EFHW Core Magic

A reader sent me this video because he had doubts about the content. After watching it, I understood why: it mixes a few real transformer concepts with hand-wavy folklore about cores, capacitance, distortion, and “magic” winding choices.

EFHW transformers attract mythology because they live at the crossroads of ferrite behavior, winding geometry, common-mode current, and real-world convenience. That also makes them easy to oversimplify. A few turns changed here, a different mix there, and suddenly people start talking as if the core itself contains some secret wisdom.

It does not. What you really have is an impedance transformer with finite magnetizing inductance, finite coupling, stray capacitance, self-resonance, losses, and thermal limits. Once you frame it that way, most of the “magic” disappears and the actual tradeoffs become much easier to understand.

Turns are not vibes, they are a design variable

One of the recurring problems in EFHW videos is the casual “we tried some turns and this seemed better” approach. Experimentation is fine, but turns are not an aesthetic preference. They directly affect magnetizing inductance, leakage behavior, self-resonance, and ultimately loss and temperature rise.

If the lowest band is under-supported, magnetizing reactance is too small, magnetizing current rises, and the transformer can run hot. If the winding count is pushed too far the other way, winding capacitance rises and the high-frequency end can become messy. That is the real trade.

So before judging a winding count, you have to define the goal. Are you optimizing for low-band operation, wideband impedance behavior, lower heating, better coupling, less stray capacitance, or simply a pretty SWR curve? Without stating the target, “best number of turns” is just storytelling.

Mix choice is a trade, not a slogan

Another thing that gets oversimplified fast is ferrite mix selection. Saying one mix is “better” for a range of bands without naming the core size, power level, turns, and winding style leaves out most of the actual engineering.

In practical terms, lower-permeability mixes often need more turns to reach the same low-band magnetizing reactance. More turns often means more winding capacitance. That can help one part of the design while hurting another.

So the real statement is not “mix 61 is better” or “mix 43 is better.” The real statement is: for this core size, this turns ratio, this winding style, this power level, and this operating range, this mix gave the best compromise.

Inductance is not “just permeability”

This is a classic conceptual slip. Permeability matters, but inductance is not the same thing as permeability. Inductance depends on the magnetic material, the core geometry, and the square of the number of turns.

The clean mental model is that the material and geometry define the A_L value, and the winding count then gives you the resulting inductance through L = A_L · N². Once that is clear, a lot of the confusion about “small cores need more inductance” or “less secondary somehow fixes it” falls away.

If you want more magnetizing inductance, the most direct lever is usually turns on the driven winding. Changing the secondary count affects the ratio and parasitics, but it does not magically rescue an under-inducted primary.

The winding behaves capacitively, not the core itself

Another place where the language goes off the rails is when the core gets blamed for “acting like a capacitor.” The core is not the capacitor. The winding system is where the parasitic capacitance lives: turn-to-turn, layer-to-layer, and winding-to-environment.

That matters because the real high-frequency limitation of many EFHW transformers comes from the interaction between inductance and that stray capacitance. As frequency rises, the transformer approaches self-resonance. Past that point, the behavior stops looking like the nice inductive transformer you hoped for.

This is why a single “inductance number” from a casual measurement often tells you very little. Once parasitics matter, you need impedance magnitude and phase versus frequency, not a comforting single-value screenshot.

Phase shift is not automatically distortion

One of the more dramatic claims in the video is the idea that if the transformer starts looking capacitive or shifts phase, you somehow get “distortion outbound and inbound.” That is far too loose.

Linear phase shift by itself is not distortion. Delay is not distortion. Distortion requires nonlinearity or frequency-dependent behavior across the signal bandwidth significant enough to reshape the signal.

What is true is that once a transformer is pushed into a poor operating region, whether because of self-resonance, saturation, poor coupling, or excessive heating, its impedance transformation becomes less predictable and losses can rise. Under real power, that can absolutely degrade performance. But that is not the same as saying “phase equals distortion.”

Good SWR does not prove good efficiency

This is where many EFHW demonstrations become misleading. A transformer can show a calm SWR curve while still wasting a meaningful amount of power as heat. Loss can make a network look comfortably resistive.

In other words, your SWR meter can smile while the transformer quietly behaves like a partial dummy load. That is why “it looks good on the analyzer” is never enough on its own.

If the design is meant for serious use, especially digital modes and other demanding duty cycles, then the real checks are insertion loss or dissipation, temperature rise, and impedance behavior across the actual operating range into representative loads.

Self-resonance is not a universal magic number

Another myth that sneaks in easily is the idea that a certain core type “self-resonates around 9 or 10 MHz” as if that were a fixed property you can quote without context.

It is not. The self-resonant behavior depends heavily on the exact winding count, spacing, lead length, fixture, enclosure, and surrounding capacitance. The core contributes, but the finished winding system defines the practical result.

That is why self-resonance should be treated as a measurement outcome for a specific build, not as a universal folklore number that applies to every winding someone might put on that core.

Heat, saturation, and Curie are different things

Videos on ferrites often blend these concepts into one dramatic story. They should not be merged carelessly.

Saturation means the material becomes increasingly nonlinear and incremental inductance falls. That can drive higher currents, higher losses, and more heating. Curie temperature is something else: the point where ferrite’s magnetic behavior collapses strongly as temperature rises.

In practice, most sane designs should never get anywhere near Curie temperature. Real-world failure usually shows up earlier through excessive loss, heating, insulation stress, and poor current distribution. Curie is not a design milestone. It is a warning siren that something has already gone badly wrong.

Winding layout is a coupling versus capacitance compromise

There is also a grain of truth in the observation that bunching windings too tightly can create local hot spots and more capacitance. Spreading the windings out can help reduce some of that capacitance and distribute thermal load more evenly.

But that is not a universal win button either. As you spread turns, leakage inductance can rise and coupling can worsen. So once again, there is no magic wrap pattern that automatically wins on every band at every power level.

Winding geometry is always about tradeoffs. Better high-end behavior often costs you something somewhere else.

EFHW transformers are unbalanced, so common-mode is part of the design

One of the more solid points in the video is that EFHW systems are unbalanced and need a return path. If you do not provide that path intentionally, the coax shield often volunteers.

That is why a proper common-mode choke at or near the feedpoint is usually the cleanest solution. It helps stop the feedline from becoming the unintended second half of the antenna.

Moving the choke farther down the line can sometimes reduce RF in the shack, but it often just relocates the common-mode current distribution rather than eliminating it. From a design point of view, that is a very different thing.

What a non-mystical EFHW transformer workflow looks like

• Choose core size and material so the low band does not force excessive turns.
• Verify impedance magnitude and phase versus frequency instead of trusting a single inductance number.
• Balance magnetizing inductance, coupling, and stray capacitance through measured winding geometry, not folklore.
• Check insertion loss or dissipation and temperature rise under realistic duty cycle.
• Design the return path intentionally with proper choking and counterpoise strategy instead of letting the coax improvise.

The actual takeaway

• EFHW transformers are not magic. They are lossy, parasitic, real-world RF components with compromises.
• Turns, mix, core size, and winding layout interact. None of them can be judged in isolation.
• Inductance is not the same thing as permeability, and a single inductance number is not enough once parasitics matter.
• Good SWR does not guarantee good efficiency or low heating.
• Self-resonance is a property of the finished winding system, not a universal folklore number for a bare core.
• Heat problems usually show up through loss, current, and insulation stress long before anyone should be talking about Curie as if it were normal operation.
• An EFHW needs an intentional return path and proper choking, otherwise the coax often becomes part of the antenna.

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.