Why We Won’t Build General-Purpose 100 W Transformers

A 100 W transceiver does not create a 100 W transformer requirement. It creates a transformer requirement that must survive 100 W after the antenna, frequency, impedance, SWR, duty cycle, feed line, weather, and installation have had their say.

That distinction is why RF.Guru does not offer compact 100 W impedance-matching transformers for the 4:1, 9:1, 49:1, 68:1, and 70:1 applications discussed here. Most of these designs are rated above 2 kW ICAS (Intermittent Commercial and Amateur Service). The large number is not there because we expect every customer to run an amplifier. It is there because the transformer must remain efficient and reliable when a real antenna stops looking like the convenient resistor used on the test bench.

The radio’s wattmeter tells only the beginning of the story

Into a perfect 50 Ω load, 100 W corresponds to about 70.7 V RMS and 1.41 A RMS. Those figures look harmless. An impedance transformer deliberately changes them.

The table below shows the ideal high-impedance-side voltage when 100 W is accepted by a purely resistive, correctly matched load. It does not include reactance, mismatch, winding parasitics, or installation effects.

A high-ratio end-fed transformer can therefore be operating close to a kilovolt peak at only 100 W, even when the antenna presents exactly the intended resistance. Once the installed impedance moves, the transformer can see considerably more voltage or current.

These are deliberately simplified resistive examples. A real multiband antenna also presents reactance, frequency-dependent current distribution, standing waves, and parasitic effects. Those do not make the design problem easier.

A tuner controls impedance only at its own reference plane

A tuner can make its transmitter-side port look like 50 Ω. It does not make every component elsewhere in the system see 50 Ω.

With a shack-side tuner, the feed line, transformer, balun, and antenna-side wiring still experience their local, line-transformed complex impedance. A 1:1 reading at the radio can therefore coexist with high voltage, high current, or standing waves elsewhere. A tuner placed at or beyond the stressed component can change that component’s load, so placement matters.

This is especially important for antennas that are intentionally used on several bands. Wire height, routing, nearby buildings, gutters, trees, soil, radials, coax placement, and even wet weather can change the feed-point impedance. A transformer designed with no margin may work beautifully in one garden and run hot in the next.

The power rating is a conditional limit, not one universal number

A responsible RF power rating needs conditions. At minimum, it depends on:

• Service: ICAS (Intermittent Commercial and Amateur Service) is not the same thermal problem as CCS (Continuous Commercial Service) or a long digital transmission.
• Duty cycle: FT8, FT4, RTTY, AM, FM, and long key-down tests allow heat to accumulate.
• SWR and load: the local complex impedance can increase winding voltage and/or conductor current. On the low-frequency side, inadequate magnetizing reactance can add magnetizing current and loss.
• Frequency: core loss, magnetizing inductance, leakage inductance, and distributed capacitance change across HF.
• Temperature and enclosure: a sealed black box in summer sun cannot reject heat like an open bench assembly.
• Connectors and insulation: the core is not the only possible limit; spacing, dielectric strength, terminals, and connectors also matter.

Our 3.6 kW dual-core 4:1 UNUN is a useful example. Its headline rating is 3.6 kW PEP ICAS below 2:1 VSWR. The same product is rated 1.8 kW CCS and 1.25 kW for FT8/FT4 under a tighter mismatch limit. Above 3:1 VSWR, the page recommends staying below 900 W PEP; above 5:1, below 450 W PEP.

That is derating in practice. A “3.6 kW transformer” is not a promise of 3.6 kW on every band, at every duty cycle, into any load. Conversely, using it at 100 W is not absurd overkill. It places normal operation well inside the electrical and thermal envelope instead of balancing it on the label.

A transformer labeled for exactly 100 W into a favorable resistive load can fall below a 100 W safe operating limit as soon as digital duty cycle, mismatch, reactance, temperature, or an unfavorable band forces derating. That is the central commercial problem: the product would stop being a dependable “100 W transformer” precisely when the customer most needs its margin.

Why our common ratios start in the kilowatt class

The different ratios solve different antenna problems, but they share one design requirement: they must remain useful after installation changes the ideal impedance.

• 4:1: our EFOC17 antenna system uses an integrated low-loss 4:1 UNUN and comes in 2.4 kW single-core and 3.6 kW dual-core ICAS versions below 2:1 VSWR. Their FT8/FT4 limits are 850 W and 1.25 kW respectively, below 1.5:1 VSWR. The same ratio is useful in multiband rigid vertical and loop systems where feed impedance varies with the installation.
• 9:1: our non-resonant EFLW51 is a 2.4 kW ICAS / 1.2 kW CCS / 600 W digital antenna system with an integrated 9:1 UNUN. It needs a tuner on most bands and is typically below 4:1 VSWR. This is exactly the kind of variable load for which a generic perfect-load 100 W transformer would provide too little margin.
• 49:1: our EFHW4020 antenna system is rated 3 kW ICAS, 1.5 kW CCS, and 1 kW FT8/FT4 below 2:1 VSWR, with an integrated 49:1 UNUN. Its current product page derates the system below 100 W PEP above 5:1 VSWR—a particularly clear demonstration that the headline rating is conditional.
• 68:1 and 70:1: our EFHW16080 and EFHW8040 are 4 kW ICAS / 2 kW CCS / 1.5 kW digital antenna systems with integrated transformers selected for their low-band Inverted-L geometries. The current pages derate them above their stated VSWR limits; above 5:1, the specified PEP limits fall below 1 kW for the 68:1 system and below 100 W for the 70:1 system.

You can explore the same design approach across our end-fed wire antennas, rigid HF verticals and V-dipoles, rigid HF loops, and dipoles, doublets, and Inverted-V wire antennas.

A 100 W antenna rating does not imply a 100 W transformer

Our TermiLoop is a useful example. It is a terminated antenna fitted with a 2.4 kW 4:1 UNUN, yet the complete system is rated for 100 W FT8/FT4 and up to 200 W SSB/CW because the terminating dummy load—not the transformer—sets that limit.

This is why the rating of an antenna product must not be mistaken for the rating of every component inside it. The transformer retains substantial magnetic, electrical, and thermal headroom at the antenna’s permitted power, while the termination safely absorbs the power that a terminated design intentionally dissipates.

More core is not merely “more watts”

Adding ferrite volume can provide both magnetic and thermal headroom. In a sound design, spreading the work across more core material can reduce flux density and temperature rise. Larger conductors reduce resistance. Better spacing and high-temperature dielectric materials improve voltage margin. Lower temperature also helps the transformer remain stable instead of drifting toward still greater loss.

There is an important warning hidden in that sentence: size does not rescue a poor RF design. The correct ferrite mix, number of turns, transmission-line impedance, winding placement, coupling, compensation, insulation, and lead length still determine bandwidth and insertion loss. A large bad transformer is still a bad transformer.

Our policy is therefore not “add cores until the wattage looks impressive.” It is to design for low loss first, then provide enough magnetic, thermal, current, and voltage margin for the loads the product is likely to encounter.

Why a kilowatt-rated transformer can be excellent at QRP

A passive transformer does not need a minimum amount of power before it starts working. A 3 kW rating is a maximum operating envelope under stated conditions, not a request for 3 kW from the radio.

At QRP and 100 W, the useful specification is insertion loss. Several integrated antenna transformer assemblies are specified below 0.2 dB on their product pages under stated test conditions. This is not a universal figure for every ratio, band, or load. A loss of 0.2 dB corresponds to approximately 95.5% power transfer:

This is valuable at QRP because every fraction of a watt belongs in the antenna, not in a warm enclosure. It is valuable at 100 W for the same reason, with the additional benefit that the transformer usually runs cooler and retains substantially more thermal and magnetic margin when the match is less predictable than it was in the workshop.

The 1:1 coaxial choke is the important exception

A conventional 1:1 coaxial common-mode choke is not trying to transform 50 Ω into 200 Ω, 2,450 Ω, or 3,500 Ω. The wanted differential RF travels inside the coax. The ferrite is primarily used to oppose unwanted current flowing on the outside of the shield.

In a well-balanced installation, common-mode current may be small, so the core sees far less differential-mode magnetization than a ratio transformer. The choke conductors still carry the full differential line current, however, and leakage flux, turn-to-turn voltage, insulation, and actual common-mode current remain design limits. This is why a genuine 100 W ICAS single-core 1:1 choke can be a rational product.

Forward power or SWR alone does not determine choke stress. Severe imbalance, unintended return paths, or a mismatch that changes the installation’s common-mode excitation can increase common-mode current. Placement also changes the common-mode circuit. Once installed common-mode current is acceptably low and the choke provides sufficient impedance relative to the actual common-mode path on every required band, additional impedance at the same location often gives diminishing suppression gains. Extra ferrite then mainly adds thermal and magnetic stability, although it can also reshape or broaden the impedance response.

A doublet can make a 1:1 balun a difficult job again

It is easy to hear “1:1” and assume the component has the same easy life as a coaxial line isolator on a well-matched dipole. That is not always true.

A multiband doublet with open-wire line can present widely different impedances at the tuner on different bands. Feed-line length changes those impedances again. On one band, the tuner-side balun may see high voltage; on another, high current. It may also need to preserve balance and suppress common-mode current while carrying the full differential RF.

The differential voltage and current chiefly set the balun’s conductor, insulation, spacing, connector, and transmission-line limits; only uncanceled common-mode current directly excites the ferrite. High line SWR alone therefore does not predict choke-core heating.

The tuner can make the radio happy, but it cannot make those currents and voltages disappear. This is why our 600 Ω open-wire doublets are extremely attractive for low-loss QRP and multiband use, yet a coax-to-open-wire 1:1 interface still needs generous headroom. For example, one of our coax-to-open-wire tuner baluns is specified at 3 kW ICAS / 1.2 kW CCS as a single unit; its 10 kW CCS configuration requires two units installed in the specified series arrangement with spacing. The open-wire line minimizes feed-line loss under high SWR; the balun still has to survive the impedance presented at its actual location.

Our design policy

We would rather build one properly engineered, low-loss transformer that works superbly from QRP through 100 W and retains real installation margin than build a smaller unit whose “100 W” label is true only into a resistor, on a favorable band, for a short transmission.

That does not mean a large transformer makes bad operating practice safe. Always tune the antenna, use the correct ratio, provide the intended return path, measure common-mode current where possible, reduce power under severe mismatch, and respect the product’s ICAS, CCS, digital, SWR, and frequency limits.

In summary

We do not offer 100 W versions of our 4:1, 9:1, 49:1, 68:1, or 70:1 impedance-matching transformers because “100 W” does not describe the load they must survive. High-ratio units can already approach a kilovolt peak under an ideal 100 W match. For an ideal transformer and the same 100 W of accepted power, a purely resistive 4:1 VSWR can double either the relevant winding voltage or current, depending on which side of nominal the impedance falls. Multiband operation, reactive loads, long duty cycles, installation changes, and temperature reduce the usable power margin further.

A normal 1:1 common-mode choke is less directly stressed by differential transmitter power, so a well-designed 100 W version can be sufficient in a clean installation. A tuner balun feeding a doublet is different: even at 1:1 transformation, it may encounter extreme differential voltage or current and still needs generous headroom.

For QRP and 100 W, our high-power transformers are not inefficient giants. They are deliberately low-loss designs operating comfortably inside their envelope—where more of your transmitter’s power reaches the antenna and less becomes heat.

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

Questions about transformer selection, derating, or your antenna installation? Feel free to contact RF.Guru via our RF.Guru contact page.