HF Ferrite Selection: Chokes, Transformers and EFHW Limits
HF Ferrite Selection: Chokes, Transformers and EFHW Limits
A choke needs useful complex common-mode impedance. A transformer needs efficient energy transfer, adequate magnetizing impedance, controlled flux and manageable parasitics. Mix numbers describe materials—not completed designs or watt ratings.
Choose ferrite from the electrical job, the exact manufacturer material and part data, and measurements of the finished winding. Common-mode suppression rewards useful complex impedance; a power transformer needs adequate magnetizing impedance, controlled flux, low delivered-power loss, safe voltage spacing and stable temperature.
Scope: this guide uses Fair-Rite material numbers and current Fair-Rite data. A number such as “43” is not a universal cross-manufacturer specification. Even within one material, core part number, dimensions, coating, effective area, magnetic path, AL tolerance, winding and thermal environment determine performance.
Start with the Electromagnetic Job
Common-mode choke
On a coaxial choke, wanted differential currents ideally create cancelling flux. Unwanted current on the cable exterior sees the choke’s common-mode impedance:
Zcm(f) = Rcm(f) + jXcm(f)
Pdiss ≈ Icm,rms² × Rcm at a defined frequency and temperature.
A substantial resistive component can damp resonances and absorb common-mode energy, but that resistance produces heat while current remains. Choke design therefore needs measured R and X, not merely |Z| or a mix name. The transmission line, connector and insulation still carry full differential voltage and current even when core flux cancels ideally.
Power or impedance transformer
A transformer is supposed to deliver power to the load with low dissipation. Its magnetizing branch should not consume excessive current, and the intended flux swing should remain in a low-loss region. Leakage inductance, winding capacitance, copper loss, dielectric loss and transmission-line behaviour bound the upper frequency.
The same complex permeability that is useful in a choke can be loss in a transformer. If a winding’s complex inductance is written L′ − jL″, its series impedance is approximately ωL″ + jωL′: the loss term becomes real resistance. The balance of those terms moves with frequency, temperature, field level and bias.
A balun label does not settle common mode
A 4:1 or 9:1 network may be a Ruthroff autotransformer, Guanella transmission-line transformer or another topology. They do not have the same voltage distribution, isolation, balance or common-mode behaviour. Impedance transformation and common-mode choking are separate specifications; a transformer called a “balun” does not automatically provide enough feedline isolation.
Manufacturer Applications and Boundaries
The table below reports manufacturer descriptions, not automatic amateur-radio recommendations. Initial permeability is a low-flux small-signal value; it is not an HF loss or power rating.
| Fair-Rite material | Published initial permeability | Current manufacturer description | Engineering implication |
|---|---|---|---|
| 31, MnZn | 1,500 | Designed specifically for EMI suppression from 1 to 500 MHz | Strong HF choke candidate, but turns, part geometry, resonance and heating still need measurement. |
| 43, NiZn | 800 | Popular for conducted-EMI suppression from 20 to 250 MHz and high-frequency common-mode chokes | Multiple turns can move useful impedance downward, but the page does not certify a universal 1.8–30 MHz power transformer. |
| 52, NiZn | 250 | High-frequency material combining high saturation flux density and high Curie temperature | A candidate requiring core-specific loss tests; high saturation and Curie values do not create a watt rating. |
| 61, NiZn | 125 | Developed for inductive applications up to 25 MHz; also EMI suppression above 200 MHz | Relevant to upper-HF inductive/transformer work; lower permeability increases the turns needed for low-band magnetizing impedance. |
| 67, NiZn | 40 | Intended for broadband transformers, antennas and high-Q inductors up to 50 MHz | A documented transformer candidate, especially toward upper HF; low-band turns and parasitics remain design constraints. |
| 73, MnZn | 2,500 | Supplied only in small cores to suppress conducted EMI below 50 MHz | Useful suppression material in available sizes; not evidence for a QRO HF transformer or large coax choke. |
| 75, MnZn | 5,000 | Intended for broadband and pulse transformers and common-mode inductor designs | Potential low-frequency transformer/choke material, but its complex-permeability, loss and temperature curves must cover the actual MHz operating point. |
| 77, MnZn | 2,000 | Inductive designs at high and low flux density for frequencies up to 100 kHz | Its stated application ceiling is far below 1.8–7 MHz; HF use requires direct loss, flux and thermal evidence for the exact winding and load. |
The selection logic is functional rather than categorical. Material 31 is a manufacturer-supported suppression candidate; materials 61 and 67 carry explicit inductive or broadband-transformer descriptions; material 77’s stated range ends at 100 kHz. The completed winding must still meet its frequency, load, loss, flux, voltage and thermal limits.
The Fair-Rite 17th-edition catalogue provides complex-permeability, temperature and bias information and makes clear that component data and application ranges matter. Its small-signal material curves do not validate a completed winding at transmitter power.
Impedance Ratio Is Not a Material-Selection Rule
For an ideal transformer, impedance ratio is the square of voltage or turns ratio. At 50 Ω:
| Impedance ratio | Ideal turns/voltage ratio | Ideal high-side resistance | High-side voltage at 100 W | High-side voltage at 1 kW |
|---|---|---|---|---|
| 4:1 | 2:1 | 200 Ω | 141 V RMS | 447 V RMS |
| 9:1 | 3:1 | 450 Ω | 212 V RMS | 671 V RMS |
| 49:1 | 7:1 | 2,450 Ω | 495 V RMS | 1.57 kV RMS |
| 64:1 | 8:1 | 3,200 Ω | 566 V RMS | 1.79 kV RMS |
Those are ideal sinusoidal, purely resistive reference-plane examples. An EFHW’s installed impedance changes with frequency, height, wire route, surroundings and return path; reactive loads and standing waves can produce different internal maxima. The ratio alone says nothing about core material, safe power or efficiency.
A high-ratio winding commonly needs more total turns and more high-side insulation distance. That tends to increase leakage inductance, inter-turn and winding-to-core capacitance, transmission-line delay and self-resonance risk. Those parasitics—not only low-band inductance—often end an attempted 80–10 m response.
Magnetizing Inductance and Flux Are Separate Checks
For a specified core part in its small-signal region:
Lm ≈ AL × Nexc²
Xm = 2πfLm
Bpk ≈ Vrms / (4.44 f Nexc Ae) for a sinusoid, using the turns actually excited by that voltage.
AL and effective area Ae belong to an exact part number, not just a mix. For example, Fair-Rite’s current 61 mm Mix 52 toroid 5952003801 publishes AL = 325 nH ±25% at 10 kHz and Ae = 1.58 cm². Its tolerance alone changes calculated magnetizing inductance materially; its 10 kHz test value is not proof of identical behaviour at HF or at high flux.
A magnetizing-reactance target such as five or ten times 50 Ω can be a starting criterion, not a universal proof. The correct value depends on acceptable magnetizing current, phase and insertion loss in the topology and load. More turns raise inductance but also parasitics, copper length and high-side voltage between winding sections.
The sinusoidal flux equation is an estimate. Non-sinusoidal voltage requires integrating volt-seconds; modulation, harmonic content and transient or fault conditions change the result. In an autotransformer, identify the actual excited turn section rather than inserting the total turns blindly.
Saturation Is Usually Not the First HF Limit
A published flux-density point on a low-frequency B-H curve is not a permissible HF operating flux. Core loss rises with frequency, flux swing, waveform and temperature, and can cause thermal runaway or property drift well before classical saturation. Fair-Rite explicitly warns on its 61 and 67 pages that strong magnetic fields or mechanical stress can irreversibly change permeability or loss.
Curie temperature is where ferrimagnetic behaviour collapses; it is not a safe case or winding temperature. Likewise, Mix 52’s high Curie temperature and high saturation flux density make it interesting, but do not prove that it is “higher power” in an arbitrary transformer. Power capability also depends on volume, surface area, cooling, winding copper, duty cycle, ambient, enclosure and the chosen maximum temperature.
Stacking cores can increase effective magnetic cross-section and thermal mass when flux is shared correctly. It does not multiply a universal watt figure. Uneven winding distribution, air gaps between cores, local electric fields and reduced cooling inside a stack can still set the limit.
Choke Power Is Not Transmitter Power
An ideal common-mode choke carries transmitter differential power with cancelling magnetic flux. Its core heating is driven mainly by residual common-mode current and loss, while the cable and connector carry the differential current. Consequently, “1 kW choke” is incomplete without:
- line impedance, differential voltage/current, SWR and connector rating;
- measured common-mode
R + jXover every band; - assumed or measured common-mode current and voltage;
- turns, cable type, bend radius and winding self-capacitance;
- duty cycle, ambient, enclosure and equilibrium temperature; and
- dielectric withstand from winding to winding, core and enclosure.
More resistive choking impedance can be desirable because it damps a common-mode resonance, but it also dissipates more power for a given current. A design succeeds by reducing that current while staying cool—not by maximizing resistance without a current and thermal boundary. Steve Hunt, G3TXQ’s measured HF choke work demonstrates why resistance, reactance and useful frequency span should be inspected separately.
Transformer Power Needs Complex-Load Tests
A 4:1 transformer tested only into 200 Ω, or a 49:1 tested only into 2,450 Ω, has not been qualified for a real antenna. Measure the actual intended complex load range. High SWR does not uniquely determine stress because the load phase and reference plane control voltage and current distribution.
Small-signal insertion loss is useful but insufficient. At power, record input and delivered power with calibrated reference planes, return loss at both ports, waveform, frequency, duty cycle and temperature until equilibrium. A back-to-back pair can expose problems, but simply dividing total dB by two assumes identical, non-interacting devices and appropriate terminations; verify that assumption.
For high-ratio EFHW transformers, inspect winding and terminal voltage, corona/arcing, winding-to-core capacitance, common-mode current and hot spots. A low SWR can coexist with ferrite loss. Enclose kilovolt-class terminals, provide strain relief and weather protection, and never touch or reconfigure the system while transmitting.
A Practical Selection and Validation Sequence
- Name the function and topology. Common-mode choke, isolation transformer, Ruthroff unun and Guanella transformer are different circuits.
- Define frequency and loads. Include the lowest and highest frequency and the full complex source/load range, not only a nominal resistance.
-
Select exact core candidates. Use manufacturer application, complex-permeability and part data; record material, part number, dimensions,
AL,Ae, volume, coating and tolerances. - Calculate first-order limits. Check magnetizing current, flux waveform, ideal voltages/currents and required insulation before winding.
- Build for parasitics. Control lead length, interleaving, transmission-line impedance, turn spacing, leakage and winding capacitance. Measure self-resonance.
- Characterise small signal. For chokes, measure complex common-mode impedance plus differential loss/match. For transformers, measure input/output match, transfer and balance/isolation as applicable.
- Test at power. Use representative complex loads, waveform and duty to thermal equilibrium; monitor cores, winding, connectors and dielectric stress.
- Test the installed antenna. Recheck common-mode current, temperature, loss and match after the actual feedline, return path and environment are connected.
Engineering conclusion: there is no reliable lookup table from 31/43/52/61/67/73/75/77 to “choke,” “4:1,” “9:1,” “49:1” or “64:1.” Manufacturer application data identify candidates; complex impedance, magnetizing current, flux, parasitics, delivered power, voltage withstand and equilibrium temperature decide whether the completed design works.
Primary Manufacturer and Measurement References
- Fair-Rite 17th-edition catalogue — material properties, complex permeability and application data
- Fair-Rite Mix 31 current material page
- Fair-Rite Mix 43 current material page
- Fair-Rite Mix 52 current material page
- Fair-Rite Mix 61 current material page
- Fair-Rite Mix 67 current material page
- Fair-Rite Mix 73 current material page
- Fair-Rite Mix 75 current material page
- Fair-Rite Mix 77 current material page
- Fair-Rite 5952003801 current core-part data
- Fair-Rite toroidal-suppression-core comparison flyer
- G3TXQ — measured common-mode choke resistance and reactance
Mini-FAQ
- Is Mix 31 automatically the best common-mode choke material across all HF bands? No. Fair-Rite designed it for EMI suppression from 1 to 500 MHz, making it a strong candidate, but useful resistance, reactance, resonance and heating still depend on the exact core, turns, cable and installation.
- Is Mix 43 the universal material for 4:1, 9:1, 49:1 and 64:1 HF transformers? No. Its current manufacturer page emphasizes 20–250 MHz EMI suppression and high-frequency common-mode chokes. A particular transformer winding needs loss, magnetizing, parasitic and power validation.
- Does Mix 52’s higher saturation flux density and Curie temperature prove higher power handling? No. Those properties are useful inputs, but completed-transformer power also depends on core geometry and volume, frequency, flux swing, winding, load, duty, cooling, insulation and allowed temperature.
- Is Mix 77 a proven default for 160/80 m EFHW power transformers? No. Fair-Rite describes Mix 77 for inductive designs up to 100 kHz. Use at 1.8–7 MHz requires measured loss, flux and thermal evidence for the exact winding and load.
- Why should Mix 67 be considered in an HF transformer discussion? Fair-Rite explicitly intends Mix 67 for broadband transformers, antennas and high-Q inductors up to 50 MHz. Its low permeability can require more turns at low frequency, so parasitics and flux still have to be designed.
- Does the impedance ratio alone determine the ferrite mix? No. Ratio sets an ideal turns and voltage relationship. Frequency, topology, excitation turns, core geometry, complex loads, magnetizing current, flux, loss, parasitics, temperature and insulation determine a viable material and winding.