I have a project of building a furnace operated by induction heating. I have previously made a small induction heater which lets me get the metal glow red when placed in, but now i want to melt it.
Previously i used this design based on royer oscillator https://markobakula.wordpress.com/power-electronics/500w-royer-induction-heater/

I want to ask what changes i should do to make it work for 2000W or greater. I'm going to buy a 2000W power supply for this but first i need to show the schematic and designing of heater to some extent to people who are going to pay for this.
Also I've been on more than 5 hours of YouTube videos on this thing. The Radio Mechanic guy has best explanations.

 

I have a Chinish device of 35 kW, the water cooling, brick-type of igbt and 4046-based frequency adapting oscillator. Capacitors are water-cooled. The price was not much over 1k$.

 

I have previously made a small induction heater which lets me get the metal glow red when placed in, but now i want to melt it.

One thing to look at is the switching frequency. Most of these are made to heat not melt, and have a too high frequency. To melt you need to drop the frequency so the power goes deeper into the work piece, it has to do with the skin effect.

 

One thing to look at is the switching frequency. Most of these are made to heat not melt, and have a too high frequency. To melt you need to drop the frequency so the power goes deeper into the work piece, it has to do with the skin effect.

How much lower frequency we are talking about here?

 

How much lower frequency we are talking about here?

If I remember correctly they said the most of the circuits show them at ~150KHz, and to heat through a rod the recommended ~70KHz. The 150KHz will eventually get it done but adds more heat into the induction coil and time to the job. Most of the heat in the coil is from convection from the work piece.

 

RE:<< To melt you need to drop the frequency so the power goes deeper into the work piece>>
Not yet very truth.
The effect of temperature is some two orders hiher instead.
Just initially all materials are warming slowly, and then when magnetic factor is fell down, it (masters say) starting to TAKE power. You ought to know the Focault deapth is heavily dependant on magnetic factor.

But about frequency, simply the role of CD4046 is to maximize the current multiplication factor into the tank resonator, let the temperature change not offset the resonant frequency too much.

Yet generally, the difference between say 20 kHz and 13 MHz system is simple - first are giving deep and uniforms warm with rather strong mechanical mixing effect (turbulence in the melted zone), whilst second is giving more like outside warming effect (like for hardening needs) but the melted zone practically is not mixing - mechanical effect is near zero.

 

60kHz...350 kHz.
0...2 kW.
+Levitation melting.
See pictures in attached PDF.