Introduction: Make a Joule Thief Coil Without a Ferrite Toroid

Picture of Make a Joule Thief Coil Without a Ferrite Toroid

There are many different articles and videos to be found online that will teach you how to build a Joule Thief circuit and power LEDs off of dead batteries. Most of them, however, call for some form of coil made using a ring, or toroid, made of ferrite. This can be rather daunting to newcomers to electronics, as "ferrite toroid" sounds rather impressive and exotic at first.

The truth is that a ferrite toroid coil is just a simple coil with two windings wrapped around a ring made of ferrite (iron). Toroids are often used because the ring shape is very effective at storing electromagnetic energy, but you don't actually need a toroid to make a Joule Thief. A simple coil with two windings around a central core will work just fine.

In reality, ferrite toroids are actually fairly easy to come by, as they can be purchased online from a site like Mouser for less than a dollar or found in the guts of such items as old microwaves and ceiling fan control panels, but just in case your dad won't let you tear the microwave apart and you don't feel like paying shipping on a $0.70 item, I'm going to show you how to make a simple coil for a Joule Thief using a few simple items you can find around the house.

Note: If you have any common sense, you will likely not get hurt in the making or usage of this project. If you do, however, it is you own fault and I accept no responsibility for any harm incurring as a result of your attempt at this project.

The image on the Intro step is not mine. It comes from this video here, which provides a good tutorial on how to make a basic Joule Thief.

Step 1: Materials and Tools

Picture of Materials and Tools

The items you'll need to make this coil are simple:


A length of thin magnet wire. I used about 25 feet of 30 gauge.

A magnetic core. Anything that will stick strongly to a magnet will work. I used a hacksaw to cut the end off of an old drill bit, but a bit of a nail or other iron object will work. You might try the piston from an old sprinkler valve solenoid. Try and make it about a centimeter long, cutting it if you have to (obviously, be careful).

Two buttons. I took these off of a dress shirt. You could also use some little discs or squares of plastic or even carboard if you want. These will serve as endstops to keep the wire wrapped tightly around the core.

Superglue. Shouldn't need to elaborate there.

You might also want to use a bit of adhesive to secure the wire on the coil after you finish winding it. Nail polish topcoat works very well for this. You could also use a dab of silicone or hot glue. I would advise against using superglue, at it will get everywhere and some people claim it eats through the ceramic insulation on the wire.


Needlenose pliers, or small wire pliers. If they don't have a wire cutter built in, you'll need wire snips or a way to cut the wire (I would advise against simply tearing at it like a caveman until it breaks).

Sandpaper. This will be used to strip the ends of the magnet wire. You can also use nail polish remover or a razor blade to do this.

A hacksaw or some other means to cut your core down to size (assuming you need to, of course).

Step 2: Build the Core for Your Coil

Picture of Build the Core for Your Coil

To do this, we're essentially just going to glue the buttons onto the ends of the iron core.

Firstly, take a hacksaw, razor knife (be careful!) or wire snips and make a small notch in one of the buttons. This will help when winding the wire around the core.

Next, use a drop of super glue to glue the button onto the end of the core. Try and get it as centered as possible.

When the glue dries, flip the core over and glue the second button to the other side, again making it as centered as you can.

Let the glue dry, and you have your core.

Step 3: Winding the Coil.

Picture of Winding the Coil.

To start, cut a piece of magnet wire about 20-25 feet long, and bend it exactly in half so that you have a small loop with two longs ends trailing from it. Be careful, as these will tangle and kink very easily.

Place the loop into the notch you made in one of the buttons on your core, with about two inches sticking out. This will make it easier to wind by keeping the wire from moving.

Wind the two trailing wires side-by-side around the core. Wrap them tightly, moving up and down the core so that they evenly cover the length of the metal with several layers. Try to keep the wires from twisting or kinking to ensure smooth windings (not terribly important to the overall function of the coil, but makes it easier to wind and fix errors and helps prevent the insulation from wearing off the wire, causing a short).

When you have about two inches left of the wires, take your wire snips and cut the loop of wire coming out of the side of the coil in half. You will now have two separate wires wound around a single magnetic core. Twist them gently together in order to secure them.

Finally, use sandpaper, acetone or a razor blade to remove the insulation from the ends of the four wires. You can now use a bit of nail polish or silicone to glue them in place and keep the coil from unraveling.

Step 4: Implementing and Troubleshooting

Picture of Implementing and Troubleshooting

You now have a coil that will work in a Joule Thief circuit. Use a multimeter or continuity tester to determine and mark which wire end is connected to which and to make sure there isn't a short that has all four wires connected together. If everything checks out, then connect the coil to your Joule Thief using alligator clips, a breadboard, soldering iron or some other method. This should work fine for a simple Joule Thief circuit, but you can always make one with more wire and windings to get more inductance and more power if you need to.

Step 5: Defy Physics and Power Your Devices With Dead Batteries!

Picture of Defy Physics and Power Your Devices With Dead Batteries!

You get the idea.


alaafprojs (author)2017-06-04

ANYIron core????

dric19 (author)2016-09-15

can i replace the drill bit with few piece of round magnet which has the same size?

El Mano (author)dric192016-09-18

No, it probably won't work very well. For the coil to work, the current in the wire has to create a magnetic field. If you're trying to create a coil using a magnetized bit of metal, the field produced by the current will be fighting the field from the magnet.

GWSC (author)2015-08-29

They will not work without a core made from a magnetic material. Its operation depends on the magnetic material becoming "saturated", that is it cannot have its magnetic field increased any further.

GWSC (author)2015-08-29

I have made a number of these using a TV balun core. Cheap and effective, simply wind using the 2 holes. I have only used the 6mm ones, but the 14mm should also work. The 6mm ones make things smaller and are cheaper. They should be available at most good electronics suppliers. I have put a link so you can see what they look like. Good luck.

richardebueramosii (author)2015-08-07

This is so helpful. Thanks for this! I may sound dramatic but this gave me more hope on making a Joule thief for my personal project. Well, I really am having hard time finding a ferrite toroid... and thanks, now I can finally make the circuit. Kudos!

argha halder (author)2015-06-25

Also I tried this one with air core, an AA battery as the core, and also a screw as the core, and it worked with all of them. (But probably not as good as the drill bit core). So anyone who can not find any nails or screwdriver tip, try using the batter as the core, or use an air core which is basically an empty core. By the way, really nice and clear instructions, so kudos for that!

Wow, thanks for that info!

El Mano (author)argha halder2015-07-24

Thank you.

Georjutty (author)2015-07-22

Should I care about connecting the four wire from the core to which or where parts of the circuits?

El Mano (author)Georjutty2015-07-24

Yes. They have to be wired in a specific way in order for the circuit to work. The easiest way to achieve this in my experience is just to wire the coil in and test it. If it doesn't work, take the two ends of one of the wires (ex, the two wires I marked black with a sharpie) and switch them, leaving the other two the same. I find usually this fixes any problem. You also want to make sure you have your transistor orientated properly, with the collector and emitter connected to the right parts of the circuit. If you're not sure which is which, google a schematic of the particular transistor you're using.

Ploopy (author)2015-07-08


holtt (author)2015-06-20

This has been my go-to Joule Thief circuit. I'm currently making about 100 of them for basic lighting for an event, and couldn't imagine trying to wind that many coils/toroids!

El Mano (author)holtt2015-06-21

No, that would be ridiculous. You'd spend more on wire than on the event lighting.

billgeo (author)2015-06-17

"Defy physics" with a well known and understood circuit from the 1910s!

Yeah... all joule thief instructables should have at least a link to a page

that shows how the thing works!

Its fun to make something, but it's even better if you learn something along the way...

El Mano (author)billgeo2015-06-17

There is an amazing tool called Google that will teach you everything you need to know about this and other circuits. Use it. You can access it at This Instructable was made to address a very specific issue that people have when making joule thief circuits, not to teach how to make them. If you are here reading this, I'm going to assume that at some point you stumbled upon a page titled "How To Make A Joule Thief," and so have a basic idea about what it is and how it works. I'm also going to assume (though of course I could be wrong) that you are literate and internet-savvy enough that if you don't know what a joule thief is or how it works, you can do a bit of research and figure it out. (Seriously, Check it out.)

tomas75a (author)El Mano2015-06-21

Good! Ppl should know how to search for things themselves!

El Mano (author)El Mano2015-06-17

I apologize. That was uncalled for. The point remains, however, that this article is not about the joule thief. This article is about building a coil, not a circuit. I wrote it for people who have already researched the circuit and want to build one but are having difficulty obtaining the traditional ferrite toroid. If a reader is interested in learning about the theory behind a joule thief or wants an in-depth look at the implementation of one, there are several other Instructables advertised on the right hand side of this page that will help them far more than I could. In all honesty, I really don't care if you think this article should have a link to The-Ten-Thousandth-Page-On-How-To-Make-A-Joule-Thief, as I think most intelligent people can simply navigate to one themselves if they're interested. In fact, I only included the link to the YouTube video on the first page because I used his photo as my cover and I thought he deserved credit for it. As for your comment about defying physics, please don't come here and ruin the party by playing the I'm-Intelligent-And-I-Know-Science-Game. This isn't the lecture hall. We get that there is an explanation behind it. We also get that it's really freakin' cool, and we don't need a stuffy know-it-all taking the fun out science.

chintans (author)2015-06-10

sorry for double comments, will delete when online from pc. Have a question in mind. Can I use the neodymium magnets I got from my old dvds lens part? Its currently in its casing and its square. So need your advise on if I can use it.

El Mano (author)chintans2015-06-10

I don't think it would work as a core. The joule thief works by using current to create a magnetic field in the coil, which is then used to induce current in the second piece of wire on the coil and open the transistor. The magnetic field it creates has to be able to change constantly in order to function, and if you try to use a permanent magnet as a core, especially such a strong one, my guess is it will fight and neutralize the weak magnetic field created by powering the coil. Imagine trying to affect a moving freight train with a push from your hand: if you're pushing in the same direction the train is moving, you're likely not going to make any noticeable increase in it's speed, and if you're pushing against it it's going to run right through you. In the same way, I think that when you power the coil, the magnetic field from the neodymium magnet would completely overpower the field created by the circuit.

chintans (author)El Mano2015-06-11

Thanks buddy.

Thats awsome information.

I have couple of neodymium magnet.
Wondering any real world application.

MakeBromyard (author)chintans2015-06-19

Glue one to your dead battery and make a (relatively) giant LED throwie that uses a cheap dead battery and lasts months!

LegendofPedro (author)El Mano2015-06-11

Yes, your core should be magnetisable (large permeability), but non-magnetic; as El Mano says, when the magnetic field vector changes direction the aligned magnetic domains will resist the change, resulting in hysteresis losses in the core.

zacker (author)2015-06-11

what are these coils used for?

El Mano (author)zacker2015-06-14

I made it for a joule thief, but it's just two coils around a shared core. You could use it for anything you needed an inductor for, even an electromagnet.

ElectroFrank (author)2015-06-11

Cutting a 1/4 inch drill ? Drills are made of very hard metal ! Ordinary nails or bolts (or just about anything else steel) will be easier to cut.

El Mano (author)ElectroFrank2015-06-14

That is true. I used a drill bit because that's what I had and it stuck more strongly to a magnet than some of the other bits of trash I had lying around. It was a pain to cut, I admit, but it works fine for the purpose.

Sembazuru (author)2015-06-12

In the materials section you mention almost off-handedly that the magnet wire has ceramic insulation. Not true. It's enamel insulation, basically painted on so it can be as thin as possible so when winding a magnet the loops can be as close as possible.

Carefull using nail polish to fix the coil. The solvent in the nail polish may attack the enamel of the wire insulation before it volitilizes off.

El Mano (author)Sembazuru2015-06-14

You're right, it is enamel, not ceramic. I didn't have any problems with nail polish, however. I made one using super glue that ended up shorting, though I don't know if the glue was the cause. Nail polish topcoat worked fine for me, however.

ElZorro (author)2015-06-11

Sorry for the dbl
The reason the spikes get lost is due to their brief duration and the esr of the battery which gets much higher as it drains thereby decreasing efficiency the longer it runs. By pass this and you get to keep the spikes. Which are actually much higher voltage than the fwd drop of the led. The power supply filter should need no explaining. But make it too big and you actually start losing again.

throbscottle (author)2015-06-11

I like this - nice inventive take on making the magnetic core from what you have lying around. Of course with the head end of a nail you could dispense with one of the buttons. It also occurs to me that a bundle of thin steel wires (wire twistys, paperclips, flower wire) glued together would work better - less eddy currents. Could be tricky to make though.

ElZorro (author)2015-06-11

And those spikes do get lost.
Just scope it

ElZorro (author)2015-06-11

And those spikes do get lost.
Just scope it

ElZorro (author)2015-06-11

I did not say increasing frequency increases efficiency.
I actually said I have them running at 40hz and at 250mhz.
You are not reading. I also said I used a filter and a bypass.
Those are both capacitors. Again you are not reading.
Using all these methods I mentioned increases my efficiency by over 300%. I also said I hvae them running for years on the same AA.

psychotron (author)2015-06-11

A better and cheaper (also easier to cut) core material would be very low carbon steel, like used in nails, because it has less hysteresis (loss). Get a spike or two from a home improvement store and you could make dozens of these.

colin55 (author)2015-06-11

"There are very fast spikes that get lost." You are confusing the issue by introducing "red-herrings." In this circuit, the frequency is not all that high and the SPIKES ARE NOT LOST.
"Use a low loss fast core a filter and a bypass" another "red-herring."

Ferrite cores reduce to a factor of 1 when the frequency increases. You simply cannot increase the frequency to increase the efficiency.

See Talking website for details on how these circuits work and how to reduce the current by 50% by adding a capacitor - something you failed to mention !!!!

ElZorro (author)2015-06-11

Oh and different gauge wires for pri and sec. Will help also

ElZorro (author)2015-06-11

I guess I should have said
I have proven it myself. :)

ElZorro (author)2015-06-11

Not as fast and more lossy though. Sure you will light your led
There ar very fast spikes that get lost.
Use a low loss fast core a filter and a bypass
You won't lose these spikes and your efficiency increases enormously. I have JTs running over 250 MHz and some running 40hz have them running for years on same batt.
Big superbrites

chuck787 (author)2015-06-11

More documents in the world need the truth of life as in your Note. Enuf said

colin55 (author)2015-06-11

" As others have said you are using a steel core which will be very lossy
in this app. Instead of explaining what ferrite cores actually do and
how they do it. You just used a very inferior core material."

Actually, that is totally untrue.
At low flux densities, steel is just as good in this application as ferrite. That's why I don't say anything until I have proven it myself.

See Joule Thief circuits on Talking website for a steel bolt and air core Joule Thiefs.

colin55 (author)2015-06-11

You can actually use a nut and bolt and it works just as well as ferrite material.
You can also just place 30 turns plus 30turns next to each other without any core material and it works perfectly.
See Joule Thief on Talking website for all these details.

ElZorro (author)2015-06-11

As others have said you are using a steel core which will be very lossy in this app. Instead of explaining what ferrite cores actually do and how they do it. You just used a very inferior core material. You could have even used air.
But then....
Step five
Defy physics
And power your devices with dead batteries.
Maybe you should have (but are probably not up to the task) made an instructable on the now ancient technology of fly back transformers.
I'm going out now to try and "defy physics"
Or at least win back my ex girlfriend.
Wish me luck.

billbillt (author)2015-06-11

great.. very inventive..

BurgersBytes (author)2015-06-11

You don't need near that big a coil with that many loops. There are also 4 pin solar chips that can run using normal coils.

andrej (author)2015-06-11

I love the crocks!

Ploopy (author)2015-06-04


About This Instructable




More by El Mano:Connect To Google Public DNS On Linux MintMake A Joule Thief Coil Without A Ferrite ToroidCreating A Pixel Screen In Batch
Add instructable to: