[Solved] Coil electromagnet geometry effect on field

[igendel]

Hi all,

Let's assume I have a magnetic field sensor (hall effect), and a hand-made copper-wire coil at a fixed distance from it (i.e. fixed distance from the sensor to the nearest part of the coil).

Except for number of windings, and voltage/current, is there a way to shape the coil itself so that the sensor will "feel" a stronger field? Will an iron core help? nail or plate? tight windings or spread out? a reasonable geometry to make the field "directional"?

I've been searching for information about electromagnets on the web and here, but couldn't find specific details about the effect of coil geometry, and my math isn't strong enough to calculate it myself...

Thanks in advance!

[Kleinstein]

Without an iron core there are mainly three shapes of interest:
1) a simple coil, about as long as the diameter - this gives the strongest field in the coil, but going down to the sides quite fast. Its a little hard to calculate but there are approximate formulas.
2) the long cylindrical coil: here it is easy to calculate the field, as just current times turns divides by the length. In the canter its reasonable constant. At the end of the coil it drops to half the strength.
3) The so called Helmholz coils. Two short coils with a distance equal to the radius. The field inside is rather homogeneous, but not very strong.

A iron core can help to make the magnetic field stronger. The simple case is a core with an air gap and constant coss section every where. Unless the air gap is very small (e.g. less than 1% of the core length) the filed is calculated just the same as with the long air coil, only the length of the coil is replaced by the thickness of the air gap. So there is more space to put the windings. As they can be anywhere around the core.

[igendel]

A iron core can help to make the magnetic field stronger. The simple case is a core with an air gap and constant coss section every where. Unless the air gap is very small (e.g. less than 1% of the core length) the filed is calculated just the same as with the long air coil, only the length of the coil is replaced by the thickness of the air gap. So there is more space to put the windings. As they can be anywhere around the core.

Got it, thank you! I'll try it as soon as possible.

[T3sl4co1l]

You get the most field, at a point away from the end of the coil, by putting the most coil towards that end.

That is, a ring winding where all the wire is piled up in the same place, as much as possible (given that the amount of wire is limited by how much field you want to get from it, and how hot you can allow it to get..).  You don't want a pancake (the turns span too much radius), nor a solenoid (the turns span too much length).  A dense ring or hoop or torus is best.

The best radius will be on par with the distance from sensor to coil.  The coil can be made larger than this distance, which will give somewhat better field (it looks like the sensor is closer to the middle of the coil, where the field is strongest), but will cost much more wire, and power, for the same strength.  Make it smaller, and less wire is needed, but the field is smaller (spacially), making the field at a distance much weaker.

Volts/amps and number of turns are a free variable. Simply use whatever size of wire (and number of turns) gets you the required power dissipation and field strength.

Tim

[igendel]

A dense ring or hoop or torus is best.

Thank you too - now, if I understand Kleinstein's answer above, putting an iron core in a dense hoop (at the center of the circle, not along the wire) will not help much, because the hoop's "length" is quite small to begin with... is that right?

[T3sl4co1l]

With a core, you don't need the large ring; that can be replaced with a long rod instead.  In that case, the rod needs to have length comparable to the remaining distance.  (As long as most of the magnetic field from the coil goes into the core, it doesn't much matter what the shape of the coil is.)

You can get further advantage by tapering it to a point, so the field becomes increasingly intense towards the sensor, though you still won't do too much better.

In either case, the rod needs to be large enough around to carry the flux inside. All core materials are limited in this, so you need enough cross section to handle it (otherwise the flux leaks out, and the effective core length is shorter than you were expecting).  It also needs enough dimension to keep the reluctance down, otherwise the flux leaks away anyway.

Tim

[Kleinstein]

To make the field extend to one side, something like a rod as a core is good. A pointed tip is good to get a high strength right at the tip, but the field if very localized. So keep that end flat if the distance is larger than about half the core diameter.  For the other end, it can help to add a kind of plate to spread out the field more even and thus reduce the field at that end. For the winding is does not matter much where they are on the core, so a tight spacing over most of the length is usually best.

Something like a EI core transformer with the core cut to T shape could be a relatively good form.

An other alternative form would be U shaped or horse shoe form. So that the distance of the ends is somewhat larger than the distance to the sensor. The field direction will be perpendicular though.

[igendel]

Okay, it's time for me do lots of experiments  Thanks again!

[f5r5e5d]

you could also add a iron/low carbon steel concentrator to the hall sensor too

free FEM for magnetics http://quickfield.com/qf_editions.htm could help - the student version worked for me but there is a learning curve and model node limits cut in much quicker than you might expect

probably need some physics, a few books on magnetic device/circuit design, experience with CAD and sim software...

[CatalinaWOW]

Might be faster and easier to get a few nails, some iron filings, some white paper and couple of magnets and start playing around.

It doesn't sound like you need a precise measure of field strength, just a way to understand how fields orient themselves around coils and how those fields are changed by highly permeable materials like iron.

Another option would be to describe what you are trying to accomplish in more detail, along with any limitations you have on your implementation.  Then some of the experience hands can suggest approaches.  These might guide your experimentation and reduce the time required to get where you are going.

[igendel]

Might be faster and easier to get a few nails, some iron filings, some white paper and couple of magnets and start playing around.

That may be a good approach for my needs... except I'm afraid that iron filings flying around by mistake will be the end of my little home electronics lab 

This electromagnet issue is neither a big nor a critical part of my application. I only need the occasional short pulse as a means of crude communication, and already I have a reasonably working solution. It's just that the whole thing is powered by battery, and I was wondering whether that's a more elegant way to achieve larger range than pumping in more power.

[CatalinaWOW]

If you are detecting if the coil is energized how about acoustic or optical transmission?

This question may or may not have anything to do with what you are trying to accomplish, but does indicate how hard it is to answer simple questions with no information.

[igendel]

If you are detecting if the coil is energized how about acoustic or optical transmission?

This question may or may not have anything to do with what you are trying to accomplish, but does indicate how hard it is to answer simple questions with no information.

I know this problem is extremely common in beginner forums, but you'll just have to trust me on this one - this is not a case of an asinine approach to a problem that can be solved in another way 

 This is part of a project for a commercial client so I can't give full details; however the electromagnet and sensor are the starting point, and all I'd like to have is better efficiency for them.

[f5r5e5d]

but there must be some motivating limits that are simple to describe that don't tell anything about the whole product

you need to "communicate" across some gap? - what dimension(s), constraints like non magnetic physical barrier

does the coil and detector have relative motion, is motion the actuation principle? or turning on/off the coil current?

what sort of interference do you have to reject, signaling rate/bandwidth, energy budget, mechanical volume, price...


abstracting the "problem description", especially paring it down to actual functional essentials, rather than assuming a particular technology is "the solution" is a valuable skill for a designer

[igendel]

abstracting the "problem description", especially paring it down to actual functional essentials, rather than assuming a particular technology is "the solution" is a valuable skill for a designer

Again, there's no need to press this issue. This is not my first project, I got good answers, everything's cool 

[f5r5e5d]

just hoping you would actually model good practice for the rest - and its hard to go into magnetic circuit design with such vague "requirements"

[igendel]

...so, thanks again to anyone who answered here, sometime things are as simple as they seem, see you on other posts.