Introduction: DIY Induction Heater

Picture of DIY Induction Heater

When you think of a way to heat up a metal object, you think of fire -right? Fire is an inefficient, old fashioned, and slow way to heat up metal objects. It wastes lots of energy as heat and creates lots of dirty smoke. Well, what if you could have a way to heat up metal objects that solves all these issues -it would be great, right? In this instructable, I am going to show you how to build a ZVS induction heater. This is a device that heats most metals using a ZVS driver circuit and electromagnetism. It is very efficient, produces no smoke, and can heat up objects like paperclips in a matter of seconds. The video below gives a demonstration of this induction heater in action, along with another type of instruction on how to build it.

Lets get started.

Step 1: How It Works

Picture of How It Works

Many of you who are reading this may be asking "What is a ZVS driver"? Well, it is an extremely efficient oscillator circuit that is able to create an extremely powerful electromagnetic field that heats up the metal. It is the backbone of the induction heater that this instructable is showing you how to make.

To understand how this power supply works, I will explain the different sections of it. The first section is the 24 volt power supply. The power supply needs to produce 24 volts at a current of 10 amps. For my power supply, I will be using two sealed lead acid batteries wired in series. The power is then fed into the ZVS driver board. The ZVS oscillator pushes and pulls current though a coil around the object that is being heated. This constant changing of the current's direction creates a fluctuating magnetic field. This induces many small eddy currents in the metal(refer to the diagram above). All of these currents are relatively high, and because of the low resistance of the target metal, heat is generated. According to ohms law, power converted to heat in a resistive circuit is P=I^2*R.

Now, the metal type of the object that is being heated is very important. Ferrous metals have a higher magnetic permeability, so they are able to harness more energy from the magnetic field. This allows them to be heated quicker than other materials. Metals, like aluminum, have a lower magnetic permeability, so it takes longer for them to heat up. Things that have a high resistance and low magnetic permeability, like a human finger, will not be heated at all by an induction heater. The resistance of the material is also very important. If you have a higher resistance in the target metal, then less current will flow, so the power converted to heat gets exponentially smaller. If you have a metal with a lower resistance, then the current will be higher, but power loss will be lower due to ohms law. It is a little bit complicated, but because of the relationship between resistance and power output, the highest power output is achieved when the resistance of the object approaches 0.

The ZVS oscillator is the most complex part of this circuit, so I am going to explain how it works. First of all, when the current is switched on, it flows though 2 inductive chokes into each side of the coil. The choke is to make sure the circuit does not draw to much amperage on start up. The current also flows thought the two 470 ohm resistors into the gates of the two Mosfets. Now, because no component is perfect, one Mosfet is going to turn on first. When this happens, it hogs all the gate current from the other Mosfet. It will also draw the drain of that Mosfet that is on to ground. This will not only let current flow though the coil to ground, but it will also let current flow though one of the fast diodes form the other gate of the other Mosfet, locking it off. Because there is a capacitor in parallel with the coil, it creates a resonant tank circuit that starts oscillating. Because of this resonant action, the drain of the other Mosfet will swing back and forth in its voltage, eventually reaching 0 volts. Once this voltage is reached, the gate charge from the Mosfet that is on will discharge though the fast diode into the drain of the opposite Mosfet, effectivly shutting it off. With this Mosfet off, the other Mosfet has the opportunity to turn on. After this, the cycle repeats thousands of times per second. The 10K resistor is meant to deplete any excess gate charge on the Mosfet, because it is like a capacitor, and the Zener diode is meant to keep the gates of the mosfets at 12 volts or under so they do not explode. This high frequency high power oscillator is what allows metal objects to be heated.

Now, its time to build this thing!

Step 2: Materials

Picture of Materials

To build this power supply, you will need a few parts, luckily, most can be salvaged for free. If you have ever seen an old CRT TV laying on the side of the road, pick it up, because it has most of the parts needed for this project in it. If you want higher quality components, you can buy them at the LCSC online store. Click the parts to bring up the product links in LCSC.

You will need:

Step 3: Tools

Picture of Tools

For this project, you will need:

Step 4: Transistors and Cooling

Picture of Transistors and Cooling

In this circuit, because the transistors switch at 0 volts (Hence the name, Zero Voltage Switching ZVS) they do not get very hot, but they should still be mounted on a heat sink if you are planning on running this circuit longer than 1 minute. I mounted both of the transistors on one heat sink. When you do this, make sure you isolate the metal backs of the FETs from the heat sink. If they both touch, it will short out and blow out your FETs. My heat sink was from a computer power supply, and already came with a piece of insulating silicone, so my transistors are isolated. To make sure that your transistors are isolated, touch your multi-meter to the middle pin of both transistors, the drain. If you get continuity, then your FETs are not isolated.

Step 5: The Capacitor Bank

Picture of The Capacitor Bank

In this circuit, the capacitors get very hot. This is due to them always having current flowing thought them. Now, the capacitor value we need to make this circuit work properly is 0.47uF, so we will need the most amount of capacitors together to reach this same value, but have a larger surface area for heat dissipation. You also need to get the voltage rating of them above 400 volts due to inductive voltage spikes in the resonant circuit. What I did, is make a ring of copper, and add 10 .047uF capacitors in parallel around it. This makes the combined capacitor bank have a capacitance of .47uF, with lots of surface area for air cooling. This capacitor bank will be in parallel with the work coil.

Step 6: The Work Coil

Picture of The Work Coil

This part of the circuit is what generates the magnetic field. It is formed using copper wire. It is very important that you use copper. I started this project using a steel work coil. It did not work very well. When it was running with no load, it was drawing 14 amps!! When I switched it out with a copper coil, it drew only 3 amps. I think that this is because the ferrous material in the steel coil had eddy currents induced into it. Its high magnetic permeability made the coil the subject of the induction heating, which wasted the power and stopped it from heating the material inserted. I am not sure if this is the exact reason this wasn't working, but it is the most logical argument based on the evidence provided.

To make your coil, use stranded copper wire, and wrap it around a piece of PVC pipe about 9 times.

Step 7: Building the Circuit Trial and Error

Picture of Building the Circuit Trial and Error

Building this circuit took a lot of trial and error. My number one issue was my original power supply and coil. The power supply is an 55 amp 12 volt switching supply. I think that this power supply drove the ZVS circuit with too high of an initial current, this blew out the mosfets. They exploded, like in the first picture. This probably could have been fixed by adding larger inductors, but I decided to just use lead acid batteries.

My second issue was the coil. In step 6, you saw that the steel coil did not work. This high current draw due to the steel coil blew a few mosfets too. In total, I lost about 6 mosfets to explosion. This may be bad, but I learned from my mistakes.

Over the course of this project, I built the circuit over again many times, but I will just explain how I built the most successful version.

Step 8: Building the Circuit

Picture of Building the Circuit

To build this ZVS driver circuit, you will need to follow the above circuit diagram. I first took the zener diode and twisted it together with the 10k resistor. You can then take this pair of components, and solder it between the gate and ground of the mosfet. Make sure that the black end of the zener diode faces the gate. Then, solder the mosfets to a piece to perf board. Use the bottom side of the perf board to solder two fast diodes between the drain and the gates of each fet. Make sure that the white line faces the drain(pin 2). Then, attach the VCC wire -from your power supply- through 2 220 ohm resistors to the gate of each transistor. Ground both sources. Then, solder the work coil and capacitor bank in parallel with each other and solder each end to a different drain. Finally, run power to the drains of each mosfet through 2 50uh inductors. These can be toroidial cores with 10 turns of wire. With that, your circuit should be ready to use.

Step 9: The Base

Picture of The Base

The base of your induction heater is just to support all of the components. I used a piece of 2x4 wood scrap. The circuit board, capacitor bank, and work coil were all hot glued to the wood. I think this setup makes it look cool.

Step 10: Testing Your Induction Heater

Picture of Testing Your Induction Heater

To power up your induction heater, just connect it to the power supply you have. Then, insert the part you are trying to heat up into the coil. It should start heating. I was able to get a paperclip to red hot temperatures in 10 seconds. Other things, like a nail, took about 30 seconds. With these objects inserted, the current draw rises by about 2 amps. This is a fun circuit to build an play around with. It can also be used very practically. It can heat up objects without any of the soot that comes from smoke. It can even heat up isolated metal objects, like the getter material in vacuum tubes. It is also human safe, so you will not get burned by putting your finger inside the coil. It will, however, burn you if you touch an object that has already been heated.

This heater has many possibilities, tell me in the comments what you plan to use it for!

Thanks for reading!


schreib (author)2017-10-17

What an amazing kid. What. . . ? 14, 16 years old? I am glad I chose mechanical engineering instead of electrical engineering half a century ago . . . I could never have competed against the sparky's-- who have since saved my bacon many a time. Tanner is a perfect example of the cool determination and confidence of top-end electrical engineers. I especially love the part showing the trial and error learning and solutions. Excellent video presentation. and yes, this could be used for knife blades but likely requires longer / bigger coil and other components to handle the amp draw.

smooth_jamie (author)2017-11-16

Nice instructable. However I really wish people would stop using the word "amperage". You wouldn't look outside and say "wow the Fahrenheitage is high today" otherwise I really enjoyed reading. Sorry for pointing that out!

zar011 (author)2017-10-17

Would it be possible to wrap this around a copper/steel pipe, and have water run through it to heat up? Or would it melt the copper/steel pipe? Could one make the coiled area longer to achieve this goal? It seems like the object to get heated will only heat up right in the center of the coil, is this correct? Thnx

JohnC430 (author)zar0112017-10-26

are u trying to make hot water?

zar011 (author)JohnC4302017-10-30

Yes, that is my goal to find something to heat up water, for use in a camping/trailer situation.

JohnC430 (author)zar0112017-10-20

make the coil with copper tubing and run water thru the tubing. check them out on eBay

zar011 (author)JohnC4302017-10-30

Thnx. I will check it out.

ericj77 (author)zar0112017-10-17

If you haven't seen them before, search for "induction cooktop". It's the same principle but it's used to heat metal in your pots and pans internally, rather than transferring heat via flame, etc. I guess what I mean is, it's commonly used to cook so I'm sure you could control the heat to prevent melting.

JohnC430 (author)2017-10-29

Did the 12V/50 Amp supply have any current control? i.e. current limit adjustment? the batteries do not have any current control and could possibly source even more than the max 50 Amps that the power supply can source. so really, what caused the FETs to blow? over voltage or over current? you have a scope there and can check the drain voltages of the FETs, how high do they go?

lochraceness (author)2017-10-18

Great and Very Impressive work!!! I was never able to do achieve such feats at that age (probably still can't).

I am in need of an induction heater for working on old rusty cars. I know they can be bought, but they are very expensive. Will this work, or will the power be too weak? Is it possible to easily make different 'mobile' coils if one can scale up the power?

I have been thinking of doing something similar my self, but I simply do not have the knowledge. My thought was to use an induction heater unit from a stove. Anybody think this will work? What would one do for the coils?

Hope someone has comments.

Congratiolations on the great work to tanner_tech!

JohnC430 (author)lochraceness2017-10-20

you can buy one on ebay for about $35 and then there are some that are even cheaper but smaller. if ure thinking of heating big parts u need the more expensive one.

gravityisweak (author)JohnC4302017-10-24

I saw these but then realized they came with no power supply. What should I look for? I'm asking since you seem to know your stuff.

JohnC430 (author)gravityisweak2017-10-25

what are you planning to do with them? there are small ones "rated" at 200 Watts and then there is one that they tout at between 1000 to 1200 Watts. all working within the range of 12 to 36VDC input power and all requiring an external DC power supply. In order to run the highest power one you need a power supply rated at 24V and 50 Amps = 1200 Watts. however if you are only going to load it with about 200 Watts then you can do with a smaller power supply. 200W / 12V = 16.7 Amps (approx 20 Amps) if you have a 20 Amp battery charger it will work for this. but of course it is an additional cost over the cost of the ZVS induction heater that you can buy from eBay.

gravityisweak (author)JohnC4302017-10-27

I was looking for something powerful enough to try forging some small knives. Railroad spikes, or similar size.

JohnC430 (author)gravityisweak2017-10-29

you need the 1KW unit.

i dont think 200 Watts will be enough

lochraceness (author)JohnC4302017-10-21

Thank you very much for that piece of information! This is highly appreciated!

richard62 (author)2017-10-24

This project brought back memories of the big induction heaters that we used to heat and form the cast iron nose of 500 & 750 lb. bombs that were used in Vietnam. That was in my old college days working the nightshift in the bomb plant. They used copper tubing for the coils with water running through them for cooling. Same principle, just scaled up several hundred times.

tpersson3 (author)2017-10-18

Hey, this is such a great project, and amazing comments to it too. I have a quesiton, as Im a bit of a newbie to this, in terms of Mosfets, what are the crucial elements of them, for instance, I have a bunch of IRF520's lying around, what would the implication be if I were to use those instead? Thanks!

JohnC430 (author)tpersson32017-10-20

put them in parallel. all the gates together, the drains together and all the sources together; i.e. connect all the same pins together but remember you need to also cool them so first mount them on a heat sink and then start wiring them together. try to keep the interconnecting wires short.

tpersson3 (author)JohnC4302017-10-23

Great, many thanks for the reply! Ill give it a shot!

JohnC430 (author)tpersson32017-10-21

also one more thing i forgot to mention. IRF520 has a max Vds of 100 volts and the IRF250 are 200 volts. we dont know why his transistors blew up, whether due to overvoltage or overcurrent. so...

samayaraj (author)2017-10-21

very nice :)

TechShopJim (author)2017-10-17

Great project!

Do you think multiple work coils would heat up the target metal if the work coils were connected simultaneously to the single driver circuit output?

I have an application in which I need to heat up 4 spots of 16 gauge sheet metal just a little bit (heated to raise the temperature by 10 or 15 degrees F) and then stop, so the work coils would only be on for a few seconds. Alternatively, I would connect and disconnect each work coil to the driver circuit. I'm trying to avoid having to have 4 driver circuits for this project.

JohnC430 (author)TechShopJim2017-10-21

you can make several coils and connect them in series and place one on each spot you want to heat. coils will work without you having to insert something into them. just wind them flat like a pancake and connect the start of the first one and then end of the last one to your driver circuit

TechShopJim (author)JohnC4302017-10-21

OK, series is good. That sounds like it would work. I was thinking parallel, and hadn't really considered series.

The reason I don't want to use the eBay modules for this particular project is that the scale I need is actually much smaller, and I can scale this Instructables circuit much easier.

Do you think making the flat "pancake" load coils with copper traces on a PCB would work? Remember, the scale of my project is very small, and only designed to heat up the target sheet steel by a few degrees for a very short amount of time.

If so, which of the coil designs in the attached image would work the best for a copper PCB-based load coil?

Coil #1 coils around on itself, so I'm wondering if the direction of the current in adjacent traces would cancel each other's eddy currents out. The advantage of Coil #1 is that the ends can be connected into the rest of the circuit on a single-sided PCB and no jumper wires.

Coil #2 is just a regular coil, and I'm pretty sure it would work the same as a wire pancake coil, but this coil would need to be made on 2-sided PCB in order to be able to connect the inside end of the coil to the rest of the circuit, or a jumper wire would be needed.

What do you think?


JohnC430 (author)TechShopJim2017-10-21

for such a simple application you can even use a Variac and reduce the output voltage to a couple of volts and use the 60Hz ac. or a variable DC supply and make series connections across the areas you want to heat.

welders use 60 HZ AC and what you want is only a few degrees increase in Temp.

Ok i see the picture now. #2 looks like it should work, but the traces are too thin and will burn up. you want a few amps flowing.

JohnC430 (author)TechShopJim2017-10-20

you can buy small heaters for a few $ each on eBay.

TechShopJim (author)JohnC4302017-10-20

Yes, I actually did buy some of those and they work well. However, I like the simplicity of this circuit.

I haven't tried running multiple load coils on the eBay drivers, but I will do that.


JohnC430 (author)TechShopJim2017-10-21

"simplicity of this circuit"? its exactly the same circuit of the heaters that are being sold.

TechShopJim (author)JohnC4302017-10-21

Yes, I know, but I want to scale my circuit down even more than that.

JohnC430 (author)2017-10-21

did you show a picture somewhere? "which of the coil designs in the attached image would work the best for a copper PCB-based load coil?Coil #1 coils around on itself, so I'm wondering if the direction of the current in adjacent traces would cancel each other's eddy currents out."
i did not see any attached image. if you want you can email something to me at it may be easier to communicate.
i needed to heat a crystal for a very very stable radio frequency and used induction heating. a small coil about 5/8" diameter. one turn and ran it at 1MHz. so a lot depends on your ability to do a design.
you dont need the resonant capacitors that everybody is building with; the caps are there just for this particular circuit to function. however you definitely need to have a fast changing field; i.e. oscillator of some kind. so just build one. the problem is that you need to know how much energy is required to heat the metal and then you work from there. if it is ferrous metal and is bare, you can just use DC with two contacts across the area you want to heat up.

manicmonday (author)2017-10-17

Are you sure it won't heat copper. I'm sure I have seen induction heaters heat copper. It would be great to use this to heat copper pipe to solder them together when a flame would be too dangerous.

JohnC430 (author)manicmonday2017-10-20

Compared to Iron, copper has very low resistance. I^2xR = P. when R is small you need more current to heat it. So yeah, your power supply must be capable of supplying a lot of current just to get it hot.

manicmonday (author)JohnC4302017-10-21

There are many such videos.

bpark1000 (author)2017-10-17

What frequency are you running at?

JohnC430 (author)bpark10002017-10-20

he is not controlling the frequency. it is self oscillating and the frequency is based upon the combination of the inductance of the heating coil and the bunch of capacitors. F = 1/(6.28*(L*C)^0.5) L = approximately 1-2 uH and C is whatever the guy put in. the caps are not cheap and i noticed they usually have 4 to 6 caps of 0.33uF in parallel. thats the frequency you get when u make those connections.

bpark1000 (author)JohnC4302017-10-20

I know that, but circuit determines frequency range. What is that range?

BossyRangs (author)2017-10-16

Good job! Could it be powerful enough to get knife blades red hot? Then it would be a very interesting device for knife making or for making other cutting tools (like chisels) yourself.

JohnC430 (author)BossyRangs2017-10-20

yep!!!! easy

ArturZ9 (author)BossyRangs2017-10-17

I am sure, to harden a knife blade you would need much more powerful device. As this one just heats part of the nail red. But the scheme can be rather easily scaled up to pretty much any power.

NikyN2 (author)2017-10-17

I'll probably save this for future reference (to build my own induction forge once life permits), but I would like at least some insight into how would you transform this into something that can get a somewhat large piece of steel to non-magnetic point in a reasonable time (yes, I'm interesting in blacksmithing too).

i.e. how do you add moar powah? just "sturdier" components that can handle higher currents, or should the focus be voltage, or... ?

JohnC430 (author)NikyN22017-10-20

there are many videos on YouTube and u can buy small ones like this one on eBay.

Dr.Bill (author)NikyN22017-10-17

Ya might have ah look at an induction stove I use the one I have to heat small metal parts that are too big for my solder iron and its 60 watts.

jimvandamme (author)NikyN22017-10-17

Increase the voltage (and the ratings on the switching devices) so that you can run directly off the power line, eliminating the power supply except for the diodes. But then you have to carefully insulate it for safety.

XinixE (author)NikyN22017-10-17

P= U*I, so yes, higher voltage and higher current would increase the power. Also, the tighter the air gap between coil and heated object, the more efficient the heat transfer will be, so build a coilto suit your needs.

Kalle Klæp (author)2017-10-18

Great project.You are a bright kid that can accomplish a lot.

One suggestion, though. Ditch the PVC pipe, as it generates toxic fumes when Chlorine and stuff. Make a coil out of copper, anything ferrous won't work. Actually, you can make a coil out of some copper tubing. Bend it carefully around some water pipe or the like, using a bend and draw motion to get it into shape without kinking. It shouldn't be a problem and if, just anneal the copper prior to bending it and it will be like soft butter. Remember it work hardens...which means it will be rigid again when bend´. Therefore get it right the first time and you'll have an excellent coil.

The PVC doesn't heat at all.

eslipak (author)2017-10-17

Great Work!. But the image of the schematic is a bit blurred. Can please post the schematic as PDF? Thanks a lot.

jontrav44 (author)eslipak2017-10-18

You can find the schematic on Markos Science Site, just Google 'Markos Science 500W Royer Induction Heater'

eslipak (author)jontrav442017-10-19

Will do. Thanks a lot.

About This Instructable




More by tanner_tech:Stepper Motor Key-tar: a New Kind of InstrumentDIY Induction Heater12 Volt 50 Amp Bench Integrated Power Supply
Add instructable to: