The Stacked Hybrid 4:1 Balun

Working Is Not the Same as Good Engineering

Updated May 23, 2026.

A hybrid 4:1 balun is, in practical terms, nothing more than a 4:1 voltage transformer followed by a 1:1 current choke. The transformer performs the impedance transformation. The choke tries to stop the outside of the coax shield from becoming part of the antenna system.

That concept is not wrong. In fact, separating impedance transformation from common-mode suppression is often the better engineering approach. A 4:1 transformer does one job. A current choke does another. When both jobs are required, combining them electrically can make sense.

The real question is not whether the hybrid concept can work. It can. The question is whether it is good RF engineering to stack the voltage transformer and the common-mode choke directly on top of each other, in the same physical plane, inside one compact assembly, especially when QRO operation is claimed or implied.

My answer is no.

Not because the design will never work. Many antenna systems work despite compromises. A poor antenna works. A warm balun works. A marginal choke works until the day it does not. For 100 W SSB, with a reasonably matched antenna and low duty cycle, a compact stacked assembly may be perfectly serviceable.

But “it works at 100 W SSB” is not the same as “this is sound QRO engineering.”

The topology is not the problem

The objection is not to the hybrid idea itself. A voltage transformer followed by a good current choke can be a very practical solution, especially in real-world antenna installations where the antenna is not high, symmetrical, and isolated from everything around it.

The objection is to the stacked physical implementation. When the transformer core and choke core are placed directly on top of each other, three avoidable risks are combined in one compact package:

• thermal coupling,
• electromagnetic coupling,
• high-voltage proximity.

Each of these may be tolerable at low power. Together, at QRO or under high-duty-cycle operation, they reduce engineering margin.

Thermal coupling turns two heat sources into one heat problem

A ferrite device in a transmitting balun is not a magic passive lump. It can dissipate real power.

In the 4:1 transformer, loss comes from winding resistance, leakage fields, dielectric loss, and core loss caused by the applied RF voltage and flux. In the 1:1 current choke, heat is produced when common-mode current flows through the resistive part of the choke impedance.

For the choke, the heating term is essentially:

Pchoke = ICM2 × RCM

That is an important point. A choke can have an impressive impedance curve on the bench and still heat badly if the antenna system forces enough common-mode current through it.

For the transformer, the magnetic flux density rises with applied voltage and falls with frequency. A simplified transformer relationship is:

Bmax ≈ VRMS / (4.44 × f × N × Ae)

That means the transformer is most stressed by high RF voltage, low frequency, too few turns, small core area, or reactive load conditions. Multiband off-center-fed antennas can create exactly those kinds of operating points on some bands.

When the transformer core and choke core are stacked directly together, the designer has turned two thermal systems into one shared thermal system. Heat from one core preheats the other. The contact faces lose airflow. The two cores radiate and conduct heat into each other. If the assembly is mounted in a small sealed plastic enclosure, the thermal resistance becomes worse again.

The basic thermal equation is simple:

ΔT = Ploss × Rθ

Stacking increases the effective thermal resistance. The same loss then produces a higher temperature rise.

This matters because ferrite materials are temperature-dependent. Permeability, loss, impedance, and voltage distribution do not remain perfectly stable as temperature rises. A ferrite device near its limit does not politely stay linear. It changes behavior, and that change often appears exactly when more margin is needed.

This is why a QRO balun should not be judged only by SWR. A balun can show acceptable SWR and still be cooking internally.

SSB hides problems that QRO and digital modes expose

A 100 W SSB station is a forgiving test case. The average power of normal speech is much lower than PEP. With light speech processing and normal voice duty cycle, the heating in a ferrite device may be far below what the PEP number suggests.

That changes with CW, RTTY, FT8, FM, AM carrier, long tuning periods, contest abuse, or QRO operation. Heating follows average power and duty cycle, not marketing optimism.

Going from 100 W to 1.5 kW is not a small increase. It is a 15:1 power increase. Voltage and current rise by the square root of that ratio:

√15 ≈ 3.87

Losses that follow current squared rise roughly with power. Voltage stress rises with the square root of power, but insulation failure, corona, and dielectric heating are not friendly linear effects.

That example is already significant, and it assumes a clean resistive 200 Ω load. Real antennas are not always that polite. With mismatch, reactive feedpoint impedance, feedline transformation, tuner interaction, OCFD imbalance, rain, nearby metal, or band-edge operation, the RF voltage at some points can be much higher.

So yes: 100 W SSB may be fine.

QRO is a different animal.

Stacking ignores a basic RF layout principle

The second problem is unintended coupling.

A toroid is better behaved than an air-core coil because much of its flux is confined in the core. That should be acknowledged. Two stacked toroids do not automatically become a perfect transformer by accident.

But real baluns are not ideal toroids. They have windings, leads, coax turns, leakage fields, electric fields, and parasitic capacitances. When two RF magnetic components are placed directly on top of each other, in the same plane, with minimal spacing, the layout maximizes whatever unintended coupling is available.

The physics is basic mutual inductance:

M = k × √(L1 × L2)

The induced voltage is:

v2 = M × di1/dt

The coupling coefficient depends strongly on geometry, distance, loop area, and orientation. This is why RF and analog layout practice normally tries to keep inductive components apart when they are not meant to be coupled. If they must be close, they are often rotated or arranged to reduce mutual coupling.

A stacked balun does the opposite. It places two magnetic RF structures close together and broadly aligned.

Again, this does not prove instant failure. But it removes a layer of engineering isolation. The 4:1 transformer is supposed to handle differential-mode impedance transformation. The 1:1 choke is supposed to add common-mode impedance. These are different jobs. Mechanically stacking them risks re-coupling the very functions we tried to separate electrically.

Capacitive coupling can bypass the choke

At HF, a few picofarads are not nothing.

The current choke is there to create a high impedance in the common-mode path. But if the physical layout creates stray capacitance around the choke, some RF common-mode voltage can bypass the intended choking impedance.

The capacitive reactance is:

XC = 1 / (2π × f × C)

At 28 MHz, the numbers are already very relevant:

That is not trivial. A tiny stray capacitance between the transformer side and the coax side of the choke can become significant on the higher HF bands. This is especially relevant when windings, coax turns, ferrite cores, high-voltage nodes, and enclosure walls are all physically close.

This is one of the hidden dangers of a very compact stacked layout. It may look neat. It may measure acceptably in one configuration. But the construction itself can create a parallel RF path around the choke.

A common-mode choke should not only have good impedance as a separate component. It should maintain that choking behavior after it is installed in the final enclosure, next to the actual transformer, with the actual wiring geometry.

High-voltage points deserve physical clearance

The third objection is voltage stress.

A 4:1 Ruthroff-style voltage transformer has points of significant RF voltage. In an off-center-fed dipole system, the feedpoint impedance and common-mode voltage can vary widely by band, height, feedline length, and surrounding objects.

That is exactly the environment where many OCFD baluns live.

Stacking a voltage transformer directly over or under a choke places high-voltage RF points close to other windings, coax shields, ferrite surfaces, mounting hardware, and enclosure walls. This increases the chance of:

• arcing or corona at QRO,
• dielectric heating,
• insulation stress,
• capacitive coupling into the choke,
• unpredictable common-mode behavior,
• failure under wet or dirty outdoor conditions.

High-voltage RF layout should use generous spacing, smooth routing, and predictable field geometry. A stacked ferrite sandwich is compact, but compact is not automatically good.

At QRO, compactness can become the enemy of voltage margin.

Low-power measurements do not prove high-power margin

A balun can pass a low-power VNA sweep and still be a poor high-power device.

A VNA sweep can tell us useful things: return loss, small-signal impedance, and sometimes insertion behavior. It does not tell us whether the ferrite runs hot after ten minutes at QRO. It does not tell us whether the choke voltage exceeds insulation margin on a particular band. It does not tell us whether the two stacked cores interact thermally after a contest exchange or a long digital transmission.

It also does not tell us whether a neat laboratory load represents the real antenna system. A small floating resistor is not the same thing as an OCFD in the air, with asymmetrical legs, wet surroundings, feedline transformation, and a shack full of alternate RF return paths.

That criticism applies strongly here. A stacked hybrid balun should not be defended only by saying “the SWR looks good” or “it worked in my field test.”

The correct QRO questions are different:

• How hot does each core get?
• Where is the highest RF voltage?
• What is the common-mode current on the coax?
• Does the choke impedance remain high after the transformer and choke are mounted together?
• Does the result change when the cores are separated?
• Does the result change when the cores are rotated?
• Does the enclosure layout add a bypass capacitance around the choke?
• Does the balun survive high-duty-cycle operation, not only SSB voice?

Without those answers, the design is not proven. It is only demonstrated.

The better engineering choice

The better solution is not complicated.

Keep the hybrid concept if desired, but do not stack the two cores directly. Give them physical separation. Give them thermal separation. Rotate them if practical. Keep high-voltage nodes away from the choke winding and away from enclosure walls. Use enough core volume for the transformer and enough choking impedance for the common-mode problem. Derate for the antenna and the duty cycle, not only for optimistic SSB PEP.

A more robust layout would use:

• a physically separate 4:1 transformer and 1:1 choke,
• spacing between the cores,
• orthogonal orientation where practical,
• larger cores or multiple cores for QRO,
• adequate airflow or thermal mass,
• proper creepage and clearance at RF high-voltage points,
• measured common-mode current in the final installation,
• high-power thermal testing, not only low-power VNA testing.

If the assembly must be compact, then the burden of proof becomes higher, not lower.

The fair challenge

The fair statement is this:

The hybrid 4:1 balun topology can work. A 4:1 voltage transformer followed by a good 1:1 current choke is often a sensible architecture.

The stacked implementation is what should be challenged.

For 100 W SSB, it may be fine. For QRO, high duty cycle, severe imbalance, wet outdoor service, or multiband OCFD use, stacking the voltage transformer and common-mode choke directly on top of each other is not best practice. It creates avoidable thermal coupling, avoidable electromagnetic coupling, and avoidable high-voltage proximity.

That is the core argument.

Not “it does not work.”

Rather:

Interested in more technical content? Subscribe to our updates for deep-dive RF articles and lab notes.

Questions or experiences to share? Feel free to contact RF.Guru for antenna and RF engineering support.