Induction Heating Circuit

[Windows9Professional]

I'm quite into induction heaters since a few weeks so I of course wanted to build a simple induction heater. But I didn't want to cheaply pick a schematic from the internet, so I tried to design my own. I'm here to ask if my schematic (Picture1) would actually do its job or if I can basically throw it in the garbage.

It's working behaviour is simply putting the work coil straight across supply voltage till the output voltage across the voltage divider is higher (caused by saturation of the inductor) than the forward voltage of the diode which activates the transistor and pulls the gate of T1 to ground and the gate of T2 to supply voltage. This discharges the capacitor and the cycle repeats.

I didn't bother calculating the component values because I don't even know if it works.

btw: I start to think that I don't even need T2 at all because the capacitor will discharge through the work coil if T1 is off.


What do you think about this schematic?

[Weston]

I don't think that circuit will work. For steady state operation the average voltage across an inductor is zero, but your inductor is connected to ground on one side and to a switch that will have a body diode on the other side so it can never swing negative. If the top side switch did turn on / this somehow oscillates the current would probably just ramp until something blows up.

Beyond the more nuanced point of the average voltage across an inductor, your feedback divider is connected to ground on both sides so it will not do anything.

Designing oscillators from scratch can be a bit tricky you might want to start with some existing design first, like the common ZVS design https://spaco.org/Blacksmithing/ZVSInductionHeater/1000WattZVSInductionHeaterNotes.htm

For validating your own designs you should probably start with simulating them, using something like LTSpice. LTSpice is a pretty powerful tool and would be useful to learn how to use for projects in the future beyond this https://www.analog.com/en/design-center/design-tools-and-calculators/ltspice-simulator.html

[Windows9Professional]

Thank you for your reply,

After looking at my schematic for a longer time I can only agree with you that my voltage divider is just shorted out.

One more question (just for fun): Would an H-bridge inverter with just a modified square wave do anything in terms of induction heating (with the right frequency)?

[Weston]

H bridge inverters are often for induction heating. You need some sort of matching network though and the control circuity is more complicated (with more potential to make the power electronics explode) over something that is self oscillating.

This page has a pretty good example of a H bridge based induction heater: http://www.richieburnett.co.uk/indheat.html

The ZVS based induction heater kits are pretty cheap on ebay, you should consider getting one to experiment with first and then try making your own driver. You could probably re-use most of the components too.

[Berni]

Yep if you want a self osculating circuit for induction heating it is hard to beat ZVS:
https://i.pinimg.com/736x/0b/ae/af/0baeaf6209c14eb40db5e88d8ad5649c.jpg

Not only is it very simple but it is also very efficient because the transistors are switching at zero crossing. Only thing to watch out for is that it needs a fast sharp turn on of the power supply to get into a oscillation (Slow ramp up of the supply voltage makes both transistors turn on and explode), but there are ways of improving the circuit to guard against that.

If you want to design your own then get yourself LTSpice or similar circuit simulation software as Weston has already suggested. The simulator can show you how the circuit will work and let you play around with it.

[Windows9Professional]

I've downloaded LTspice yesterday and I've had quite a lot of educational fun. I've rebuildt different kinds of induction heaters and am pretty fascinated by how the software gets basically every physical aspect right.

Thank you all for your replys

[Berni]

I've downloaded LTspice yesterday and I've had quite a lot of educational fun. I've rebuildt different kinds of induction heaters and am pretty fascinated by how the software gets basically every physical aspect right.

Thank you all for your replys

Yep this is how most people design analog circuits. Its so much easier to play around in SPICE to figure out how it all works before actually building it for real.

But be careful about how you interpret the results from LTSpice. Just because it works in the simulator does not mean it will work in real life. In simulation components can never blow up no matter how overloaded they are, wires have zero resistance and zero parasitic inductance/capacitance, components have perfect values with no tolerance... etc. So if you do get odd behavior in simulation its likely due to one of those aspects of the simulation being too perfect compared to real life.