I am trying to design some toroidal electromagnets. 1

[HughS]

I am trying to design toroidal ring electromagnets (radially magnetised) - but am having considerable problems with finding a suitable core material - pure iron is prohibitively expensive, and the core is something that could be produced easily with powder technology - but I find I am batting zero with attempts to find a suitable material/process. Laminated cores are not appropriate.

Any ideas - materials/suppliers??

HughS

 

[eromlignod]

How about ferrite?

Don
Kansas City

 

[Laplacian]

This company has readily accessible application engineers:

http://www.mag-inc.com/

I don't think they do one-off designs though. May be worth a check.

 

[MagMike]

What do you need the electromagnet to do? A radially oriented toroid is fairly unusual. Why wouldn't plain low carbon steel work? You mentioned that laminated cores aren't appropriate which means you might have a low frequency application. Low carbon steel (annealed) might work well in that instance.

Mike

 

[HughS]

Thanks for the suggestions - the application is for a linear actuator - and it is coaxial. This is why the need for the odd electromagnets. Plain low carbon steel will probably be OK - as it is a low frequency device - but trying to get hold of 3" latheable bars in Australia is a bit of a challenge - any ideas?? (lots of mild steel - but I think that has too much magnetic remnance for this job). The other possibility is powder metallurgy - but I am trying to get a prototype built! - and I suspect set-up costs may be a bit horrendous.

keep the suggestions coming! and many thanks - I'm impressed with this site!

Cheers Hugh

 

[eromlignod]

The mag-inc.com site suggested by Laplacian includes standard sizes of torroids, which appear to be available in boxes of 24 in approximately your size. This might be worth looking into.

Don
Kansas City

 

[MagMike]

My suggestion of low carbon steel may not be suitable. Here in the USA, Low Carbon Steel is also considered to be a mild steel.

If a particular material has too much remanance for your job, couldn't you just use less of it, to compensate?

Otherwise, Don's suggestion of standard size toroids is a good one.

 

[HughS]

Hi - thanks for the responses..

Unfortunately conventional toroids won't work -

the magnetic axis has to be radial - which means that a coil that is effectively circumferentially wound, has to be inserted in the toroid-like form. This is what makes things difficult. We've figured out the topology - but stuck on the materials and fabrication - I can machine the bits out of bar stock - assuming I can find it (250 dollars - never mind the freight) for a 1' bar of 3" pure iron. - cost about 500 bucks to freight it here! - and i'd need about 3 bars. - so sintered powder metallurgy would appear to be the only way to go - but- what materials - and who could do it.

We may be '1st world' here - but I sometimes wonder!

Thanks - all suggestions welcome

Hugh Spencer

 

[Laplacian]

Maybe you can find a supplier for the powder names listed here. They also have application information.

http://www.amidoncorp.com/aai_guide_to_powder.htm

or here

http://www.coilws.com/Powder Core Inductor/PowderCore_Guide.php

 

[eromlignod]

When I first read your original post I was going to point out that a toroidal electromagnet won't work (the magnetic field is contained entirely within the core), but then I noticed that you were saying "radially magnetized".

What exactly does the core look like? I'm picturing something like a giant lockwasher with protrusions all about the circumference, each one having its own winding...like a motor stator. Is that what you're talking about?

Don
Kansas City

 

[HughS]

well - it is effectively the same thing - but it is for a linear actuator - so each magnet ring has to be equivalent to the radial magnets on a Fisher and Paykel 'smart drive' washing machine motor (bloody fantastic bit of technology those!) - so the solenoid pole facing the armature magnet has a unipolar field. Imagine a donut with pink icing (N) on the circumference and blue icing (S) around the hole.

This is why I'd kill for an e-mail friendly (linked) sketch pad - sort of like the old Mac Draw vector drawing program - the whole application is 50K - and I'm sure some smart cookie could re-visit it and make it work on a wider platform. (If wishes were horses......) On forums like this it would be invaluable - 'Hey Joe - check out this sketch - have I got the bias resistor right on Q3??") !!

[img]http://www.mysite.com/images/thumbsup2.gif[/img]