When Ferrite Runs Hot: Core Loss, Flux and Real Power Limits
When Ferrite Runs Hot: Core Loss, Flux and Real Power Limits
Ferrite in a choke or transformer can heat because of magnetic loss, unwanted current, excessive flux, winding loss or poor thermal conditions. The transmitter’s wattage and the colour of the core do not tell us which mechanism is responsible.
Most of us have installed a ferrite choke, balun or transformer and assumed it would keep doing the same job forever. Then a contest, a long digital transmission or a difficult antenna load makes the enclosure warm, tuning drifts or RF returns to the shack. That is not ferrite magic wearing off. It is the complete magnetic and thermal circuit telling us that one of its limits has been crossed.
The practical rule: do not rate a ferrite assembly from core size, mix number or transmitter power. Measure the electrical function, drive it with representative loads and waveform, wait for thermal equilibrium, and respect the lowest limit among core, winding, cable, insulation, connector, enclosure and environment.
Heat Has More Than One Source
A hot enclosure does not identify a hot core, and a hot core does not identify one loss mechanism. The complete loss budget can include:
| Loss or stress | What drives it | Evidence that separates it |
|---|---|---|
| Magnetic core loss | Material, frequency, flux excursion, waveform, bias, temperature and core geometry. | Material/core data under comparable excitation plus powered core-temperature and transfer measurements. |
| Winding and connector loss | RMS current, conductor resistance, skin/proximity effects, contacts, joints and local current crowding. | Four-wire DC checks where useful, differential insertion loss, local thermal imaging and connector inspection. |
| Dielectric and insulation stress | Voltage distribution, electric-field concentration, material loss, contamination, moisture and spacing. | Voltage and field analysis, insulation/material limits, visual inspection and appropriate withstand testing. |
| Common-mode dissipation | Unwanted-mode current flowing through the resistive part of the choke impedance. | Installed common-mode current, complex impedance at the operating point and temperature rise. |
| Environmental heat | Ambient temperature, solar gain, nearby equipment, enclosure, airflow and thermal interfaces. | Ambient and internal temperature records with the RF drive removed and applied. |
For a transformer, core loss and winding loss are unwanted because the intended job is to transfer differential RF power. For a common-mode choke, the resistive part of its impedance can deliberately dissipate unwanted-mode energy, while the reactive part also opposes current without the same real-power loss. Both components matter; “ferrite converts RF to heat” is not the whole model.
A Choke and a Transformer Excite the Core Differently
In a transformer, applied voltage, frequency, number of turns, effective core area and waveform govern the magnetic flux excursion. Too few turns at low frequency, excessive voltage or an unfavourable waveform can drive high flux even when the output power sounds modest. A reactive or mismatched load can add voltage and current stress that a resistive bench load did not reveal.
In a common-mode choke, equal-and-opposite differential currents ideally cancel their magnetising effect. The exterior common-mode current does not cancel and excites the core. That is why transmitter output power alone cannot predict choke heating: two 1 kW stations can have radically different common-mode current, and a badly controlled return path can heat a choke at much less transmitter power.
The distinction also explains why a ferrite material suitable for one transformer circuit is not automatically a good suppression material, and why a lossy suppression material is not automatically suitable for efficient differential power transfer. The exact manufacturer data, winding and operating conditions decide.
Saturation, Loss and Temperature Are Not Synonyms
Saturation is the high-excitation region where increasing magnetising force produces progressively less increase in flux density. Effective permeability and inductance change, so the circuit waveform and impedance can change sharply. But a ferrite component can overheat from core loss or conductor loss before it reaches a simple saturation threshold.
Likewise, a saturated core does not merely become a fixed low-value resistor. The device becomes nonlinear: inductance, current waveform, harmonics, voltage distribution and loss can all change. The result depends on the source and load circuit.
The generic TDK ferrite summary treats saturation flux density, permeability and core loss as different properties. TDK’s magnetic definitions also separate hysteresis, eddy-current and residual contributions to core loss and require winding loss to be added to the core term. That is the right discipline for an HF assembly: do not use one material number to stand in for the entire device.
Curie Temperature Is Not the Operating Limit
The Curie temperature is the material transition above which ferromagnetic behaviour is lost. It is not permission to operate a finished choke or transformer anywhere near that temperature. Useful permeability, saturation flux density and loss already vary with temperature, while cable dielectric, wire insulation, coatings, adhesives, spacers, solder joints, connectors and the enclosure may have lower limits.
The maximum allowable component temperature is therefore the lowest applicable limit in the completed assembly, with margin for measurement error, hot spots, ambient extremes, solar loading and ageing. A core surface measurement can also underestimate a hot winding or enclosed internal point.
Do not use the finger test. “Too hot to touch for two seconds” is neither a calibrated temperature nor a safe operating criterion. De-energise the system, prevent accidental transmission, allow hazardous charge to discharge as designed, and use a suitable contact sensor or correctly configured thermal method.
Why the Simple I²R Example Can Mislead
For a linear resistor carrying known RMS current, P = I²R is valid. A ferrite choke under real RF drive is more complicated. The resistance in its small-signal impedance trace may change with frequency, current, temperature, winding capacitance and the impedances on both sides. The current distribution can also change after the choke heats.
A calculation such as “2 A times 2 A times 2 Ω equals 8 W” is useful only if 2 A RMS is actually measured in the relevant common mode and 2 Ω is the real part of the choke impedance at that same operating point. Scaling that current from 100 W to 1 kW without measuring the antenna system is not defensible.
Use the formula as an energy-accounting check after the operating quantities are known—not as a shortcut from transmitter watts to ferrite temperature.
Core Size and Mix Labels Do Not Create a Universal QRO Table
A larger core can provide more magnetic cross-section, volume and thermal mass, but it can also be wound differently, see a different flux excursion or be enclosed with poorer cooling. Multiple cores can distribute loss when the magnetic and thermal design supports that result; core count alone is not a rating.
The same is true of sleeves, beads and clamp-on parts. A small part is not automatically restricted to receive use, and a large toroid is not automatically safe at high power. Exact material, dimensions, number of passes, common-mode current, voltage distribution, frequency range, thermal path and duty cycle decide. The reviewed sleeve-versus-wound-core comparison shows why no universal conversion exists.
Material labels such as 31, 43, 52, 61, 73, 75, 77 or 78 belong to manufacturer-specific data families. They cannot be reduced to a single band-winner table for both transformers and chokes. The Fair-Rite technical catalogue gives frequency-dependent complex permeability, impedance, loss and temperature data under stated fixtures and conditions; the completed assembly still has to be measured.
Enclosure and Duty Cycle Change the Result
A weatherproof enclosure can protect against water while restricting convection. Direct sun, a dark box, attic heat, nearby equipment and limited conduction can raise the starting temperature before RF is applied. Conversely, an exposed core can run cooler but may create mechanical, weathering and insulation problems.
Duty cycle matters because heat is integrated over time. SSB voice, continuous carrier, RTTY, FT8 and AM can produce different thermal histories at the same peak transmitter setting. So can contest cadence, receive intervals and band changes. A short key-down check proves neither thermal equilibrium nor long-term survival.
Thermal cycling is not automatically a “silent killer” that steadily destroys every ferrite. Damage depends on whether temperatures, gradients and expansion mismatch exceed the limits of the core, winding, coating, adhesive, cable, connector or enclosure. Inspect and remeasure after representative cycles instead of assuming either permanent damage or perfect recovery.
What to Measure Before Calling It a Power Rating
- Identify every part. Record manufacturer, material, dimensions, batch or tolerance where available, winding conductor, cable, connectors, insulation, enclosure and thermal interfaces.
- Define the electrical job. State whether the device transfers differential power, suppresses common mode, transforms impedance, provides balance or combines functions.
- Map the real loads. Measure complex source and load impedance plus common-mode source/load conditions across the intended bands.
- Measure the small-signal baseline. Record complex impedance or S-parameters with fixture, calibration, reference planes and uncertainty.
- Apply representative drive. Use the intended waveform, voltage, current, mismatch and duty cycle in controlled steps.
- Wait for equilibrium. Record ambient, core, winding, cable, connector and enclosure temperatures until the relevant points stabilise.
- Watch for nonlinearity. Recheck impedance, insertion loss, SWR, current, waveform and harmonics as drive and temperature rise.
- Cycle and inspect. After representative thermal cycles, inspect mechanical and insulation condition and repeat the electrical measurements.
IEC 62044-3:2023 specifies measurement methods for core power loss and amplitude permeability at high excitation. Those are core-characterisation tools; a finished HF assembly still requires its own electrical and thermal qualification with the installed load.
Symptoms Point to Tests, Not Conclusions
| Observation | Possible causes | Next measurement |
|---|---|---|
| Core or enclosure temperature rises | Core loss, winding/connector loss, common-mode dissipation, solar or ambient heating. | Separate component temperatures, electrical loss and no-RF environmental baseline. |
| SWR or tuning drifts | Temperature-dependent permeability, winding/connector change, antenna/environment drift or instrument reference-plane movement. | Repeat complex impedance at fixed planes while logging temperatures and restoring the baseline. |
| Common-mode current returns | Choke impedance changed, another path bypassed it, current redistributed or the source changed. | Multi-position exterior-current scan before, during and after the thermal event. |
| Odour, discolouration, softened jacket or damaged insulation | A material or hot-spot limit may already have been exceeded. | De-energise; inspect and replace or professionally evaluate before reuse. |
An overheated assembly should not be returned to service merely because it cooled down. Ferrite may or may not have changed permanently, while hidden insulation, cable or connector damage can be the lower and more important limit. Electrical remeasurement and physical inspection decide whether reuse is defensible.
Primary and Authoritative Sources
- Fair-Rite Products — 17th-edition technical catalogue: manufacturer material data, complex permeability and impedance, loss, temperature and application boundaries.
- TDK — Ferrites summary: manufacturer definitions for magnetisation, saturation flux density, permeability and core loss.
- TDK Electronics — Magnetic Design Tool guidance: measured core-loss dependence on frequency, flux density and temperature for declared materials and waveforms.
- IEC 62044-3:2023: current high-excitation measurement methods for power loss and amplitude permeability of soft-magnetic cores.
- IEC 60401-3:2015: catalogue-data presentation, measurement conditions and reliability guidance for magnetically soft ferrite cores and devices.
Joeri’s Bottom Line
Ferrite is a component, not a talisman. A choke that behaves well at one station, band and duty cycle has not earned an unlimited QRO label. A transformer that stays cool into a resistive test load has not proved its behaviour into a reactive antenna. And a material mix that is useful in one circuit is not a universal answer for another.
When a core runs hot, keep the warning—but diagnose it properly. Separate magnetic loss from winding loss, flux from common-mode current, temperature rise from Curie temperature, and core data from the limits of the finished assembly. Measure the real load, drive and thermal equilibrium. That is how we stop “cooking alive” from becoming damaged cable, unreliable suppression or a stressed transmitter.
Mini-FAQ
- How hot is too hot for ferrite? Use the lowest specified limit of the complete assembly with engineering margin. Curie temperature and a finger-touch test are not allowable operating temperatures.
- Does a hot choke mean the ferrite saturated? Not necessarily. Core loss, winding or connector loss, common-mode dissipation and environmental heat can raise temperature before or without a simple saturation event.
- Can transmitter watts predict choke heating? No. Choke heating depends on actual common-mode current, complex impedance, voltage, frequency, duty cycle, temperature and the installed source/load path.
- Are larger or multiple cores always safer at QRO? No. More core volume can help a verified design, but winding, flux, current, voltage, parasitics, cooling and the lowest assembly limit still decide.
- Can I reuse a ferrite assembly after overheating? Only after de-energised inspection and repeat electrical testing show that the core, cable, winding, insulation, joints, connectors and enclosure remain within specification.
- How should a ferrite power limit be established? Measure the completed device with representative complex loads, waveform and duty cycle; verify transfer or impedance, nonlinearity, hot spots and equilibrium temperature.