complex loads. Record:
- complex common-mode impedance versus frequency;
- differential insertion and return loss over the required load domain;
- common-mode voltage and current at the intended installation point;
- transmitter waveform, accepted power and duty cycle;
- temperature after thermal equilibrium; and
- the cold electrical response again after the powered cycle.
Core material, diameter, count and turns are construction inputs. None independently establishes broadband behaviour, saturation margin, insulation clearance, cable survival or an ICAS rating.
Receive Noise Requires an SNR Test
The coax exterior can carry local electric and magnetic fields toward the antenna, where imbalance converts part of them into the receiver’s differential path. A choke can reduce that route. It can also alter the intended antenna current or move the conversion point.
Do not declare success from a lower noise floor alone. Freeze receiver bandwidth, gain, attenuation and AGC; compare wanted signal and noise together; change one boundary; and restore the original arrangement. The useful receive result is improved signal-to-noise ratio with the intended antenna response retained.
A Measurement-Led Choke Plan
- Freeze the installation. Record wire, return conductors, coax route, mast, bonds and every parallel cable.
- Mark the intended boundary. State which exterior section belongs to the antenna and where it should end.
- Measure the differential load. Save R + jX at one calibrated antenna-side plane on every operating band.
- Scan the baseline current. Use a repeatable probe orientation at several marked positions from feedpoint to equipment.
- Characterise the candidate choke. Record fixture, calibration, ZCM, differential transfer, exact construction and frequency range.
- Change one boundary. Insert or move one choke without changing wire, cable route, tuner, grounding or power.
- Repeat the full scan. Look for current reduction, redistribution and a new maximum on both sides of the choke.
- Restore and repeat. Use A/B/A trials to expose cable movement, heating, receiver drift and propagation changes.
- Verify the wanted outcome. Compare current, accepted-power field data, receiver SNR and station symptoms as the claim requires.
- Run the powered qualification. Check voltage, current, temperature and cold post-test performance at the required duty cycle.
Larry Lamano, WA0QZY, demonstrates why several current-probe positions matter: exterior current can have maxima and minima along the cable. A single reading beside the choke is not a map.
Primary Technical Sources
- Roy Lewallen, W7EL — Baluns: What They Do and How They Do It: foundational current analysis and experiments on feedline imbalance, exterior current and current-balun action.
- Larry Lamano, WA0QZY — Common-Mode Current and Common-Mode Chokes: multi-position current measurement and frequency-dependent choke impedance.
- Tom Rauch, W8JI — End-Fed Half-Wave Matching-System Analysis: measured and modelled end-fed return paths, matching loss and SWR-test limitations.
- Tektronix — ABCs of Probes: CMRR definition, decibel relation and frequency/test-system dependence.
- Fair-Rite Products — 17th-edition technical catalogue: manufacturer data for complex, frequency-dependent ferrite behaviour.
- IEC TR 61000-5-1:2023: coordinated EMC installation and mitigation measures.
Joeri’s Bottom Line
I want the choke to define a measured electrical boundary, not decorate a cable. “High CMR,” “QRO” and “wideband” are incomplete until I know the ports, fixture, complex impedance, differential loss, installed current and powered temperature.
Give the end-fed wire a deliberate return. Use the transformer the measured load requires. Put a separately characterised choke where the intended return should end. Add another only when another measured path justifies it. That is how the feedline becomes a feedline again.
Mini-FAQ
- Is a high-CMR label enough to choose an antenna choke? No. Ask for the defined ports and fixture, frequency range, complex common-mode impedance, differential loss, installed current reduction and powered stress limits.
- How much common-mode impedance is enough? Enough to meet the allowable current in the measured exterior loop without unacceptable differential loss, voltage or heat. There is no installation-independent ohm target.
- Should the first choke always be beside the end-fed transformer? No. It belongs where the intended return branch should end. If coax exterior is a declared return conductor, that boundary can be away from the transformer.
- Do I automatically need chokes at the antenna, entry and radio? No. Each choke should control a demonstrated current path. Habitual placement can create unexamined cable sections and move current into other conductors.
- Can two series chokes simply have their impedance values added? Not safely from separate plots. Interconnects, coupling and capacitance can change the combined response, so measure the complete physical assembly.
- Does a lower receive noise floor prove the choke helped? Not by itself. Compare wanted signal and noise with fixed receiver settings and a restored baseline; the relevant outcome is SNR with the intended response retained.