DL4ZAO’s “Balun-Workshop”: Where the Boundaries Matter

A reader pointed me to DL4ZAO’s 62-page “Balun-Workshop” handout, linked from his technical-download page. I read the current v2 PDF proposition by proposition. This is not a generic balun primer: it is my technical response to what that handout actually teaches.

The Handout’s Strongest Engineering Stays

Slides 4 and 10–13 give a useful coax model: the inner conductor and inside shield carry equal, opposite current for the intended TEM transmission mode, while the outside shield can carry a separate current. Roy Lewallen, W7EL, develops the same practical distinction in “Baluns: What They Do and How They Do It”. That separation is the heart of antenna-system troubleshooting.

Slides 12–20 also make an important functional distinction. A current balun or common-mode choke opposes net current that would otherwise use the feed line and surrounding installation as another path. A voltage-balun topology establishes a voltage relationship, but it does not by itself guarantee equal-and-opposite load currents when the load is asymmetric. The exact voltages are relative to the circuit reference created by that topology; “symmetrical to earth” is not a universal property of every device sold as a voltage balun.

Slides 46–55 correctly ask several separate questions: Does the device pass the wanted differential signal with low loss and acceptable reflection? Does it impede common-mode current? Does it maintain the intended current relationship under a declared load? These are different measurements. One attractive SWR trace cannot answer all three.

Wanted Dipole Current Is Not Feed-Line Common Mode

Slide 5 says that differential current on the line becomes common-mode current on the dipole and that this current produces radiation. I would not use that label in an antenna-feed discussion. At the dipole terminals, the wanted arm currents are equal in magnitude and oppositely directed with respect to the two feed terminals. The unwanted feed-line component is the non-cancelling current: outside-shield current on coax, or the vector sum of the conductor currents on a balanced line.

Radiation does not require calling the intended dipole current “common mode.” Antenna conductors radiate because their time-varying charge and current distribution produces fields; mode labels are defined by conductor currents and reference, not by whether radiation occurs. Keeping that vocabulary clean preserves the troubleshooting map: radiator-arm current is wanted, while current on the feed-line exterior, mast, control cable or station wiring may be an unintended part of the radiating and receiving structure.

The Outside Shield Is a Current Surface, Not a Fixed Cable Specification

Slide 4 describes the outer shield as a separate one-wire line with a lower velocity factor. The separate-current-surface picture is useful. The fixed line-parameter implication is not. The intended coax mode has geometry and dielectric confined between inner conductor and inside shield, so the manufacturer can specify its characteristic impedance and velocity factor. Outside-shield current closes through the antenna, mast, soil, protective-earth network, equipment chassis, other cables, nearby structures and displacement current. That external geometry defines its propagation and impedance.

Slide 10’s path from one dipole arm, along the shield exterior, to earth is therefore one possible circuit, not the only circuit. “Earth” may be part of the return, but a home station can also close the path through capacitance, bonding, mains protective earth, control wiring, tower, gutters or a second feed line. Protective-earth conductors must keep their safety function; RF mitigation is designed around the complete installed current path, never by removing protective bonding.

Current Balance, Voltage Balance and Transformation Need Topology

I agree with Mandel’s central warning on slides 18–20: a voltage-balun function is not automatically a common-mode choke. I would make the naming more explicit:

Separate cores, as shown in the hybrid examples on slides 40–42, can make the transformation and common-mode functions easier to develop and verify independently. “Two cores are required” is still too categorical. A Guanella transmission-line transformer can combine impedance transformation and current balance through its line connections and isolation; an autotransformer-style circuit behaves differently. The circuit, coupling and measured ports decide the function—not the number of cores or the word balun on a drawing.

The Geometric-Mean Rule Has a Narrow Home

Slides 13, 17, 46 and 53 repeatedly specify the wound line’s characteristic impedance as the geometric mean of input and output impedances. That relation belongs to particular matching structures—for example, a single quarter-wave transformer between real terminations. It is not a universal rule for every 1:1 choke, every electrical length or an impedance that varies behind a tuner.

In a 1:1 choke on a nominal 50 Ω system, a wound transmission path near 50 Ω can indeed help preserve differential return loss and insertion loss. That is a differential-mode requirement. Common-mode performance is a separate network involving the outside surfaces, ferrite, winding capacitance, enclosure and external return. As the winding becomes electrically longer, its differential characteristic impedance and phase matter more; neither a good differential match nor a high common-mode impedance proves the other.

The 50 Ω dB Table Is Valid Inside Its Fixture

Slides 48 and 61–62 place the choke as a series impedance between a 50 Ω generator and a 50 Ω receiver. The appendix explicitly derives the result for that circuit. Within that boundary, the arithmetic is useful: adding 200 Ω of purely series impedance gives about 9.5 dB more voltage attenuation than the direct 50 Ω-to-50 Ω connection, while 1 kΩ gives about 20.8 dB.

The limit is portability. In an antenna installation, Z_S, Z_L and Z_CM are complex and frequency-dependent, and the choke changes the standing-wave solution. A “20 dB choke” in the workshop fixture is not a promise of 20 dB less braid current in another network.

Report the measured vector impedance Z_CM = R + jX, the frequency range, fixture and assembly. Then verify current in the installation. Even impedance alone is not a power rating: common-mode voltage is approximately |I_CMZ_CM|, while ferrite dissipation contains the real component I_CM,RMS²R_CM. Power, duty cycle, temperature, material nonlinearity, winding insulation and enclosure cooling still need limits.

S21 Is a Measurement Coordinate, Not the Wrong Mode

The handout’s series-through test is not invalid merely because it uses transmission. Keysight’s E5061B impedance guidance explicitly supports reflection, series-through and shunt-through methods for different impedance ranges. The RF mode comes from how the DUT and return are connected, not from the letters S21.

For a choke test, configure the conductors so the fixture current represents the common-mode path, define the return, calibrate at the DUT plane, and characterize residual series impedance and shunt admittance. A full two-port conversion can be valuable. For an ideal floating series element with no shunt paths, Y₂₁ = −1/Z_series. In a real fixture, Y₂₁ also sees stray coupling, so −1/Y₂₁ is a model result, not a spell that removes fixture error.

Keysight requires open/short/load calibration and fixture compensation for its impedance methods. Rohde & Schwarz likewise treats fixture characterization and de-embedding as necessary when adapters and lead-ins sit between the VNA and DUT. At high impedance, a few picofarads can dominate; cable dressing, rotation and a second fixture geometry are useful sensitivity checks.

A Core Test Is Not a Completed-Choke Qualification

Slide 48 says that a simple wire on the core is enough to test choking action before winding the full assembly. That is fair as an early material-and-turns screen. It cannot establish the final upper-frequency response, differential match, voltage withstand or thermal margin. The complete winding adds conductor spacing, crossover geometry, end-to-end capacitance, connectors, enclosure coupling and heat flow.

Fair-Rite’s impedance-measurement note shows that even test-wire position and vector subtraction of the wire’s complex impedance affect a ferrite measurement. Its ferrite-selection guidance makes material choice frequency- and application-dependent. TDK’s ferrite application guidance similarly treats common-mode impedance as a material-and-geometry result rather than a universal mix ranking.

Placement Is an Installed-Network Decision

Slide 15 recommends putting the choke at a common-mode current maximum. A series impedance can have strong leverage where the unmodified network carries large current, but installing it changes that current distribution. The maximum also moves with band, line routing, mast, ground, nearby conductors and station connections.

The defensible method is measurement-led: map exterior current on the bands and operating states of concern, add or move one choke, then repeat the same measurements. The ARRL’s common-mode-current article likewise begins with measuring current and retesting after installation. Feedpoint, station entry or an intermediate position can each be appropriate, but none is an unconditional two-choke recipe.

My Carry-Forward From “Balun-Workshop”

I would keep Mandel’s inside-versus-outside coax model, the functional warning about voltage baluns, the separation of differential transfer from common-mode blocking, and the insistence that ferrite and winding geometry are compromises. I would attach a firmer label to every result:

• Name the mode: wanted differential current, unwanted current sum, or a specified transformer port relationship.
• Name the topology: current choke, voltage-transforming circuit, Guanella or autotransformer arrangement, and every reference connection.
• Name the fixture: source and load impedances, calibration plane, conductor configuration, parasitic model and uncertainty.
• Name the assembly: material, core geometry, winding, connectors, enclosure, frequency, power, waveform, duty cycle and temperature.
• Name the installation: antenna, feed line, mast, bonding, protective earth, other cables, nearby conductors and measured exterior current.

That preserves the workshop’s practical value without turning a classroom fixture or one topology into a universal balun rule.

Engineering References

• Günter Fred Mandel, DL4ZAO: Balun-Workshop v2
• Roy Lewallen, W7EL: Baluns—What They Do and How They Do It
• Keysight: E5061B Impedance Measurement Methods and Fixture Compensation
• Fair-Rite: Notes on Impedance Measurement
• TDK Electronics: Ferrites and Accessories—Application Notes
• Rohde & Schwarz: Accurate Test-Fixture Characterization and De-embedding
• ARRL QST: Common-Mode Current and Common-Mode Chokes

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