Induction Heating

[gary350]

I finally found a circuit that has all the parts values but it is not like any of the other circuits I have seen. No information about power rating either. I guess people that design these circuits must think because they know how it works other people should too. I need an induction heater that will heat the end of a 7/8" diameter solid steel rod red hot. This circuit has no information about coil diameter.

**broken link removed**

 

[cachehiker]

I used a circuit just like that to build a Jacob's Ladder with a pair of microwave oven transformers years and years ago. It's somewhere in my 1001 Practical Electronic Circuits book dated 1975. I tuned it to run at 250Hz with 480V supply so the transformers would output 30KV with their secondaries in series.

2N3055 implies something with huge heat sinks, maybe 100W each, and maybe 10A peak current in the coils before switching. The inductance of the coil will of course be different from the one used to draw the circuit. I don't know anyway to tune it without building it, starting with a low voltage, and then adjusting the capacitance to get it oscillating at a frequency high enough to keep the transistors from overheating and low enough to get maximum power.

With any luck, 24V*10A = 240W input will get you over 120W output coupled into the steel rod. Is that enough?

I'd personally wonder how best to wind it as tight as possible for minimum reluctance and still manage to prevent a red hot steel rod from shorting out against the coil.

 

[gary350]

I watched several videos on YouTube 500 watts heats up slow 1000 watts heats up faster. I don't need it to heat up in 5 seconds if it takes 1 minute that would probably be fine.

 

[Toe Cutter]

I've used induction heating systems, but never got involved with the detailed design of them. I would agree that you would need something more like 1000+ watts. Operating frequency might also play a role in how effective your induction heater would be. I don't know how fast that 2n3055 circuit might be, but 2n3055 doesn't seem to be the fastest transistor. It would probably have lots of switching losses. Also, the 3055 would be limited to 5-10 amps conduction. These limitations of BJT are why IGBTs are used in induction systems. The circuit you posted looks like a variant of the Royer converter, which is often found in CCFL inverters. Induction systems typically have a resonant LC circuit switched with IGBTs, which lets many 10's of amps circulate in the inductor. With this circuit the current in the inductor would be limited by the conduction capability of the transistors.

 

[gary350]

I don't know anything about induction heating circuits. The circuits I find online often show no parts values or only a few parts values. Coils often have no information about diameter or what it is made out of. Some circuits have no information about the power supply. So far I have not found a circuit that I can build. I can't just grab a part out of the air and hope it works. If a circuit shows no value of a capacitor can I use a .001 UF or should it be a 45,000 MFD? If no power supply voltage is listed can I connect it to 120 VAC? After I graduated from college 1975 I never worked in electronic, technology moved forward but I didn't. I don't know what a 1GBT is. I know what a 1B3GT is.

Here is a link to a 100 circuit drawing but they all have something missing, no power supply voltage, no parts values, if a coil says 5 turns can I make it 2" diameter 2" long will that work? Sometimes a coil says, 5h no information about number of turns. I guess I need to know how to make a coil that is 5 henrys.

https://www.google.com/search?q=ind...weeoIDhBQ&sqi=2&ved=0CGYQsAQ&biw=1280&bih=815

 

[Mosaic]

Hmm, won't sticking ferrous metals into the coil change the resonant frequency? Increasing L means decreasing Hz. Thus if you're not careful this thing my over run the transistors with no load.

 

[Toe Cutter]

IGBT = insulated gate bipolar transistor. They are a hybrid of bipolar transistor and field effect transistor. They are used because of the high current, high voltage ratings and fast switching.

Yea, I know what you mean about not finding the information you are looking for on the web. The web has its limits. Induction inverters are a more obscure technology for the hobbyist. Lots of simpler circuits available with good explanations. But for something that is more complex, difficult to build, hard to implement, high power, prone to failure, more expensive... the web has its limits. It could be the people building induction inverters on the web have little knowledge themselves, or are just too lazy to detail everything or assume other people know what they are talking about.

I spent some time looking for a design for a variable frequency drive. VFD shares some characteristics (in a rough sense) with induction inverters. But there really aren't may people on the web building VFDs and publishing all the details. I still want to build a VFD, so the only thing I can do is research the basic theory and start designing. If a "canned" solution for what you want is unavailable, you have to design and build it yourself. If you are really interested in researching, then visiting a good college library would help.

At some point you will have to know how to design something and have a good sense for what is reasonable. At least have enough knowledge to get in the ballpark. A 5 Henry coil is a huge amount of inductance, for example. Maybe useful as a 60hz power supply ballast.

One idea for you: Not sure if it would work, but there are cheap induction cooktops, ~1500 watts, made in China, of course. You could probably buy one of these for about $100. Hack it to your own purposes. Rewind the coil, if necessary, to a shape that can heat your iron rods. I won't say this would work, or that it would be easy, or if 1500 W is enough power, or it won't fail. Its one possibility.

 

[Toe Cutter]

Hmm, won't sticking ferrous metals into the coil change the resonant frequency? Increasing L means decreasing Hz. Thus if you're not careful this thing my over run the transistors with no load.

It tends to the opposite, putting more iron in the field acts as a shorted turn transformer, raising the resonant frequency.

 

[gary350]

IGBT = insulated gate bipolar transistor. They are a hybrid of bipolar transistor and field effect transistor. They are used because of the high current, high voltage ratings and fast switching.

Yea, I know what you mean about not finding the information you are looking for on the web. The web has its limits. Induction inverters are a more obscure technology for the hobbyist. Lots of simpler circuits available with good explanations. But for something that is more complex, difficult to build, hard to implement, high power, prone to failure, more expensive... the web has its limits. It could be the people building induction inverters on the web have little knowledge themselves, or are just too lazy to detail everything or assume other people know what they are talking about.

I spent some time looking for a design for a variable frequency drive. VFD shares some characteristics (in a rough sense) with induction inverters. But there really aren't may people on the web building VFDs and publishing all the details. I still want to build a VFD, so the only thing I can do is research the basic theory and start designing. If a "canned" solution for what you want is unavailable, you have to design and build it yourself. If you are really interested in researching, then visiting a good college library would help.

At some point you will have to know how to design something and have a good sense for what is reasonable. At least have enough knowledge to get in the ballpark. A 5 Henry coil is a huge amount of inductance, for example. Maybe useful as a 60hz power supply ballast.

One idea for you: Not sure if it would work, but there are cheap induction cooktops, ~1500 watts, made in China, of course. You could probably buy one of these for about $100. Hack it to your own purposes. Rewind the coil, if necessary, to a shape that can heat your iron rods. I won't say this would work, or that it would be easy, or if 1500 W is enough power, or it won't fail. Its one possibility.

If you can NOT buy it or can NOT afford it then you have to build it. I bought a VFD to run a stretch wrap machine at work. The VFD only cost about $125. it runs a 1/2 HP single phase motor. I would not attempt to build a VFD for that price unless I just was building it for fun and a learning experience.

Here is a circuit that would probably work for me. What is matching inductor and how many turns does it have? No power supply voltage listed either can I put a bridge rectifier on 120 VAC and power it with 169 VDC? What about that capacitor that is in parallel with the DC power supply? How many turns is a 1.5 uh coil? This is a good example of circuit drawings with MISSING information.

**broken link removed**

 

[kubeek]

Gary, I don´t think the first circuit you posted is a good idea. First, if you want 1000W from 12V, there will be 83A flowing through those transistors.

This seems like a nice introduction, have you read it? https://www.richieburnett.co.uk/indheat.html
also this https://www.scribd.com/doc/19443502/Induction-cooker-circuit-design
I just googled induction heater design...

 

[Toe Cutter]

If you can NOT buy it or can NOT afford it then you have to build it. I bought a VFD to run a stretch wrap machine at work. The VFD only cost about $125. it runs a 1/2 HP single phase motor. I would not attempt to build a VFD for that price unless I just was building it for fun and a learning experience.

Here is a circuit that would probably work for me. What is matching inductor and how many turns does it have? No power supply voltage listed either can I put a bridge rectifier on 120 VAC and power it with 169 VDC? What about that capacitor that is in parallel with the DC power supply? How many turns is a 1.5 uh coil? This is a good example of circuit drawings with MISSING information.

Right, I could buy a VFD for some cheap amount. But most of them are way over-rated for what I want to do. I want to power a 70 watt motor at 400 hz, 120v. Besides, I take an interest in building these types of things. I think a small design, to power a small motor, is within my abilities as a hobby. Maybe you have the same idea: its interesting to build and learn.

For the circuit you post. I think it is not intended to be practical. Its just a very basic outline, for illustration. All they are showing is the very basics of a full-bridge design. Its up to you on how to implement it. Maybe run it off 24VDC, maybe 208v 3-phase. They omit the gate driving and control circuitry. The L/C values they use would give a resonant frequency of 210 khz. Pretty high. Is that a good frequency to run at? Depends on the application. When I was researching VFD I found many diagrams just like this full bridge sketch. I decided that I would not be able to find a design that does exactly what I want. I would have to design it myself if I wanted to build something.

 

[johansen]

Right, I could buy a VFD for some cheap amount. But most of them are way over-rated for what I want to do. I want to power a 70 watt motor at 400 hz, 120v. Besides, I take an interest in building these types of things. I think a small design, to power a small motor, is within my abilities as a hobby. Maybe you have the same idea: its interesting to build and learn.

Grab an irs2453d and for $2 you have a full bridge driver. I have used them through multi kilowatts at 200 khz with external boot strap diodes and external complementary pnp + npn gate drivers. However you don't need any extra parts if you keep to single digit nF mosfets and below double digit Khz. (cheaper than winding gate transformers.)

With 100uf on the boot strap capacitors you should have no problem running that chip all the way down to single digit Hz if you wish, just feed a square wave into the CT pin but don't let the ct pin go below 1/6th of vcc. To drive it i just use a couple transistors and two resistors to get me a level shifted square wave going from vcc to 1/4th vcc.

If you can ensure the duty cycle stays within .5% of 50% duty cycle then you won't need a capacitor in series with the H bridge for small motors.

All the 400 hz motors i have are 3 phase, so i haven't tried to drive them with a square wave single phase inverter. 60hz induction motors don't seem to mind, but i haven't tried driving them beyond 50% nameplate ratings either.

Three phase:

I do recall there's a circuit that uses three opamps to make three phase ac, but it only works at one frequency and the three opamps make a loop.

If you were to take three 400hz sine waves from three opamps, feed them into three comparators, each receiving a 1-10khz square wave, you could get three 0-100% pwm signals which could then be sent to three half bridge drivers, pipe the output straight into a motor, or alternatively into an LC filter to kill the high frequency.
I suppose the challenge is that you really want the pwm to stay between 20% and 80% to keep everything linear; and all three sinewaves need to have exactly the same voltage to kill any dc bias from working its way in thtere.

Aside from the 6 switches needed, the parts count for a fixed frequency driver such as I just described could probably come in under $20.

 

[Toe Cutter]

Grab an irs2453d and for $2 you have a full bridge driver. I have used them through multi kilowatts at 200 khz with external boot strap diodes and external complementary pnp + npn gate drivers. However you don't need any extra parts if you keep to single digit nF mosfets and below double digit Khz. (cheaper than winding gate transformers.)

With 100uf on the boot strap capacitors you should have no problem running that chip all the way down to single digit Hz if you wish, just feed a square wave into the CT pin but don't let the ct pin go below 1/6th of vcc. To drive it i just use a couple transistors and two resistors to get me a level shifted square wave going from vcc to 1/4th vcc.

If you can ensure the duty cycle stays within .5% of 50% duty cycle then you won't need a capacitor in series with the H bridge for small motors.

All the 400 hz motors i have are 3 phase, so i haven't tried to drive them with a square wave single phase inverter. 60hz induction motors don't seem to mind, but i haven't tried driving them beyond 50% nameplate ratings either.

Three phase:

I do recall there's a circuit that uses three opamps to make three phase ac, but it only works at one frequency and the three opamps make a loop.

If you were to take three 400hz sine waves from three opamps, feed them into three comparators, each receiving a 1-10khz square wave, you could get three 0-100% pwm signals which could then be sent to three half bridge drivers, pipe the output straight into a motor, or alternatively into an LC filter to kill the high frequency.
I suppose the challenge is that you really want the pwm to stay between 20% and 80% to keep everything linear; and all three sinewaves need to have exactly the same voltage to kill any dc bias from working its way in thtere.

Aside from the 6 switches needed, the parts count for a fixed frequency driver such as I just described could probably come in under $20.

I was looking at the high-side gate drivers. Looks like the one you mention combines all four in one package. But, the frequency generator is fixed in frequency and duty cycle. Looks like it could be useful for the OP who wants to build an induction system and is considering a full bridge circuit.

This 400 hz motor I have is single phase, but needs a run capacitor. Probably designed to run off aircraft power. One experiment I tried was to use a "modified sinewave" DC-AC inverter. The type that plugs into a 12v cigarette lighter and provides 120v 60hz. I found that my inverter uses a 556 timer to generate 60 hz timing pulses. It was pretty easy to modify it to produce 400 hz. Just change the timing capacitors. This worked, but I don't know how well a motor of this type can tolerate the rectangular pulse waveform. I was thinking of building a small VFD mostly as a learning experience.

 

[johansen]

But, the frequency generator is fixed in frequency and duty cycle. Looks like it could be useful for the OP who wants to build an induction system and is considering a full bridge circuit.

This 400 hz motor I have is single phase, but needs a run capacitor. Probably designed to run off aircraft power. One experiment I tried was to use a "modified sinewave" DC-AC inverter. The type that plugs into a 12v cigarette lighter and provides 120v 60hz. I found that my inverter uses a 556 timer to generate 60 hz timing pulses. It was pretty easy to modify it to produce 400 hz. Just change the timing capacitors. This worked, but I don't know how well a motor of this type can tolerate the rectangular pulse waveform. I was thinking of building a small VFD mostly as a learning experience.

the irs2453 is whatever frequency or duty cycle you want, if you feed it a level shifted square wave into the ct pin. you can get it to self oscillate as low as 400 hz but that probably wouldn't be reliable.


if the 400 hz motor you have needs a run capacitor then i would suspect its not going to like a square wave very much. one thing you might try is experimenting with the capacitor, change its value.