Ever tried the type 0 toroids?

[Conrad Hoffman]

IMHO, the ability of toroids to contain their fields is vastly overrated. Everybody just keeps repeating what they've heard; sure it's better than a solenoid wound coil, but by no means radiation free. It's more different than an E-core iron transformer, rather than wildly better, and you can still get into trouble if the location is poorly chosen. The type 0 is, as said above, non-magnetic. Plan on using a lot of wire!

[Saturnus]

As said before a type 0 ring is unmagnetic as is implied in it's name (inductance x0), and is as such just a very expensive curtain ring. If you want to experiment with this I suggest visiting the local home depot or similar store and look for some wooden or plastic curtain rings in a usable size, no need to pay 3 more middle men a kick back.

[phase_accurate]

Soneone asked wha one shouldn't wind multiple layers onto one core. Well this way that inductor's unavoidable parallel capacitance would get too large.

If you want to avoid stray fields as much as possible and achieve a highly linear inductor then take an air-core with a magnetic path around it. One possible example would be an RM core with a huge air gap. This should be possible with an ordinary RM core, some patience and a Dremel for instance.

Regards

Charles

[wg_ski]

Originally Posted by Conrad Hoffman

IMHO, the ability of toroids to contain their fields is vastly overrated.

Yeah - the reason toroids contain their fields is because the flux stays confined and had nowhere to come out if the "end". That only works if the confining medium has a high perm relative to air. Use air and it will leak! With even windings all the way around, far-field cancellation might be a bit better than a solenoid in some orientations, but not others. Locally, in the circuit next to the coil, it's like an air core coil.

I would expect the -2 mix material usually used for class D to have higher EMI than mix 26. 10 perm is better than air, but there will still be fringing flux.

[Pafi]

That only works if the confining medium has a high perm relative to air.

This is not true. If you have a toroid inductor, with well distributed turns, with air core, and close, parallel "legs" the magnetic field must be cylindrical symmetric. For example, in the equator plan: every individual turns have a magnetic force which is paralell with the equator's plan. So if there were non-zero result, then it would have been tangential. But applying Ampere's law to this says that it requires a current which flows in the direction of the axis, wich is not true. In similar way it can be proven for every points that outside the coil none of the components can be different from zero.

The leakage comes from imperfections, for example when you wind the turns in a way that it goes around along the core, then you make a single turn choke in excess the toroid. Not equal distances are the other source of leakage.

[Saturnus]

Originally Posted by Pafi

The leakage comes from imperfections...

Problem is that imperfections are unavoidable because the very nature of winding a wire around a toroid will cause imperfections.

[Pafi]

Yes, unavoidable (at least entirely), but

- these are secondary effects, and much weaker then with a solenoid
- if you know about them, you can decrease them. If you ignore them, then they can be bad.
- if you know about them, you can get ready for it.
- the question was if the low leakage is caused by the magnetic core, or not. I answered to this question. (Not.)

[sendler]

Rather than waste time and shipping costs on the 1.5 AL type 0 cores I will wind some air core toroids to see about the remaining leakage. It should be easy to do, starting with an over value cylindrical single layer coil and shaping it into a toroid. Local leakage hasn't been a problem as the solenoidal air core inductors I'm using as a reference sound really excellent in practice. So the next hope is to keep the sound but reduce the far field radiation to a level that can be contained in a metal enclosure.

[ttako]

Originally Posted by Pafi

Yes, unavoidable (at least entirely), but

- these are secondary effects, and much weaker then with a solenoid
- if you know about them, you can decrease them. If you ignore them, then they can be bad.
- if you know about them, you can get ready for it.
- the question was if the low leakage is caused by the magnetic core, or not. I answered to this question. (Not.)

Hi Pafi,

I sent you two emails... Did you get them?

Thanks,

Tamas

[Nikolas Ojala]

If you are worried about the small leaking magnetic flux, then make a sheet-iron casing for the coil. Leave some air between the coil and steel sheet.