ZVS Induction heater 2000W 40A 50V

[strawberry]

ZVS Induction heater 15..50V 40A 2kW
Stanby loss with 6 capacitors: 6A@50V
MOSFETS: IRFP4227
Capacitors: WIMA FKP4 0.22uF 1000Vdc
Tank coil: D35mm H35mm 5w
PCB fabricator: PCBway
Power supply:  DPS-2500AB 2000W P-CLASS PROLIANT BLADE POWER SUPPLY

https://youtu.be/1Z1EJBMwy5Q
Altium source files : https://drive.google.com/open?id=0BwVLiaRx-ca4c2pJVlhLRmEyNDQ
Coil : https://www.ebay.com/itm/303569445933

[bktemp]

Nice design. What does your work coil look like?

One neat trick I have seen is using a small mosfet instead of the diodes (D2 and D3 in your design):
Connect the gate to 15V, source to the gate of the main mosfet and drain to drain of the opposite mosfet. So the gate voltage follows the drain voltage as long it is below ~13V.
Using this method all the gate drive power is derived from the tank circuit. So you don't need any pullup resistor or the addional gate drive stage.
The only drawback is, the 15V auxiliary supply must be present before the main voltage is applied, otherwise it may no start up correctly.

[strawberry]

Cu pipe

[T3sl4co1l]

Cute, but how is it different from the cheap junk one made in China?  Is this one safe from exploding when the coil is open or short circuited?  Does it have power control?  Can it be scaled up to practical sizes (10kW+)?

Tim

[RadioNerd]

It looks like you are not using any water cooling for the work coil.
Can you use it for several minutes without overheating?

[strawberry]

It have new MOSFETs , ultra low loss capacitors, 70um PCB technology, unified heatsink(safty), no need for gate resistors(simple turn on/off), integrated bimetal overheat protection, It all in small 50x160mm pcb form factor. Efficiency increased up to 6A stb. with same konfiguration as chinese 1kW version on ebay.

Work coil and capacitors are capable at more power than 2kW, When used all 10xcapacitors heating gone realy fast but coil reached ~400°C in 10s. With 6xcapacitors still it is not useable for several min. I cool it down in water bucket.

No it is not protected from short neither power control. It is first such powerfull device made. Because I made it not china.

[doktor pyta]

Few Years ago I built similar device for wireless energy transfer. I found that when I tried to turn on the power supply one of five attempts ended with high current consumption an activating overcurrent protection (10A) of the PS. This is possible when the symmetry in the circuit is "too good".  From further investigation it was because the two "drain" inductors were not coupled (as a middle-tapped autotransformer). "working coil" has no central tap. I think that making it similar to Resonant Royer topology would make it work more reliable.
Maybe someone has better explanation.

[strawberry]

Interesting drain inductors configuration. I never had startup failure before(at no load).
When turning device on with gate power then drains get voltage overshoot.

[T3sl4co1l]

Most common failure mode is lack of oscillation leading to massive power dissipation, or gate bias dropping out before drain current, resulting in massive overvoltage spike.

This is a current-sourcing circuit, and as with all current sources, it must be commutated by bringing the source to a short-circuit condition.  (The source supplying the transistors is the inductor(s), and whatever current they've been charged up to.  Discharging these, with a short circuit, while there's a constant-voltage source behind them, is a non-obvious procedure.  This highlights another of the fundamental quirks/design issues of this circuit.)

Tim

[strawberry]

PCB update

[BlackTM]

Hi strawberry.

Can you please tell me at what frequency does your induction heater work?

[james_s]

The induction heater I built from a circuit I found online uses a small 15V standby power supply to get it oscillating and then there is a relay that applies full power when activated. Hasn't blown up yet.

[strawberry]

Hi strawberry.

Can you please tell me at what frequency does your induction heater work?

It's roughly 90..100kHz( depends on work coil and capacitors)

The induction heater I built from a circuit I found online uses a small 15V standby power supply to get it oscillating and then there is a relay that applies full power when activated. Hasn't blown up yet.

My schematic are the same, but additional gate drive emitter follower gives nice square wave for MOSFET gates.

[calexanian]

Looks good. I have a couple of those cheap ones from china. I also have a big old lepel RF generator. I have always wanted to modify one of these circuits with a starter circuit that would always ensure a proper start. I think maybe a quick one shot pulse on one side to get it going, or introduce a low level recurring pulse on one side from a separate driver circuit that then just gets ignored once the power devices get going. I have mine connected to a foot switch and an LED riged up when it is running. If the light does not turn on I let up on the pedal momentarily and when I step back on it again it will usually start. I also use a variac to control the power level. Its a nice arrangement. Very accurate. Makes pretty welds in copper pipe if solder paste is used. Fantastic for copper tube wave guides. Gotta use water cooling for that though cause copper is hard to heat.

[strawberry]

Lepel RF generator is something new for me, is it using vacuum tubes? Do you have schematic?

I am planing to build mains powered PLL induction heater. I am aiming on about 3kW power

[calexanian]

Lepel is still in business. They are owned by Inductotherm now. The unit I have is tube based but they make fully modern types now. Honestly though i just use the ZVS ones now. Less trouble. Ebay is full of old Lepel tube models, but I would advise against them. Once the insulation on things starts breaking down inside they just dont work right anymore and they are truly a pain to rebuild. Way cheaper and easier to build one like yours. There is a classic circuit in the RCA transmitting tube manual for a 833 type tube based one that was popular back in the day. Its very tricky to get the main coil and tank circuit right, but they evidently worked quite well.

[kelesekrem]

hi, thanks.
I want to use this heater model in 3d printer, can I use it and can go to high temperatures for 600C

[strawberry]

Do you mean casting metal  in 3D printed molds .. Yes it could be possible, by now I have not tried to melt something with induction heater.

[3DARTjar]

Thank you Strawberry for sharing with us your project! I would be interested what quantity of nonferrous metal can you melt with this heater in a alumina crucible ? I would like to use it with a AEG SAFT NIFE 50V/34A power supply if it is suitable ?

[jlkazim]

Hello I am interested in ZVS Induction heater 15..50V 40A 2kW but now I want to order 20 PCB of this machine and the components how to do?

[CopperCone]

I wanted to build one with h bridge phase power control. This way you could use a dspic to control both phase and frequency.

This seems less parts heavy then designing a high power adjustable supply for the circuit in this thread. The only benefit i see is that it operates at the resonant frequency without tuning, but it seems to have a potential startup problem.

Is there any benefit to using dual stages vs full h bridge control with an adjustable induction heater.

However as your melt gets hot the inductance changes so the resonance frequency changes and you get power loss... So this method may be better then phase control for that reason.

I guess that you would need to add a pll tracking circuit to the mcu and have the mcu periodically sample the pll output to adjust the center frequency..

[T3sl4co1l]

Back in industry, I used a single stage, full wave, frequency offset control.  Simplest power stage, but crappy to compensate.  Wanted to try phase control, but the driver and output stage didn't have the logic to support it.  Yes, you need a PLL to track resonance, or something near it in any case.  Phase shift PWM gives you effectively a variable voltage source, exactly what you need.  With the system at resonance, the amplitude response is a (single or repeated?) pole at the tank time constant, low Q and easy to compensate.

Downside is 100% switching loss and EMI.  A tri-level inverter (a half bridge with a shorting-to-zero mode) might be preferable for this reason.

Tim

[j2000]

Hey Strawberry...thanks for the schematic...
Did u sell DIY kit complete?

Regards
Joel

[IvanLazarov]

Hi Strawberry, thanks for sharing the schematics for your induction heater. I will try to build it myself.

Can you please advice what should be the current rating of chokes L1 and L2? If the total power is 2KW, does that mean that each choke should be rated at minimum of 20A?

What is the purpose of the additional circuit on the right hand side that you included in the latest revision the drawing?

Cheers,
Ivan

[strawberry]

1/If you want do serious staff with it, then L1 L2 will be redesigned.
2/It is protection circuit idea