High current choke choosing gaped ferrite vs powder material

[Miyuki]

Hi folks

Im choosing best core for relative high current inductor 60A (with 15A ripple) inductance 10uH
I also need secondary winding with 1:8 so better to have split core than toroid for easy winding

But not sure if will be better gaped ferrite or rather use some powder core as gap for this current is huge like 6mm at ETD49 core or still big about 3mm at much bigger cores
Or use some sendust or so E core which is little more expensive but wont have big gap to worry how it affect winding

Do you have experience with this current level choke?

[duak]

I've only designed magnetics for PWM servo drivers and used only gapped ferrite cores.  There are various powdered material cores available and some are electrically conductive.  When using conductive powdered metal cores for servos or in DC power supplies with significant ripple current you have to consider that eddy current is induced in the core that causes losses, and in some cases significant heating.  eg., I was doing an experiment with a servo and added a filter to its output using powdered metal core inductors.  The cores got terrifically hot even though the current was well below their DC rating.

see http://elnamagnetics.com/wp-content/uploads/library/Micrometals/Design_Guidelines_for_Iron_Powder_Cores.pdf

A colleague designed a clean PWM laser diode driver that used powdered metal cores in the output filter.  Its output current was on the order of 50 A or so.  The windings were multiple magnet wires in parallel to handle the current and make it easy to wind.  This application had very little ripple current and thus low AC losses.

Cheers,

[T3sl4co1l]

Are you concerned about cost, or is this small quantity?

Gapped ferrite is probably easier, since cores and bobbins are available from big distributors.  Powder cores are usually available from specialty distributors ($$) or in larger quantity.

25% ripple is kind of a lot for powdered iron, but the better powder materials will handle it.

Tim

[Conrad Hoffman]

Don't get fixated on the material, but look at the data sheets in terms of frequency and losses. I find the powdered metal useful for high Q lower frequency circuits, but I use toroids. It may not be the best choice for your application.

[Kleinstein]

If size is an issue, the powder cores can have an advantage, as there maximum B field is higher (e.g. a factor of 2) than with ferrites, if not limited by core loss (e.g. high ripple and frequency).  The µ of the powder cores is limited, which corresponds to a kind minimal gap. The high µ powders can be conductive and thus higher loss. With ferrite one is flexible to adjust the gap.

[T3sl4co1l]

Depends. Core loss is usually the dominant limit; and the poor utilization of the toroid shape (i.e., you've got a gaping hole in the middle) tends to give poor power density.

Example, ferrite saturates at 0.3-0.4T (MnZn ferrite), and powder at 0.8-2.0T (various types); but ferrite might handle Bpk = 50mT at 400kHz while powdered iron might handle 20mT, for the same core loss or temp rise say.

Tim

[Miyuki]

I found I have near distributor of KDM Sendust (Kool Mu) cores who have E shapes at store and for very reasonable price even at single piece
for example
Sendust 26 EE 55 core for 3$ piece and same size Ferrite CF139 is 2$ piece
It is not so big difference in core price

[jbb]

Powder cores generally have higher core loss than gapped ferrite. Which can be just fine depending on design.

Where powered cores are great is that they have distributed gaps. This means that you can use foil windings for high current / low turn count. Foil windings really don’t like cores with normal gaps due to eddy current losses around the fringing field.

For frequencies of, oh, 30 kHz or lower, amorphous/ anocrystalline iron can be quite good (allows high flux density, but watch out for core losses and accoustic noise).