Laboratory electromagnet

[opster]

Hello,

I need to make a laboratory electromagnet that can produce an induction of ~1T over a 1"x1" area with an air gap of about 1" .

I have a small budget ~$1000. Which is why I think I'll have to make it. Can somebody give me a few pointers on what iron I should use? I know it should be of lowest carbon content as possible, but I'm afraid I won't be able to find low carbon iron at an affordable price. What commonly available steel could I use in it's place?

Also has anyone wound a coil on there own without a winder? How much should I expect to pay to have a coil wound?

Thanks for any ideas

 

[opster]

Mike,

I think you meant 18 pounds per coil and roughly 14000 turns per coil, right? Else I'll need to get some more wire.

Saul

 

[MJR2]

For an indiviual coil weighing about 18 pounds or a total of 36 for two coils.

#14 wire
110 watts
28 VDC
3.93 amps DC
Series connection
1.75 ID
6 length
4.38 OD
1764 turns
6931 NI per coil
13862 NI total
If you fill the area you described it will take more wire.

Mike

 

[opster]

Mike,

Sorry, my fault. I checked my spreadsheet and realized I used 10% oover the gauge diameter (for insulation) and then an additional 15% for random wound so that's how I got 18lbs per coil. Oh well, if the magnet doesn't end up being strong enough I'll get some more wire.

 

[opster]

I'm done!

Everything went well. I ended up with 14.5gauge wire rated at 200C. The vendor wasn't going to have 14gauge in for a few weeks, so I settled. For a recap:

Structure (frame):
"As I mentioned before I have a space restriction that limits the pole center to be 4.5" off the ground. What I ended up with was 1.5"x4" cross section steel for the base and two yokes (I think that's the right terminology) and poles with a diameter of 1.75" that taper at 60 degrees to a 1" diameter. The poles were made long enough to accommodate coils each of 6" in length and an OD of roughly 6", with a maximum air gap (between poles)around 1.0."

Coils:
Bobbins (2 of them) were ~6in in length, 1.75in ID, and about 6.0in OD. With 18lb (18lb per coil) of 14.5 gauge wire The OD of the wire wrapped on the coils turned out to be about 4.5in. So I have room to add wire if I ever need to. Also I used a lathe at 25rpm to wind the coil, and made a counter with a cheap calculator and a reed switch. The layers were packed well, except at the ends of the coil where I would stop every 2 layers and put some fiberglass tape down to even things out.

Results:
With an air gap between the poles of 0.9in :
1amp 0.2T
2amp 0.35T
3amp 0.45T
4amp 0.50T
5amp 0.55T
6amp 0.6 T
Field uniformity is better than 90%.

The coils in series have a Res of 9.1ohms cold. I'll probably be running the magnet around 3 or 4amps for my experiments with the rubber.

Cost:
The total cost was $1200, $500 for the magnet wire, and $500 for a used LHP100-10 power supply (100V@10amps, 1000W), plus $200 for misc. hardware, tape, etc..

What I would do different:
I think the steel I used wasn't a good choice. I used it because it was free.

For my situation I'm glad I built the magnet, if I had bought something off the shelf I would have had to modify it or make other fixtures to get everything to line up right. It took me around 40hrs to build.

 

[electricpete]

Just out of curiosity, would a permanent rare earth magnet pair have met your needs? It seems like you could achieve that 0.6T over that volum cheaper that way.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[opster]

I don't think so. I have two 1in diameter by .5inch length NdFeB magnets and they would give around .4T over that gap, but the field is not uniform at all, a lot stronger in the middle and weaker as you head towards the edge (by about 50%). I probably could have designed the circuit to eliminate this, but that seems like a lot of work and a little mickey mouse since I want to be able to adjust the strength.

 

[MJR2]

You're getting quite a bit of saturation somewhere. But you got the field you want so it doesn't matter.

Looks like you accomplished quite a bit here. Well done.

Mike