Building electromagnet
January 11, 2011 2:56 PM Subscribe
I'm looking to build an electromagnet for a magneto-optical system I'm constructing. I'd like to generate a field of ~5,000 Oe and I don't think a simple Helmholtz Coil or solenoid will cut it. Where can I read more about different types of electromagnets and how to build them?
Lindsay Publications has a book that includes lots of stuff on induction magnets - not sure how much info is there, but it might get you where you want to go.
Also, putting "Coin Shrinker" into Google might get you info you want.
posted by Kid Charlemagne at 5:18 PM on January 11, 2011
Also, putting "Coin Shrinker" into Google might get you info you want.
posted by Kid Charlemagne at 5:18 PM on January 11, 2011
Not sure how far into this you are but as I understand the laws it's current that builds the strength of the field, not voltage. I've been experimenting (home hobbyist, not a scientist) with very low electromag fields and the current was the chief problem. I ended up with a power supply hooked up to a strong motor from a squirrel cage unit to complete the circuit. Pain in the butt, and since field strength decreases exponentially with the distance from the "core" it's tricky to figure out the needed strength. I had it easy since I'm just using straight wire electromags.
If you can get the current high enough the next problem is melting the wires. I tried a bunch and was surprised by how quickly they glowed red hot and split apart. It seems to be a function of the wire thickness and thermal conductivity, so a thick copper or silver wire is best but the thickness works against the field strength because you can't wrap it as many times as with thinner wire. That's why they use cooling systems I guess. Rather than going with one type of material for the wire they also make wires shielded in other metals so you could split thermal from electrical conductivity if needed.
In short, unless you can find a good resource you're going to have to experiment a lot so leave plenty of time (months, not days).
posted by jwells at 6:01 PM on January 11, 2011
If you can get the current high enough the next problem is melting the wires. I tried a bunch and was surprised by how quickly they glowed red hot and split apart. It seems to be a function of the wire thickness and thermal conductivity, so a thick copper or silver wire is best but the thickness works against the field strength because you can't wrap it as many times as with thinner wire. That's why they use cooling systems I guess. Rather than going with one type of material for the wire they also make wires shielded in other metals so you could split thermal from electrical conductivity if needed.
In short, unless you can find a good resource you're going to have to experiment a lot so leave plenty of time (months, not days).
posted by jwells at 6:01 PM on January 11, 2011
Half a Tesla is getting into NMR/MRI territory. The magnets I have used with these strengths have been liquid-helium cooled copper coil electromagnets (copper superconducts at this temperature, which is handy because the wires quit doing the objectionable things you're mentioning.) The helium is kept cool because its vessel is immersed in liquid nitrogen. These sorts of magnets do have to be structurally quite sound because the back-EMF, eddy currents, etc can be sizable if they are not perfectly rigid.
The larger bore sizes get expensive extremely quickly - the human-sized magnets can go through $50,000 worth of cryofluids a month. There are cryo-free systems but they have miniscule bores and are also quite expensive due to their use of exotic alloys, vacuum canisters, etc.
posted by Protocols of the Elders of Sockpuppetry at 11:12 PM on January 11, 2011
The larger bore sizes get expensive extremely quickly - the human-sized magnets can go through $50,000 worth of cryofluids a month. There are cryo-free systems but they have miniscule bores and are also quite expensive due to their use of exotic alloys, vacuum canisters, etc.
posted by Protocols of the Elders of Sockpuppetry at 11:12 PM on January 11, 2011
To get practcal advice, you need to tell us more. What are the dimensional requirements? Is field uniformity important? How do you plan to drive it? I'll try to remember when I'm at work to ask the magnet experts for good books though.
Protocols of the Elders of Sockpuppetry: Copper doesn't superconduct at any temperature. Superconducting magnets are usually wound from Niobium or one of its compounds (NbTi, NbN or NbSn). For extremely strong fields (in magnetics research facilities), they sometimes use copper coils to complement the superconducting ones, as superconducting coils cannot produce arbitrarily high fields.
posted by springload at 6:03 AM on January 12, 2011
Protocols of the Elders of Sockpuppetry: Copper doesn't superconduct at any temperature. Superconducting magnets are usually wound from Niobium or one of its compounds (NbTi, NbN or NbSn). For extremely strong fields (in magnetics research facilities), they sometimes use copper coils to complement the superconducting ones, as superconducting coils cannot produce arbitrarily high fields.
posted by springload at 6:03 AM on January 12, 2011
Sorry, I asked the people who wind magnets around here, and they say they didn't find any books about it.
posted by springload at 10:13 AM on January 14, 2011
posted by springload at 10:13 AM on January 14, 2011
What I can say though is that electromagnets aren't really that sophisticated unless they have strong requirements on field uniformity or tailored field gradients, or strong enough field to require forced cooling. Solenoids and Helmholtz coils are what you have to work with. Copper is the commonly sued material at room temperature, varnished wire is fine, just don't get it so hot that the insulation starts to go bad. The driving may be a bigger problem than the magnet itself, mostly because of the strong voltage kickback when you turn the power off.
posted by springload at 2:48 PM on January 14, 2011
posted by springload at 2:48 PM on January 14, 2011
This thread is closed to new comments.
posted by clarknova at 3:27 PM on January 11, 2011