I am doing some experiments with pulsed magnetic fields. I am using an audio amplifier as a power source, and a simple software as my frequency generator. I use mostly square waves and sometimes sine. Both straight forward, simple waves - sort of like a clean audio test signal of different shape. Nothing fancy. I can even hear them when I run them through my coil and when it is noisy around here, I can just put on headphones and hear clearly whether the coil is working.

I am terribly uneducated when it comes to all things electromagnetic (I only know a bit about audio and that electricity going through a wire will create a magnetic field)… so I need some help. On the upside, I am a fast learner, given right direction.

So far I have been experimenting with a coil I built, that is 4" in diameter, about 1" wide and has about 1lb of 22 gauge insulated copper wire on it (I think that is some 300ft but I don't remember any more) - as a result, it is 8 ohms resistance which was done on purpose as my amplifier is rated at 8 ohms at max power. In any case, the coil works quite well, down to about 300 Hz, and up to 5000 Hz although it tends to get hot around 300 Hz after 10 minutes or so. Haven't tested it lower than that, except today at 20 Hz when it overheated in a second and I had to shut it off. Since I am using an audio amplifier as a power source, I made sure that the coil is 8 Ohms resistance, as that's what the amplifier requires.

Now, I have two questions:

  1. If I need a coil that can take low frequencies, like 20 Hz, the one that I have does not seem to like it, as it immediately gets extremely hot - within seconds literally. It is well varnished (I did it by hand, on every turn) and quite solid, but I imagine that such low frequency makes it vibrate enough to produce high heat. I tried even at very low power setting, such as 1/10 of the max that I normally use, but still it gets very hot very quickly and I am sure that if I let it run, it would melt in about a minute or two. Would using thicker copper wire be better for lower frequency? If so - instead of using 22 gauge wire, what should I use to be able to run that coil as if it is on higher frequency? And if thicker wire would work better, would that somehow weaken the magnetic field? (right now I can run 1000 Hz through it no problem, for hours on end, at max power which is around 1000 W maybe even a bit more). I imagine that since my power amplifier is rated at 8 Ohms I would just use thicker wire but with lower number of turns to achieve the same resistance. If it matters - I am using a Mackie professional power amplifier, the type that is used for concerts. It is rated at 1500 W when in bridged mode (both channels bridged together into one).

  2. How can one achieve an even higher power of the magnetic field using, say, an audio amplifier as a power source? I imagine, getting a more powerful amplifier would be a good start - perhaps a 2000 W or more. But then, how do I build a coil that will not overheat?

Any general advice you can give me in that respect is much appreciated. I think, but I am not sure, that thicker wire is better suited for lower frequencies. Having lots of power, as in - strong magnetic field, is essential. I just don't know how to do it. I understand that there is a practical limit at which coil will necessarily melt, but I don't need to go that far.

One more question: I made an observation that - the lower the frequency the stronger the magnetic field. Am I correct there? Namely, I run 1000 Hz, and I can hear some of the iron stuff on my desk vibrate - from my Luxo desk lamp to the box cutter knife, but it does not attract anything, and can run like that for hours… but at 20 Hz, that same knife stuck to the coil at 1/10th of the power (remember - the coil has no core - it is pure copper)! It was scary for a moment - I imagine that's what happens when someone brings metallic objects into an MRI machine room…

  • 2
    \$\begingroup\$ "use thicker wire but with lower number of turns to achieve the same resistance" - Resistance works the other way around: thicker wire has less resistance per meter and you'd need a longer wire for the same resistance. The magnetic field (and with it the reactance) is basically proportional to the current and the number of turns, so more turns at a give current = stronger field, more current at a given no of turns = stronger field. \$\endgroup\$
    – JimmyB
    Jun 8, 2015 at 10:35
  • \$\begingroup\$ How much B field do you need? What size area? You can use soft iron to concentrate the field. (For B fields less than ~0.5 - 1 T or so.) As other's said, way too much power. If you have a 'scope you could look at the current and voltage waveforms and get an idea of the inductance. \$\endgroup\$ Jun 8, 2015 at 15:51

3 Answers 3


Holy Cow! You're running a kilowatt into a 4-inch coil and you're wondering why it get's hot? Please, rethink this one.

The reason you can run at higher frequencies is that the impedance of your coil is equal to the resistance plus a term that increases with frequency. Look up impedance and inductance. As your frequency increases, so does the coil impedance, and the total power drops. At low frequencies this doesn't apply much, and that's why you're getting hot.

Increasing the wire size is not going to help. What you need to do is something like this. Wind your coil on a piece of 4" diameter copper pipe, and solder copper tubing all the way around both ends of the coil form, as close to the windings as you can. Then run water through the tubing to get rid of the heat.

Remember, if your amplifier is putting out 1 kilowatt of power, you have to get rid of that power. If you don't heat something like cooling water or air, it all goes into the wire. Just think about how hot a 1 kilowatt light bulb would get.

  • \$\begingroup\$ As a specific answer to that hypothetical, the metal housings on my theatre lighting fixtures with 1000W tungsten globes reach 100ºC easily after some usage. The filaments glow at approx. 3000ºC (~3300K). \$\endgroup\$
    – cortices
    Jun 9, 2015 at 12:25

This is all expected behavior, because the coil has inductance as well as resistance.

At DC, you measure 8 Ohms. But at AC, the inductance adds to this. You have a series RL circuit:

\$Z_c = R + j \omega L\$

I guess with the coil you describe, the inductive reactance probably dominates from 10 Hz or so. After that, each doubling of the frequency will double the impedance of the coil, and the current will drop by two times too.

Try an experiment with your multimeter. First, turn the audio level down, that amplifier sounds scary. Connect the meter in series with the coil and drive it at 20 Hz, then increase the frequency to 30, 50, 70, 100, 150 Hz and observe the current decreasing. This is the effect of the inductance increasing the impedance of the coil as the frequency increases.

If it's important to have the same current in the coil at different frequencies, you'll need to apply a correction to the input signal to compensate for the filter effect of the coil inductance.

If you are finding that the amplifier can't produce enough voltage to drive enough current you need through the coil, you could consider a few things.
A transformer would work to raise the voltage, or put another way, it would transform the high coil impedance done to something more suited to your amplifier. A normal mains transformer might work, you're in roughly the right frequency range.
Use a series capacitor to compensate for some of the inductance. This will make a much more narrow-band circuit, so it'll only work at one frequency, but you could completely remove the effect of the inductance at that frequency.
All of this assuming you can solve the overheating problem, with cooling or by keeping the pulses short.

One last thing to keep in mind: the circuit doesn't see square waves, it only sees sine waves. When you apply a 50 Hz square wave, you're really applying a series of sine waves at 50, 150, 250, 350... Hz. The circuit responds differently to each of them, it filters out the higher ones. So the current in the coil will be a very rounded-off square wave.


OP is clearly taking it back to 1990, where you use to be able to walk into any radioshack, and scoop up a "supercoil" for 10.00 each. If memory serves me correctly, those things always burned up when you ran more than 500 watts through them. That being said, how much power are you running through it? I also feel your pain, due to the fact I just spent a decent amount of time on radioshacks site looking for one and if you query a search for one, 99% of the returns are for ignition coils. lol Tho, I feel this might help you. https://www.parts-express.com/cat/air-core-inductor-crossover-coils/297

Since we're on old school setup, might want to consider making a bracket that allows a fan to be spaced off the MDF, about a half inch of space and then the inductor coil above the fan, this will help keep those temps down...

Either way, good luck...


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.