Discrete gap
A ground gap in the centre leg. Efficient energy storage and tight inductance control, at the price of a fringing field that pushes into the nearby turns and heats them. Keep the winding back, or grade the gap.
Home / Technology / Core Materials
Core materials, and why the choice is not obvious
ConverNova designs custom magnetics on ferrite, nanocrystalline, amorphous and powder cores, from Ferroxcube, TDK, Magnetics, VAC, Hitachi and Micrometals. Material and geometry are one decision, not two: the same inductance can be built four ways, and which one is right depends on your frequency, your ripple, your temperature rise and the space you have. This page is what we would tell you on the phone.
The four families
What each material buys, and what it costs.
| Material | Buys you | Costs you | Typical use |
|---|---|---|---|
| Ferrite, MnZn | The lowest core loss at switching frequencies, wide range of standard shapes, low cost | Low saturation flux density, hard saturation, and it is brittle | Almost every high frequency transformer and gapped inductor |
| Ferrite, NiZn | Useful impedance well into the tens of megahertz | Low permeability, so it is a high band tool rather than a power core | High frequency EMI suppression |
| Nanocrystalline | Very high permeability and high saturation, so high impedance or high energy in a small volume | Core loss rises steeply with frequency, and the material is expensive | Common mode chokes, high current inductors, low frequency transformers |
| Amorphous | High saturation and good tolerance of unbalanced current | Bulkier than nanocrystalline for the same impedance | Line frequency and low kilohertz chokes |
| Iron powder | A distributed gap, soft saturation, no fringing hotspot next to the winding | Higher core loss, so it needs thermal headroom | PFC and boost chokes with large DC bias and modest ripple |
Numbers for a specific grade come from the manufacturer's datasheet, and we work from the loss curve at your frequency and flux swing rather than from the headline permeability.
Geometry
The shape is half the decision.
Two cores of the same material and the same effective area behave differently if the window is a different shape.
Send us your spec- E and EI
- The workhorse. Generous window, easy to wind, easy to gap on the centre leg where the fringing field is furthest from the outer turns
- ETD and ER
- Round centre leg, so the winding follows a circle instead of turning four corners. Shorter mean turn length, less copper, less loss
- Planar E, ER, EQ
- Low profile with a wide flat window for PCB or stamped windings. See planar transformers
- PQ and RM
- Good volume efficiency and a partially shielded winding, for tight boards where stray field is a nuisance
- Toroid
- No air gap in the magnetic path, so almost no stray field. The right shape for common mode chokes, and the awkward one to wind
- U and C cores
- Long window for high isolation, and room to place a graded gap where it does the least harm
- Custom machined
- When a standard shape wastes window, forces a height you cannot afford, or puts the gap where the fringing field bakes your copper. We machine the core and build the production tooling
The gap
Where the energy actually goes.
In a gapped inductor the energy is stored in the gap, not in the ferrite. That single fact decides most of the design.
Stepped or graded gap
The gap machined in steps so the fringing field spreads out before it reaches the copper. More machining, and it earns its place when the winding loss from fringing is what is limiting you.
Distributed gap
The gap spread through the whole core by using a powder material. No fringing hotspot and a soft saturation curve, paid for in core loss across the whole volume.
Straight answers
Core selection, the usual questions.
Ferrite or nanocrystalline for a high frequency transformer?
Ferrite for most switching transformers, because its core loss at high frequency
is far lower. Nanocrystalline where you need very high permeability or very high saturation
in a small volume, which mostly means common mode chokes and high current inductors. Above a
few hundred kilohertz nanocrystalline core loss climbs steeply and ferrite wins on
efficiency.
When is powder the right answer?
When you want a distributed gap. It saturates softly instead of collapsing,
which gives graceful overload and removes the fringing hotspot a discrete ferrite gap puts
next to your winding. You pay in core loss, so the design needs thermal headroom.
When is a custom machined core worth the tooling?
When a standard geometry forces a compromise you will live with for the life of
the product: wasted window, a height you cannot afford, or a gap in the wrong place. We
machine the core and build the production tooling, so the geometry that wins the prototype is
the one that ships.
Which suppliers do you work with?
Ferrite from Ferroxcube and TDK. Nanocrystalline from VAC and
Hitachi. Amorphous alloys. Iron powder from Micrometals. Plus equivalents when you need a
second source approved.
Construction
Cores, wound and taped.
Stacked laminations and toroids, with the geometry chosen alongside the material rather than after it.
Start here
Not sure which core your design wants?
Send the waveform, the frequency, the peak and RMS current and the space you have. We will tell you which material and which geometry, and why, and we aim to do it inside 24 hours.