Long inductor configuration.

To a first-order approximation I'd say that stacking the cores into one long "bead" and winding that would be best.

Something that's missing from your problem statement is how close to the walls of the tube the wires are going to get. The literature that I've read wants to start by telling you that the magnetic field outside of the toroid is "insignificant" -- but those authors are assuming that you're not going to be putting the thing into close contact with a lossy, magnetically permeable material.

Even if you get well away from the walls, the leakage inductance of a toroid-wound coil is equivalent to one turn around the periphery. If I'm not mistaken, this is a geometrical effect of the way the wire is wound. So there will be some interaction with the tube wall, which will be worse the fewer turns you use.

Dunno the answer -- pot cores? Beads arranged three, four or five in a circle with the wire going through the holes? Ignore my maundering and forge ahead?

--
Tim Wescott 
Control systems, embedded software and circuit design 
I'm looking for work!  See my website if you're interested 
http://www.wescottdesign.com

I'd vote for the pot core approach. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
        NRA and NHRA... the core values of middle America

It's too bad you can't get thin-wall toroids, or ferrite tubes -- you could make up an arbitrary-length pot core.

Dunno if it's worth the machining time, but you could probably have such things made.

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com 

I'm looking for work -- see my website!

look like they have tubes:

There are some tradeoffs, as always.

A two-core inductor has more than twice the inductance of two series inductors each with the same number of turns as the stacked inductor. To get the same inductance with stacked cores you only need sqrt2 (0.707) turns of the separate inductors.

So yeah, you get more inductance/volume by stacking cores for a given filling factor than from stacking individual inductors.

You also have more core-loss heat/volume to dissipate though. Since you said lots of current, that may be significant.

As for coupling to the iron pipe, you've reduced the eight or nine-turn topological "stray" fields of each the seriesed inductors to that from a single turn around the stack.

You could try a lengthwise-slotted (to cut down eddy loss) copper or other diamagnetic liner. If you have the budget, go for bismuth or pyrolitic carbon tubes in the required size and cut your own slot.

Do you *have* to use an iron pipe?

Mark L. Fergerson

Really thick-wall though -- not what you'd expect to see in a pot-core-ish sort of thing.

Of course, you may need absurdly thick walls in a pot core if the aspect ratio gets high. Hmm.

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com 

I'm looking for work -- see my website!

How about a stack of tape-wound cores? A wrap of sticky tape would hold a cylinder together.

10 N-turn cores in series, and a stack of 10 cores with N turns will both have the same inductance.

The stack will have lower DC resistance (due to using about half as much wire as the series group) but the self-resonant frequency will be lower (due to greater inter-winding voltage and capacitance)

If you need the high frequenies maybe an assymetrical set of stacks eg: 6+2+1 will give the best performance for you.

--
This email has not been checked by half-arsed antivirus software

Pot cores would certainly be easier to wind. I suppose stacking them one on top of another would make little difference, but I wonder if they should be separated by a layer of insulation?

Cheers

--
Syd

That's an interesting point. I suppose for multiple separate toroid inductors, the 'top and bottom' portions of the winding would help in sinking heat to the outer portion, whereas for a stack of cores with a single winding, the inside wouldn't get much cooling by means of conduction through copper.

So many variables.

The pipe is a pressure housing holding back 15-20kpsi, so yes indeed.

Cheers

--
Syd

I suspect not. This is all arm-chair reasoning, but if each is already insulated sufficiently I don't see a potential building up across any one pair of cores that's more than "it's share".

--
Tim Wescott 
Control systems, embedded software and circuit design 
I'm looking for work!  See my website if you're interested 
http://www.wescottdesign.com

Just tossing this out without much thought (the caffeine hasn't kicked in yet) -- what about a long skinny circuit board with plain ol' surface mount inductors on it? Make 'em sheilded, and be careful with your return path, and there shouldn't be much AC field to excite the walls of the pipe.

--
Tim Wescott 
Control systems, embedded software and circuit design 
I'm looking for work!  See my website if you're interested 
http://www.wescottdesign.com

I haven't tried it myself but I'm told that stacking dissimilar cores provides a wider passband- one core takes over where the other leaves off. Makes sense if that's what you need, and can accept the losses of each.

Mark L. Fergerson