I have a boost converter that will convert 12V and 11A to around 20V and 6A. Since I didn't find any inductors (in my local stores) capable of such currents and frequencies and the correct inductance I decided to build one myself.

I've calculated the required inductances to be around 3uH for 11A and a couple of 710uH (for a low pass filters in 'boosted' section) for 6A, but am having doubts about saturation current.

I've seen an anwser on this site showing how to measure it, but I would like to calculate it prior to actually winding them.

I've seen some ferrite cores with 'air gapes' to achieve high saturation currents, I've seen toroidal and cylindrical cores with and without those air gaps. But I havent seen any specific numbers in their datasheets about the saturation current. So I am completely lost here...

So is saturation current dependent on core material/type/dimensions? If I know these values how can I calculate it?


1 Answer 1


Saturation of the core depends on the core material but usually for ferrite type materials it is round about 0.4 teslas (an example is 3C90 material from ferroxcube): -

enter image description here

Notice that despite an increasing H field (basically amps), the flux density is leveling out at about 400 mT. So what is the H field all about?

The H field is the thing that makes magnetism and for a regular ferrite (say a toroid), H field is amperes x turns / distance around toroid. By distance around toroid I means \$\ell_e\$ (effective length) shown below: -

enter image description here

So, a bigger toroid has a longer \$\ell_e\$ and therefore the H field for a given number of ampere-turns will be smaller. This means the toroid will saturate less for the same current.

What does a gap do? A gap reduces the effective permeability of the core material and, remembering that B = \$\mu H\$, a smaller permeability also means a smaller flux density for the same current.

In summary, it is amps x number of turns that is the driving force (magneto motive force) but when taking into account the effective length of the toroid (or E core etc..), MMF maps to magnetic field strength H and this, coupled with the effective permeability of the core dictates saturation point of flux density B.

So a bigger ferrite with a gap is a must when avoiding saturation.

It's also worthy of note that many designer's aim is to have a certain value of inductance and that introducing a gap may halve the permeability. This halves the inductance yet, increasing the turns by \$\sqrt2\$ restores the inductance to its previous value. Inductance is proportional to turns-squared.

This tells you that you can obtain the same inductance from a gapped version of a core AND lower the saturation point. However, copper losses may dictate that there is no net gain in "efficiency" without going to a bigger core.

Another method is to choose two inductors of twice the value and put them in parallel. If you look at a power inductor's spec you will see that the saturation current for the inductor of the higher value is more than half the saturation current of the inductor with half the inductance.

Reference BH curves for various materials: -

enter image description here

Magnetization curves of 9 ferromagnetic materials, showing saturation.

1.Sheet steel,

2.Silicon steel,

3.Cast steel,

4.Tungsten steel,

5.Magnet steel,

6.Cast iron,




Cobalt and Nickel are both used in ferrites. Information taken from here

  • \$\begingroup\$ In lots of datasheets I dont see this kind of graph (or any graph for that matter...). Can I assume that all of the ferrite type cores have a saturation point at about 0.3mT (a bit of margin from what you say)? \$\endgroup\$
    – Golaž
    Apr 10, 2015 at 7:07
  • \$\begingroup\$ @Golaž if you are serious about designing inductors you have to use cores that are fully specified within their data sheet. Normally ferrites saturate at between 200mT and 500mT. See the following site for several examples but note that 3000 gauss = 300mT: bnf.com.hk/new_page_8.htm \$\endgroup\$
    – Andy aka
    Apr 10, 2015 at 7:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.