Air gap of a ferrite core. How?

[imtiaz9]

I want to ask how to do the air gap for a E type ferrite core, by myself?
I know the stock ferrites can be order with the air gap already made, but it is not an option for me, as my supplier has only non-gapped cores.
The stock gaps are in the middle leg, but I have seen DIY project using spacer on the all three legs, but this will increase the EMI, right?
Also, if a need a 4mm air gap in the middle leg, how much should it be, when I divide it among all of the legs... may be in half ( 2mm) ??

 

[Starichok51]

yes, if you want to complete a non-magnetic gap of 4 mm, it is necessary to divide it by 2.
spacer enough to lay just under the two side of the core.
yes, gap on the sides of the core will result in some increase the EMI.

 

[imtiaz9]

thanks "Starichok51"

 

[rikkitikki]

Actually , with so large gaps you can not use the formulas directly but have to compensate for fringing effect. This means you have to do some testing with different gaps to achieve your wanted inductance.

 

[KX36]

Firstly, do you need a 4mm gap? are you sure you've calculated that correctly? As rikkitikki says, fringing flux is a source of inaccuracy from the formulae (and heating in any windings that pass close to the gap) and the bigger the gap, the more fringing flux you have.

Simplified inductor design procedure:

Pick a core, then

l_g=mu_0 * L * Imax^2 / (Bmax^2 * Ac)
where
l_g is gap length,
mu_0 is the permeability of free space; pi*4E-7H/m
L is inductance
Imax is peak winding current
Bmax is peak flux density
Ac is cross sectional area of core.
As you can see, l_g is inversely proportional to Ac, so a bigger core requires a smaller gap length and vice versa. The volume of the gap would be the same.

The next step is number of turns
n= L * Imax / (Bmax * Ac)

Then you might need to itterate cores to choose the right core.

Just separating the cores by half the gap length is a common way to get a gap, as you say it increases EMI. You can shield the EMI by wrapping a shorted turn of copper foil around the outside of the whole inductor so that it surrounds the mating surfaces of the ferrites, but I don't know if that's a good idea if you do have gaps at the outside legs of the core. Gapping the centre leg of an E core is better for EMI but requires careful grinding as ferrites are brittle.