A DIY Induction Heater
This great little project demonstrates the principles of high frequency magnetic induction. The circuit is very simple to build and only uses a few common components. With the induction coil shown here the circuit draws about 5A from a 15V supply when a screwdriver tip is heated. It takes approximately 30 second for the tip of the screwdriver to become red hot!
The control circuit uses a method known as ZVS (zero voltage switching) to activate the transistors which allows for an efficient transfer of power. In the circuit you see here, the transistors barely get warm due to the ZVS method. Another great thing about this device is that it is a self resonant system and will automatically run at the resonant frequency of the attached coil and capacitor.
How Does Induction Heating Work?
When a magnetic field changes near a metal or other conductive object, a flow of current (known as an eddy current) will be induced in the material and will generate heat. The heat generated is proportional to the current squared multiplied by the resistance of the material. The effects of induction are used in transformers for converting voltages in all sorts of appliances. Most transformers have a metallic core and will therefore have eddy currents induced into them when in use. Transformer designers use different techniques to prevent this as the heating is just wasted energy. In this project we will directly make use of this heating effect and try to maximise the heating effect produced by the eddy currents.
If we apply a continuously changing current to a coil of wire, we will have a continuously changing magnetic field within it. At higher frequencies the induction effect is quite strong and will tend to concentrate on the surface of the material being heated due to the skin effect. Typical induction heaters use frequencies from 10kHz to 1MHz.
The Circuit
The circuit used is a type of Royer oscillator which has the advantages of simplicity and self resonant operation. A very similar circuit is used in common inverter circuits used for powering fluorescent lighting such as LCD backlights. They drive a center tapped transformer which steps up the voltage to around 800V for powering the lights. In this DIY induction heater circuit the transformer consists of the work coil and the object to be heated.
The main disadvantage of this circuit is that a center tapped coil is needed which can be a little more tricky to wind than a common solenoid. The center tapped coil is needed so that we can create an AC field from a single DC supply and just two N-type transistors. The center of the coil is connected to the positive supply and then each end of the coil is alternately connected to ground by the transistors so that the current will flow back and forth in both directions.
The amount of current drawn from the supply will vary with the temperature and size of the object being heated.
From this schematic of the induction heater you can see how simple it really is. Just a few basic components are all that is needed for creating a working induction heater device.
R1 and R2 are standard 240 ohm, 0.6W resistors. The value of these resistors will determine how quickly the MOSFETs can turn on, and should be a reasonably low value. They should not be too small though, as the resistor will be pulled to ground via the diode when the opposite transistor switches on.
The diodes D1 and D2 are used to discharge the MOSFET gates. They should be diodes with a low forward voltage drop so that the gate will be well discharged and the MOSFET fully off when the other is on. Schottky diodes such as the 1N5819 are recommended as they have low voltage drop and high speed. The voltage rating of the diodes must be sufficient to withstand the the voltage rise in the resonant circuit. In this project the voltage rose to as much as 70V.
The transistors T1 and T2 are 100V 35A MOSFETs (STP30NF10). They were mounted on heatsinks for this project, but they barely got warm when running at the power levels shown here. These MOSFETs were chosen due to having a low drain-sorce resistance and fast response times.
The inductor L2 is used as a choke for keeping the high frequency oscillations out of the power supply. The circuit will work without it, but it is less efficient, and could lead to damage of the power supply or control circuit. The value of inductance should be quite large, but also must be made with thick enough wire for carrying all the supply current. The one shown here was made by winding about 8 turns of 2mm thick magnet wire on a toroidal ferrite core. As an alternative you can simply wind wire onto a large bolt but you will need more turns of wire to get the same inductance as from a toroidal ferrite core. You can see an example of this in the photo on the left. In the bottom left corner you can see a bolt wrapped with many turns of equipment wire. This setup on the breadboard was used at low power for testing. For more power it was necessary to use thicker wiring and to solder everything together.
As there were so few components involved, we soldered all the connections directly and did not use a PCB. This was also useful for making the connections for the high current parts as thick wire could be directly soldered to the transistor terminals. In hindsight it might have been better to connect the induction coil by screwing it directly to the heatsinks on the MOSFETs. This is because the metal body of the transistors is also the collector terminal, and the heatsinks could help keep the coil cooler.
The capacitor C1 and inductor L1 form the resonant tank circuit of the induction heater. These must be able to withstand large currents and temperatures. We used some 330nF polypropylene capacitors. More detail on these components is shown below.
The Induction Coil and Capacitor
The coil must be made of thick wire or pipe as there will be large currents flowing in it. Copper pipe works well as the high frequency currents will mostly flow on the outer parts anyway. You can also pump cold water through the pipe to keep it cool.
A capacitor must be connected parallel to the work coil to create a resonant tank circuit. The combination of inductance and capacitance will have a specific resonant frequency at which the control circuit will automatically operate. The coil-capacitor combination used here resonated at around 200kHz.
It is important to use good quality capacitors that can withstand large currents and the heat dissipated within them otherwise they would soon fail and destroy your drive circuit. They must also be placed reasonably close to the work coil and using thick wire or pipe. Most of the current will be flowing between the coil and capacitor so this wire must be thickest. The wires linking to the circuit and power supply can be slightly thinner if desired.
This coil here was made from 2mm diameter brass pipe. It was simple to wind and easy to solder to, but it would soon start to deform due to excess heating. The turns would then touch, shorting out and making it less effective. Since the control circuit stayed relatively cool during use, it seems that this could be made to work at higher power levels but it would be necessary to use thicker pipe or to water cool it.
Pushing it Further
The main limitation of the setup above was that the work coil would get very hot after a short time due to the large currents. In order to have larger currents for a longer time, we made another coil using thicker brass tubing so that water could be pumped through when it was running. The thicker pipe was harder to bend, especially at the center tapping point. It was necessary to fill the pipe with fine sand before bending it as this prevents it from pinching at the sharp bends. It was then cleared out using compressed air.
The induction coil was made in two halves as shown here. They were then soldered together and a small piece of pvc pipe was used to connect the central pipes so that water could flow through the whole coil.
Less turns were used in this coil so that it would have a lower impedance and therefore sustain higher currents. The capacitance was also increased so that the resonant frequency would be lower. A total of six 330nF capacitors were used to give a total capacitance of 1.98uF.
The cables connecting to the coil were just soldered onto the pipe near the ends, just leaving room for fitting some PVC pipe.
It is possible to cool this coil simply by feeding water through directly from the tap but it is better to use a pump and radiator to remove the heat. For this, an old fish tank pump was placed in a box of water and a pipe fitted the outlet nozzle. This pipe fed to a modified computer CPU cooler which used three heat-pipes to move the heat.
The cooler was converted into a radiator by cutting the ends off the heat pipes and then linking them with PCV pipes to the the water would flow through all 3 heatpipes before exiting and going back to the pump.
If you do cut some heatpipes yourself, make sure to do it in a well ventilated area, and not indoors as they contain volatile solvents that can be toxic to breathe. You should also wear protective gloves to prevent skin contact.
This modified CPU cooler was very effective as a radiator and allowed the water to remain quite cool.
Other modifications needed were to replace the the diodes D1 and D2 with ones rated for higher voltages. We used the common 1N4007 diodes. This was because with the increased current there was a larger voltage rise in the resonant circuit. You can see in the image here that the peak voltage was 90V (yellow scope trace) which is also very close to the 100V rating of the transistors.
The PSU used was set to 30V so it was also neccesary to feed the voltage to the transistor gates via a 12V voltage regulator. When no metal was inside the work coil, it would draw about 7A from the supply. When the bolt in the photo was added, this went up to 10A and then gradually dropped again as it heated up beyond curie temperature. It would certainly go over 10A with larger objects, but the PSU used has a 10A limit.
The bolt you can see glowing red hot in the photo took about 30 seconds to reach maximum temperature. The screwdriver in the first image could now be heated red hot in about 5 seconds.
In order to go to higher power than this, it would be necessary to use different capacitors or a larger array of them so that the current was more distributed between them. This is because the large currents flowing and high frequencies used would heat the capacitors significantly. After about 5 minutes of use at this power level the DIY induction heater needed to be switched off so that they could cool down. It would also be necessary to use a different pair of transistors so that they could withstand the larger voltage rises.
In all this project was quite satisfying as it produced a good result from just a simple and inexpensive circuit. As it is, it could be useful for hardening steel, or for soldering small parts. If you decide to make your own induction heater project, please post your photos below.
The information provided here can not be guaranteed as accurate or correct. Always check with an alternate source before following any suggestions made here.
What is the red waveform on the scope?
The red waveform is the gate voltage of the other transistor.
Yes you can.
Rusdi,
There is also the temperature hazard, and then potentially chemical hazards if components get burned. The scope is not needed, but was useful for seing the waveforms and frequency. The scope ground should be the same ground as your PSU. you can then measure at the MOSFET gate, or collector terminals. If you measure at the collector, remember that the voltage could be much higher than the supply voltage so set the scope accordingly.
Surastyo.
I can't seem to find the datasheet for that part number. What are its ratings? They are typically set as GDS, but you need to check the manufacturers datasheet. I also do not see any large inductor (L2) in your picture. I've attached a diagram showing the 7812 voltage regulator being used.
I think 60V is much too low of a voltage rating for the MOSFETs. THey are probably being destroyed by over voltage.
To get hotter temperatures, you need to use more powerful transistors and capacitors, and then use a higher supply voltage.
We don't have anything else available at the moment. You would just have to look for components with larger voltage and current ratings.
Martin, If you are not getting nice sinusoidal oscillations, this could be due to having too little, or even too much inductance in your choke (L2). You may need to experiment with different choke designs to find something that works well with your other parts. Your work coil should ideally be only a little larger than the object to be heated as this will maximise the concentration of magnetic flux. If that does not help, add a photo of your scope showing the gate voltage, and drain voltage waveforms.
Joco, I can't find that part number. Did you make a typo?
Tohooloo, C1 can be as large as you like. The combination of capacitance of C1 and the inductance of L1 form a resonant circuit. A larger capacitance or inductance will result in a lower frequency.
Joco, That MOSFET seems ok. Try using a larger choke (L2).
Neither. You need higher power.
Deyan,
It sounds like your capacitors have too much internal resistance (ESR). You need something that is better for high power, they are typically much larger in size for the same capacitance value. There are many types of transistors that may work, but it is up to you to test if the ones you have are ok. I suspect that a high voltage IGBT will have too large of a voltage drop between collector and emmiter which would mean that the opposite transistor would not switch off properly. I used these 100V, 35A MOSFETs which worked well.
Stephen,
I used these 1000V, 330nF capacitors.
I can only suggest you follow the instructions given. If you use other components, they may or may not work. That is up to you to work out.
1. All those factors will vary the performance. The number of turns will determine the impedance of the coil and therefore how much current will flow for a given frequency. Less turns will allow for higher peak current, but too few and the transistors will blow.
The coil diameter should ideally be just slightly larger than the workpiece that is to be heated. This maximises the field coupling and is more efficient.
The capacitor size will also determine peak current. larger is better, but not too large or the resonant frequency will become too low. Physically larger capacitors will have low internal resistance too so they will perform much better. The bank of caps in your picture look way to small for the amount of metal you are trying to heat.
2. You should get a nice half sine waveform. You need to experiment witht the inductor (L2). Try different numbers of turns until you see some improvement. You could also add 15V Zener diodes between the gate and source pins of each transistor to help protect them from voltage spikes.
3. V+ can be as high as you like, just make sure your components are sufficioent to handle the voltage rise in the coil. You will also need to supply the gates from a stable 12V source.
4,5,6. See other answers.
You can use a single power supply, or seperate ones. If using a single supply it should be at most 15V, or 30V if you use the regulator as shown in another post.
Your power supply is not able to deliver enough current. You need a bigger supply, or many more turns on the work coil or choke.
Yes
The current will depend on the voltage you apply, and the load impedance. More turns on your work coil, smaller capacitors, and larger choke inductance will reduce current.
Nice job. Thanks for sharing Rob.
James,
If you Google the part numbers you will find the manufacturer pdf datasheets. These will tell you which pins provide a particular function. I think it is best you try working this out yourself as you will gain a better knowledge and understanding. You can also check out the section on the site for helping to learn electronics as it will explain a lot about the components and the physics of the electronics involved.
Chanil,
Yes, if you want to limit the current to 1A, then you must increase the impedance of the circuit. Using many more turns in your work coil and the choke will limit the current.
Chanil's video...
ad,
You can reduce C, but you will need t make sure that the remaining capacitors you have are able to take all the current. The advantage of using lots of capacitors is that the current is shared between them.
Eric,
What you describes just sounds like it is due to the limitations of your PSU. Remember that voltage and current are directly related and will be affected by the impedance of your load. If your PSU shows low volts, and high amps, this indicates that the impedance of your circuit is low and your PSU is dropping the volts to keep the current within its ratings.
Yes, it would seem that you just have DC flowing at low voltages