Transformer Losses: A Reality Check

RF UNUN Transformers Explained: 1:2, 1:4, 1:9 and EFHW UNUNs

Broadband RF UNUN transformers are some of the most misunderstood parts in amateur radio. A 1:2, 1:4, 1:9, 49:1, or 70:1 UNUN may look like a simple ferrite core with wire on it, but electrically it is a compromise between impedance transformation, magnetic flux, leakage inductance, winding capacitance, common-mode current, voltage stress, and heat.

The transformer ratio alone does not make an antenna efficient. It only transforms impedance. Real efficiency depends on whether the ferrite material, core size, winding geometry, voltage stress, current stress, and antenna load are suitable for the frequency and power level being used.

Impedance Transformation and Choking Are Different Jobs

An UNUN transformer should primarily be judged as an impedance transformer. Its job is to transform one unbalanced impedance into another unbalanced impedance with the lowest practical loss over the intended frequency range.

Common-mode current control is a separate system problem. In many real amateur-radio installations, antennas are mounted low, close to houses, masts, gutters, metalwork, trees, fences, and other objects. Under those conditions, forcing one transformer to perform impedance transformation and common-mode suppression at the same time often creates a compromise.

A low SWR does not prove that common-mode current is controlled. It only tells you what impedance the transmitter sees at that moment.

Common RF UNUN Transformer Ratios

The table below shows how common HF UNUN ratios are typically used. Ratios are shown as impedance ratios, which is how most amateur-radio UNUN transformers are advertised.

A transformer ratio should be selected from the antenna impedance, not from habit. A 1:9 UNUN is not “better” than a 1:4 UNUN; it simply transforms impedance by a larger factor.

Why Transformer Loss Happens

Transformer loss is not one single mechanism. In practical HF designs, several effects add together:

• Core loss: RF energy is dissipated as heat inside the ferrite material.
• Copper loss: winding resistance increases with current, skin effect, and duty cycle.
• Leakage inductance: not all magnetic flux couples perfectly between windings.
• Distributed capacitance: the winding itself becomes a small RF structure, especially at higher HF frequencies.
• Self-resonance: the winding inductance and capacitance form resonant behaviour that can disturb the intended transformer action.
• Common-mode current: feedline current can make measurements look better or worse than the real antenna system.
• Reactive antenna loads: a transformer feeding a highly reactive impedance may heat even when the SWR looks acceptable after a tuner.

Realistic Loss Expectations

The numbers below are practical engineering estimates for well-built HF UNUN transformers using appropriately large ferrite cores and sensible winding layouts. They are not guarantees for every commercial or homebuilt transformer.

Loss in dB is calculated from efficiency: Loss dB = 10 log10(1 / efficiency). A 10% power loss is about 0.46 dB. A 30% power loss is about 1.5 dB.

1:2 UNUNs: Small Ratio, Low Drama

A 1:2 impedance UNUN is often overlooked because it is not as dramatic as a 1:9 or 49:1 design. Technically, however, it is one of the easier ratios to make efficient. The voltage step-up is modest, the number of turns can remain low, and parasitic capacitance is usually easier to control.

Use a 1:2 UNUN when the antenna or system impedance is genuinely near twice the transmitter impedance. It is not a universal matching device, but in the right place it can be cleaner and more efficient than forcing a tuner to handle the entire mismatch.

1:4 UNUNs: The Practical Workhorse

The 1:4 UNUN is one of the most useful HF ratios. It transforms 50 Ω to approximately 200 Ω, which is useful when the antenna system naturally presents a load in that region.

Its advantage is that the impedance ratio is still moderate. Compared with a 1:9 or 49:1 transformer, the voltage step-up is lower, the winding structure is usually simpler, and the transformer can often be made efficient over a wider HF range.

However, a 1:4 UNUN should not be asked to solve every problem in the antenna system. If the feedline carries common-mode current, that should be handled with a dedicated current choke in the correct place, not hidden inside the impedance transformer.

1:9 UNUNs: Useful, But Not Magic

The 1:9 UNUN is popular for end-fed random wires and non-resonant wire antennas. It transforms 50 Ω to approximately 450 Ω, which can bring many high-impedance wire antennas closer to the tuning range of an ATU.

However, a 1:9 UNUN does not remove reactance. If the antenna impedance is 450 Ω - j1200 Ω, the transformer still has to process that reactive energy. The tuner may hide the mismatch from the transmitter, but the ferrite core still sees the RF stress.

This is why random-wire systems need:

• A sensible wire length that avoids extreme impedances on the intended bands
• A defined counterpoise or return path
• A proper current choke at the correct point in the system
• A transformer with enough ferrite volume for the intended power level

A 1:9 UNUN is a matching aid, not an antenna design by itself.

EFHW UNUNs: 49:1, 64:1, 68:1 and 70:1

End-Fed Half-Wave antennas are a special case because the feedpoint impedance of a half-wave wire can be very high. This is why EFHW systems often use 49:1, 64:1, 68:1, or 70:1 impedance transformers.

These high ratios are convenient, but they are also demanding. A 100 W transmitter feeding a high-ratio EFHW transformer can produce several hundred volts of RF at the antenna terminal. At higher power, the voltage stress becomes serious. Winding layout, insulation, ferrite material, core size, and mechanical construction all become part of the design.

EFHW 49:1 UNUNs on 17–10 m: Why the Upper Bands Are Difficult

A 49:1 EFHW UNUN can work very well on its intended harmonic bands when the antenna length, ferrite material, core size, and winding layout are correct. However, the upper HF bands expose transformer limitations much more quickly than 40 m or 20 m do.

The problem is not simply the 49:1 ratio. The problem is that a high-ratio transformer requires a winding structure with significant voltage step-up, and that winding structure is no longer electrically “small” at higher HF frequencies.

At 17 m, 15 m, 12 m, and 10 m, the transformer is affected by several parasitic effects:

• Inter-winding capacitance between turns and layers
• Leakage inductance from imperfect magnetic coupling
• Distributed capacitance along the high-impedance winding
• Self-resonance of the winding structure
• Increased RF voltage stress at the antenna terminal
• Greater sensitivity to layout, lead length, grounding, enclosure geometry, and connector placement

On the upper bands, a 49:1 UNUN may show a clean match on 40 m and 20 m but become increasingly reactive or lossy higher up. The SWR problem is only the visible symptom. The deeper issue is that the transformer’s winding structure may be approaching its practical broadband limit.

This is especially important on 17 m and 12 m. These bands are not natural harmonic resonances for many common EFHW wire lengths, so the antenna impedance itself may not be close to the value the 49:1 transformer expects. In that case, the transformer is being asked to process both a difficult impedance and a high-frequency winding environment.

On 10 m, the antenna may be closer to a harmonic resonance, but the transformer is operating at a much higher frequency. The result can still be difficult: winding capacitance, leakage inductance, and self-resonance can dominate the behaviour even when the antenna wire appears electrically reasonable.

A low SWR on the upper HF bands does not automatically mean low transformer loss. The transformer may still be dissipating power as heat or relying on common-mode current to complete the system.

For this reason, a 49:1 EFHW UNUN should be evaluated as a real RF transformer, not just as a turns-ratio device. The important questions are: does it remain stable across the intended bands, does it stay cool under real power and duty cycle, and does it avoid forcing the feedline shield to become an unintended part of the antenna?

In a clean transformer-only design, the goal is to make the winding geometry, ferrite choice, core size, and physical layout good enough that no extra lumped-element correction is needed. If the transformer only works because extra components are added around it, then the system should be understood as a matching network, not as a pure broadband transformer.

Why EFHW Inverted-L Antennas Can Still Work Well

An EFHW inverted-L can be a strong DX performer because the vertical section contributes useful low-angle radiation. Unlike a short ground-mounted vertical with a poor radial system, an EFHW inverted-L can place more current in the radiating wire and less in lossy soil paths.

That advantage only holds when the transformer is efficient and the feedline current is controlled. Once transformer loss becomes excessive, the antenna may still tune, but effective radiated power drops and the transformer becomes a heater.

A good match is not the same as good efficiency. SWR tells you how the transmitter sees the system; it does not tell you how much power is being turned into heat inside the transformer.

Common Measurement Mistakes

Many transformer “efficiency tests” are misleading because they measure a simplified bench condition rather than the real RF system. Common problems include:

• Testing into a pure resistor while the real antenna is reactive
• Ignoring common-mode current on the coax
• Using back-to-back transformer tests without understanding that two transformers and two different operating directions are involved
• Assuming that low SWR means low transformer loss
• Measuring only at low power and extrapolating to high power
• Ignoring core temperature rise during long transmissions or digital modes

A more realistic evaluation combines VNA measurements, forward/reflected power under real antenna load, current measurement, and thermal observation. If the ferrite core heats significantly, power is being dissipated in the transformer instead of being radiated.

Design Rules That Actually Matter

• Use the lowest ratio that solves the impedance problem. Higher ratio means higher voltage and usually more parasitic effects.
• Choose the ferrite material for the frequency range. Low-band and high-band HF requirements are not identical.
• Use enough core area for the power level. More power requires more thermal and magnetic margin.
• Keep the winding layout clean. Lead length, turn spacing, crossing wires, and enclosure layout matter at HF.
• Separate impedance transformation from choking. Let the UNUN transform impedance, and use a dedicated current choke where common-mode control is needed.
• Do not trust SWR alone. SWR does not directly measure transformer efficiency.
• Measure heat under real duty cycle. SSB, CW, FT8, RTTY, and AM stress transformers very differently.

Summary

RF UNUN transformers are not magic boxes. A 1:2 UNUN is usually easy to make efficient. A 1:4 UNUN is a practical workhorse when the antenna impedance is genuinely near the transformed value. A 1:9 UNUN is useful for tuner-assisted wire antennas, but it does not eliminate the need for a counterpoise, current choke, or sensible wire length.

EFHW UNUNs remain valuable, especially 49:1 designs for compact resonant wire antennas and larger-ratio designs for low-band EFHW or inverted-L systems. But their high impedance ratios create higher RF voltage, more turns, more distributed capacitance, and greater core-loss risk. This becomes especially critical on 17–10 m, where the winding structure itself can dominate the transformer behaviour. A well-designed 49:1 transformer should remain a transformer, not a hidden lumped-element matching network.

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

Questions or experiences to share? Contact RF.Guru.