High frequency electromagnets.

[fritz]

If you apply 1A with a 10V voltage source - you need a (DC) coil resistance of about 10 Ohm.
This magnet consumes 10 Watts permanent and heats up.
On turning on the current will ramp up in a linear function up to 1 Amp.

The strength of the magnet is proportional to your current.
If I start to play with a permanent magnet disturbing my electromagnetic field - you will see that this 1 Amp will start to fluctuate - because we induce an extra current - which might add or subtract from the undisturbed current.


If you want an electromagnet with pretty no power - you have to bring the resistance / voltage down.
If you have a superconducting coil with 0.000001 Ohm and a current of 1Amps - your power consumption will drop from 10 Watt to 100uW - by the factor of 100.000.

If you start to think about AC - you have to keep the following in mind:
*)Losses in the core
*)Extra losses due to reactive power
*)higher resistance due to skin effect...
*)Less current due to inductivity

Additional - you will experience losses in the material you want to attract.
If you exceed the maximum flux density and rated frequency of a core material - the losses may increase dramatically - and the stuff will start to heat up. If we talk about metal this would be kind of induction heating.

So nothing gets better with AC.

[mscoffman]

I just remember that "Magnetic Field Strength of a coil is proportional
to ampere(*)turns".  The way to use less current is to emplace more
wire coil turns. The phase shift caused by the coil means that an AC
signal will be more reluctant to flow into coil above and beyond it's DC
resistance. That means to get more field strength can only occur with
more AC power excitation.  The magnetic poles switch (N/S) with AC.
You can change the reluctance (inductance) of the coil upward by
adding a core material. But there is lowest reluctance - is called the
reluctance-of-free-space which is an universal physical constant.

:S:MarkSCoffman

[gyulasun]

Hi Vidar,

One more thing that just occured to me: Nikola Tesla showed just the kind of electromagnet you may have been looking for:
http://www.tfcbooks.com/patents/512340.htm 

He 'embedded' capacitance between two parallel guided wires and made a coil from the double wires. And obviously the resonant frequency dependens on the mechanical closeness of the parallel wires and the dielectric properties of the insulator materials between the wires and of course the wires length (number of turns).  Tesla wrote this setup is good for coils in general so not exclusively for flat "pancake" shaped coils.
By choosing the AC frequency of the input current to be the same as the resultant resonant frequency of the double wires connected as he showed, the inductive reactance would not "resist" the current flow like in conventional coils. 

rgds,  Gyula

ps: I fully agree with fritz and mscoffman posts.

[fritz]

I just remember that "Magnetic Field Strength of a coil is proportional
to ampere(*)turns".  The way to use less current is to emplace more
wire coil turns. The phase shift caused by the coil means that an AC
signal will be more reluctant to flow into coil above and beyond it's DC
resistance. That means to get more field strength can only occur with
more AC power excitation.  The magnetic poles switch (N/S) with AC.
You can change the reluctance (inductance) of the coil upward by
adding a core material. But there is lowest reluctance - is called the
reluctance-of-free-space which is an universal physical constant.

:S:MarkSCoffman

On switched dc I get the "investment" in the reluctance (while switching on and field establishes) back on turning off (if I´m so clever to recycle the energy from the back EMF on turning off)
But charging and decharging the field causes extra current - which leads to  extra losses as long we have a resistive coil.

If I have a perfect supraconducting coil in vacuum  - I have still reluctance - but no losses -I invest the energy once and can take it back.

Take for example the SMES - where an inductive system is used to store energy. Normally you can do this only with capacitors - using superconductors - you can do the same with coils.

http://en.wikipedia.org/wiki/Superconducting_magnetic_energy_storage

rgds.

[fritz]

To round it up - an alternating magnetic field will induce an em field.
So even if we have a perfect coil we will experience losses due to em radiation.
This is pretty nothing at AC - which is for me somewhere between 10 - 400Hz.
If we come into the Khz range this part will increase.
Depending on the matching from our "magnetic" antenna (our coil) to the free field impedance
(120pi Ohms) this will start to suck up energy.
Typicalwise, electromagnets have a very high inductivity - so aproaching 1Mhz - the only radiating thing
will be the feed wires - and we will end up with a dipole (because the coil is high-Z).
rgds.