Inside RF.Guru 4:1 Ruthroff Transformers: Ratio, Balance and Power

Our 4:1 transformers do not start with a claim to have reinvented transformer theory. They build on C. L. Ruthroff’s transmission-line-transformer work and the practical design literature developed by Dr Jerry Sevick, W2FMI, notably Transmission Line Transformers. I want that lineage visible: established principles are an excellent foundation for a carefully built modern product.

At RF.Guru, the implementation combines a defined Würth Elektronik ferrite platform, PTFE-insulated conductors and model-specific single- or dual-core assemblies. My design point is that these are not independent items on a parts list. Core geometry affects the winding; insulation affects its spacing and electrical length; the mechanical build has to preserve that arrangement in service. That is why we develop and measure the finished assembly as one system.

Topology and Port Classification

Ruthroff described broadband transmission-line-transformer connections that can be arranged unbalanced-to-unbalanced, unbalanced-to-balanced or in other forms. “Voltage type” describes how conductor voltages are combined. It does not, by itself, tell us whether the finished ports are balanced.

RF.Guru’s current product pages classify the single-core 2.4 kW product as a 4:1 UNUN, and current antenna-system pages describe the dual-core 3.6 kW version as an integrated 4:1 UNUN. That is the appropriate name when both the source and load references are intentionally unbalanced.

The original foundation is C. L. Ruthroff’s 1959 paper, “Some Broad-Band Transformers”. Ruthroff showed both unbalanced-asymmetrical and unbalanced-symmetrical 4:1 forms. For circuit analysis, use the actual interconnection and reference nodes rather than the name alone.

What the Ideal 4:1 Ratio Means

For an ideal lossless transformer with voltage ratio n:

A resistive 200 Ω load can therefore appear as 50 Ω at the low-impedance port in the ideal model. More generally:

The approximation fails progressively as winding loss, leakage flux, finite magnetising impedance, line delay, distributed capacitance and the actual complex load become important. A load of 100 + j200 Ω does not become a perfect 50 Ω resistor; the ideal transformation alone would produce about 25 + j50 Ω.

For a simple 50-to-200 Ω transmission-line transformer, a useful starting value for winding-line impedance is the geometric mean:

The real multi-conductor winding does not have one immutable impedance. Conductor diameter, spacing, insulation thickness, ferrite proximity and bends through the core change its distributed parameters.

A Ruthroff Transformer Is Not Galvanic Isolation

The common 4:1 Ruthroff connection is autotransformer-like: source and load share an electrical conductor path. It transforms impedance but does not provide the safety isolation of a transformer with independent windings.

That shared path can be useful for static-charge drainage in some antenna systems, but it must not be represented as protective isolation. RF transformers do not replace protective earth, lightning bonding, surge protection or safe antenna-access procedures.

What Sets the Bandwidth

The low-frequency limit is strongly influenced by the shunt or magnetising impedance presented by the wound structure. If it is too low relative to the port resistance, current is diverted into magnetising and loss mechanisms rather than the intended load.

The high-frequency limit is set by line length, phase error, leakage inductance, parasitic capacitance, connector geometry and self-resonance. Ruthroff’s analysis shows why the physical transmission line must remain electrically short for the voltage-summing approximation to remain accurate.

Why a Defined Würth Ferrite Assembly Matters

We chose a defined Würth Elektronik ferrite platform so that the winding is developed around a component we can identify and source consistently. That is a production advantage, not a claim that Fair-Rite cores—or suitable cores sold by Amidon—cannot work. I do not want a visually similar replacement silently turning a finished transformer into a different RF design.

Cores that look identical can differ in:

• complex permeability μ′ and μ″ versus frequency;
• effective magnetic path length and cross-sectional area;
• A_L value and its tolerance;
• B-H behaviour, temperature dependence and Curie temperature;
• volume-loss density versus flux, frequency and temperature; and
• dimensions, coating and mechanical tolerance.

Würth’s official application note on magnetic-component loss separates winding loss from core loss. It concerns power inductors in switching supplies, not a test of these HF transformers; the useful distinction is that neither a permeability value nor a core-loss number accounts for the whole assembly.

Traceable ferrite sourcing supports repeatable magnetic geometry and material behaviour from unit to unit. RF.Guru qualifies the completed assembly—core set, winding, terminals, enclosure and thermal path—because the finished transformer, rather than a catalogue permeability value alone, determines ratio accuracy, loss and operating limits.

One Core Versus Two Cores

The 2.4 kW model uses one core; the 3.6 kW model uses two. The engineering reason for the additional ferrite is magnetic and thermal margin in the completed design. In a stacked arrangement linked by the same winding, both cores share the excitation: the effective magnetic cross-section and volume increase while path length remains similar. For the same turns and applied volts-per-hertz, this can reduce flux density and distribute loss over more material.

That does not make “two cores equals twice the power” a physical law. The result depends on:

• whether both cores actually share flux as intended;
• turn count and the voltage applied to the magnetising path;
• frequency-dependent ferrite loss;
• conductor current, proximity effect and solder-joint heating;
• voltage spacing and electric-field concentration;
• enclosure thermal resistance and ambient temperature; and
• transmission duty and time allowed for cooling.

Extra ferrite adds thermal mass, but steady-state temperature also depends on how heat reaches the enclosure and ambient air. Mating surfaces between stacked cores do not automatically double useful cooling area.

Why PTFE-Insulated Conductors Are Used

PTFE-insulated conductors are part of our build because the wire needs a substantial electrical barrier and controlled mechanical protection through the ferrite apertures. PTFE offers useful temperature and dielectric properties, while tinned stranded copper supports flexible routing and repeatable terminations. I want the electrical geometry to survive assembly, handling and outdoor service—not only look correct on a schematic.

The winding system is integrated with PV coating, nylon spacers, a polycarbonate enclosure, 316 stainless-steel hardware and compression sealing hardware. These material choices control conductor spacing, mechanical strain, moisture paths and long-term outdoor assembly integrity; additional PTFE sleeving is used at relevant high-stress locations in the separate 10 kW variants where required. That construction detail is not an extra power rating for the 4:1 models described here.

The insulation is also part of the RF geometry. Its thickness and dielectric constant change conductor spacing, winding-line impedance and distributed capacitance. PTFE does not create the 4:1 ratio, cancel common mode or establish a product voltage rating by itself.

PTFE can creep under pressure, stranded wire is not automatically Litz wire, and solder can wick into strands and create a rigid stress point. Bend radius, clamping, edge protection and strain relief remain essential.

What the Published Power Numbers Mean

RF.Guru publishes the following model-specific operating ratings for the relevant single- and dual-core platforms:

The single-core values are stated on the 2.4 kW 4:1 UNUN product page. The dual-core values are also listed on the EFOC17 system page. These are measured transformer-version limits under their stated operating conditions, not generic ratings for a loose ferrite core. An antenna package can have a lower limit because its supplied choke, cable or another component has a different operating envelope. In particular, the EFOC17 transformer table does not override the separate supplied-choke limit on that system page.

A device rating also does not authorise that transmitter power. The operator must obey the licence, band and RF-exposure rules that apply at the station.

The Matched-Load Voltages Are Already Serious

Even the ideal 50-to-200 Ω case illustrates why terminals, insulation and enclosure clearances matter:

These are ideal sinusoidal matched-load values, not worst-case internal stresses. Reactive loads and standing waves can place a voltage or current maximum at the transformer and can increase circulating energy. Never touch, open or adjust the assembly while transmitting; discharge and isolate the antenna system before service.

Insertion Loss Is Also Heat

The single-core product page specifies insertion loss below 0.1 dB; this is not a loss figure for every RF.Guru 4:1 assembly. To see why even a small loss matters at high power, take 0.1 dB as a worked example. At that boundary, transmitted power is about 97.7% of the incident power under the stated test conditions:

At 2.4 kW that difference is about 55 W; at 3.6 kW it is about 82 W. These are arithmetic examples, not measured heat outputs for either model. Some apparent insertion loss can include fixture mismatch rather than dissipation, so a valid test must separate mismatch and true loss. Conversely, a low-power VNA result does not prove that permeability, loss and temperature remain unchanged at high power.

Report insertion loss with frequency, source and load impedances, calibration plane, fixture uncertainty, temperature and power. For reactive antenna loads, repeat with representative complex impedances rather than only a 200 Ω resistor.

Voltage Balance Is Not Current Balance

A voltage-type transformer can create approximately equal and opposite terminal voltages relative to a stated reference. If the two load arms see different impedances, equal voltages do not produce equal and opposite currents:

When Z₁ differs from Z₂, the feedline exterior, mast, counterpoise and nearby conductors may carry part of the system current. That is why “BALUN” cannot be inferred from a nominal voltage split and why an UNUN is not a common-mode choke.

If outside-coax current must be controlled, add a 1:1 common-mode choke whose complex common-mode impedance is suitable across the bands and power involved. Place it according to the intentional return path and verify current at several coax positions. A choke is part of the installed antenna boundary, so adding it can change feed impedance and SWR.

From a Design to a Rated Assembly

Our ratings come from measurements of completed assemblies, not from counting cores. The useful engineering questions cover the circuit, electrical stress, heat and the conditions under which the result can be repeated:

1. Port schematic. Exact Ruthroff interconnection and BALUN/UNUN references.
2. Small-signal sweep. Complex S-parameters or impedance from the low to high band edge with fixture de-embedding.
3. Load matrix. Several resistance ratios and representative reactive loads, not only 200 + j0 Ω.
4. Power protocol. Frequency, waveform, PEP or average power, duty, key-down time, recovery time, ambient and enclosure orientation.
5. Thermal evidence. Core, winding, terminal and enclosure temperatures through equilibrium or the defined ICAS cycle.
6. Electrical stress. Terminal voltage, current and breakdown margin at the worst VSWR magnitude and phase.
7. Common-mode test. Impedance or current suppression measured separately from differential transformation.
8. Production tolerance. Results across component lots and completed units with acceptance limits.

This qualification keeps each watt rating attached to the tested model, frequency range, load and duty conditions. The product manual supplies the operating envelope; the installation then preserves the specified VSWR, cooling, current path and safety conditions.

Design Statements and Their Engineering Boundaries

Proven Principles, RF.Guru Construction

Ruthroff and Sevick remain part of this story because the established transformer principles are still useful. RF.Guru’s contribution here is the implementation: a selected ferrite platform, a winding and insulation system developed around it, and mechanical construction that preserves the electrical geometry. The single- and dual-core versions are deliberate assemblies, not the same schematic filled with arbitrary parts.

Core count is one part of the design. The published rating also accounts for frequency, complex load, VSWR phase, duty cycle, thermal environment, winding current, insulation stress and the completed assembly’s acceptance criteria. Port classification and common-mode behaviour remain separate from the nominal 4:1 ratio.

That is why I care about repeatability as much as the first good measurement. Use the version whose operating envelope fits the actual load and duty cycle, keep common-mode control explicit, and preserve the installation conditions that make the design work. The result is a practical transformer built on proven principles—not a new theory and not a promise that one box can match any antenna at any power.

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