I understand... I thought that you are making some type of equalizer where inductors are used too. If you have not a toroid core, so you can use EI core too. Could you tell me which inductance you have to have?
EI cores have less efficiency compared to toroidal cores. However I don't think is so easy to build a (efficient and reliable) toroidal core by yourself. Someone tried to mix iron powder with some type of epoxy, but I don't know the results.
Yes, they have less efficiency then toroidal cores, but if you dont have any toroidal core, you have a "no choice".
To mix iron powder with epoxy is not good idea for me, because you have to mix properly so as the mixture was homogeneous. And I think that efficiency will be lower than EI core - but maybe I am wrong, I havent tried it.
EI cores have less efficiency compared to toroidal cores. However I don't think is so easy to build a (efficient and reliable) toroidal core by yourself. Someone tried to mix iron powder with some type of epoxy, but I don't know the results.
I wonder what's the exact meaning of "efficiency" in this statement?
I don't think that it generally holds. The answer depends on the frequency range and other application parameters.
Laminated steel ("iron") cores (e.g. EI) have about the fourfold saturation flux compared to ferrite cores and a higher relative permeability. On the other hand, they have higher hysteresis losses and stronger non-linearity. Power electronics with frequencies up to 10 or even 20 kHz mostly uses laminated steel.
Inductors, both laminated steel and ferrite need an air gap to be "effective". An exception applies for low Er high frequency ferrites, e.g. used for EMI filters. Iron powder cores have an "inbuilt" distributed air gap by their fill-factor. If you have a torroid core used as storage inductor, it will be most likely a iron powder core, because a high Er ferrite core would be saturated before it reaches reasonable current values.
I think in general the toroidal shape helps to have a magnetic flux as ideal as possible so the magnetic losses are minimized. In fact other shapes, like EI have the problem of the grain alinement (in this case at the interface between E and I). Probably you are right, there are application in which shape other than toroidl are a better choice.
Of course the air gap is almost always needed to stabilize the magnetic permeability; a mix between epoxy and iron powder should gave a distribuited gap (very difficult to control in homogeneity so I don't think an homemade toroidal core could work weel).
However this is an interesting document that clarify some aspect.
If you play around with practical inductor designs, you'll realize that the magnetical material properties and basic design parameters like peak flux matter more for the overall performance than considerations of ideal shape. If the application requires a ferrite core with air gap, it's e.g. very unlikely that you'll end up with a toroid core.
It should be also considered, that the winding technique for toroidal inductors or transformers is difficult and that it's nearly impossíble to make high current coils with good fill factor.
I'm not generally against toroid cores, I have e.g. designed 60 kW toroid transformers. It's just one of several design options.