3kW Induction Heater Build

Waar kan ik her kopen voor PCB 3 Kw?

Download KiCAD, it might assist you in the PCB files.

I have same issue? Can Anyone help?

Facing the same issue, no idea what's wrong in this design.

One thing about ZVS circuits is that they need near instant operating voltage. Switch mode power supplies tend to ramp up to operating voltage.

As such you need a high current rated switch (usually solid state) to turn the circuit on when the supply is at full voltage.

Connecting directly to a power supply and just turning on the supply will kill the mosfets every time

2.2 microfarad is the most important part second is the 100 volt or higher study Ohm's law you will understand

R4 is part of the RC snubber circuit designed to help reduce ripple before the MOSFETs gate pin. If you were to reduce the resistance of R4, it may cause some issues with oscillation? or be completely fine? Also keep in mind the maximum power dissipation for zeners D1 & D2, as reducing R4 resistance will mean more power dissipation for the zeners. R5 is just a gate pulldown resistor, so unless I'm overlooking something? I can't see how that will reduce the inrush current upon startup. If we were to assume your powersupply is hitting current limit on startup, then perhaps adding one or 2 large capacitors on the output of your power supply would solve the problem?? Keep in mind you have to charge up the caps before connecting the induction heater otherwise the additinal caps will exacerbate the problem further

hi i have exactly the same issue i can find no issues with my build followed exactly and it has blown 3 of the mosfets only q2 gets warm the only difference i can see is the power supply which is a huawei R4875G1 55v75a could that be the problem

Just remember the tank circuit at the output is capacitive inductive tuned so keep in mind the size of the induction is important and proper in being a perfectly turned tank circuit for lagging and leading reaction for the output to be tuned null...

Now i got my 50khz Wave but its No sinewave. Its an Mix of sine and sawtooth.

hola buenos dias, queria saber a que temperatura llega el fierro o lo que le ingreses a la bobina y con que fuente de alimentacion lo estas alimentando, desde ya muchas gracias.

You are probably saturating the coils... check the workaround in the 3kW version.

Hello, If I had to choose between those 2, I'd go with the wima MKS4 330n 1000vDC caps. Just be aware they have a maximum AC voltage rating of 400vac. 😀

@Marcin Skwierczyński I know, that is not simple. I need max temperature 1000 celsius degrees. I need have about 10 kHz (I will use this inductor for heat treatment, not for melting metal. This is important for the frequency of operation of the inductor itself). But I need start from one point, and I decided to take the first step here.

I’m in almost same situation like you then, although my approach took me trough building small zvs driver first, then I realized I need to gather proper equipment to analyze how it works and understand the basics to be able to push it further and make proper induction heater. In your case if you need to heat steel up to 1000 C, then you will need to adjust frequency during heating process, near 770 C you’ll be at Curie point. Try out building most basic zvs induction heater, power it trough 5-12v DC and start analyzing. It’s great journey and a lot of fun, i hope I’ll finish mine in 6-12 months, depending on amount of problems I’ll encounter. Fingers crossed Ps. Part of research was choosing proper equipment to analyze it, hit me up if you’ll need advice what to choose on low to medium budget.

@Marcin Skwierczyński Of course. What is the power of your inductor? Send me a list.

Don't worry, your English is very good 😀. I would probably start by soldering all the components to the PCB. After that, I would tin the high current traces with a small amount of solder (this will help make a good electrical connection between the PCB track and bus-bar later). Next, I would fit the bus bars onto the PCB. This step would be a little difficult due to the fact all the leads would have to be perfectly aligned to allow the bus bar to be fitted. Once the bus bar has been fitted, it's just a matter of soldering the leads to the bus-bar. Although a bus-bar has a nicer appearance to using copper wire as I showed in my video. Using wire is much quicker & easier in my opinion. I've built 3 of these heaters using the copper wire method, and I've had no issues/failures yet. 😄

@Schematix Is it the "third" one the 10kW one? 😄

Hi Edward,

1. Download the gerber.rar file from the link provided

2. Goto JLCPCB.com

3. Click on "quote now" or "Order now"

4. Click on "Add gerber file"

5. Select the gerber.rar file you previously downloaded

6. Wait for the file to be processed. Once completed you'll see a preview of the PCB

You can customize the design options to your choosing. However if you're unfamiliar with certain options, leaving them at their default option is perfectly acceptable for this particular project.

Hope this helps 😁

@paul K any idea how (and if) L1 and L2 affect the frequency? Or is it all in the capacitor bank and the working coil? If it's the working coil, perhaps something involving Kapton tape, some stainless eyes + wire, and varied tension to adjust the spacing of the coil windings (and a black and tackle for more fine grained control)? Alternatively, could different inductors (in series) be switched in and out through a bank of relays, similar to how ATUs in ham radios work to tune antennas? I think the ATUs also switch different caps in and out as well for more combinations. A few select banks and L values might yield a large range of operating frequencies, particularly if more than one inductor is switched in (or simply shorted across to remove it from the circuit).

But also, having said that, the effect would not be as large as being able to increase the power in the LC tank, which does of course mean the ability to handle more current. It's also worth noting that coupling is an important factor i.e. trying to heat a small workpiece in a large coil isn't helpful. The resource link below deals with some theoretical considerations of induction heating. It's quite readable and is well worth reading. In particular it talks about the size of workpieces and the effect on heating and on page 16 gives three bullet points indicating items which influence the rate of heating:

• Field strength of the magnetic flux field

• Coupling of the induction coil to the workpiece

• The electrical and magnetic properties of the material being heated

http://www.elegron.pl/files/TheoryHeating.pdf

I'm currently researching what influences the temperature which can be achieved when heating a workpiece, and I suspect that it will be down to the amount of power which the circuit can generate (which influences the magnetic field intensity) and the efficiency with which that power can be transferred to the workpiece e.g. size of workpiece wrt the coil. So coil size matched to workpiece size is good for power transfer. The nature of the workpiece also has significant effect.

For a simpler design, using a high/low side driver vs the gate drive transformer could give you the ability to tune the circuit in a straight forward way:

Adding auto tune via PLL has been done here:

http://inductionheatertutorial.com/inductionheater/induction9.html

You'll need a power supply to go with it...

If you want to build one yourself, I would estimate roughly $150 USD in materials for the heater alone. A suitable power supply will cost at least $60 USD and require significant effort to prepare for use. I would caution you based on your question about soldering in the other thread that someone who is not comfortable enough with electronics to build the heater using the provided instructions may not be prepared to safely operate a high-power device with exposed conductors and thermal constraints. Self-contained commercial heaters for "operator use" start at around $5000 USD.

The temperature will be highly dependent on input voltage, coil geometry, workpiece size, thermal insulation, and duty cycle. I have been running the previous 1.5 kW model and have been able to melt aluminum and copper. The latter required continuous runtime capability supported by a cooling fan on the board and water cooling of the coil. Melting iron in an insulated crucible at 3 kW seems plausible, but iron becomes more difficult to heat inductively once it attains red heat which manifests as a drop in current.

hola como andas, queria saber si armaste este modelo de 3 KW y si pudiste fundir aluminio, desde ya muchas gracias.

Yes, it seems there is a supply shortage at the moment. Mouser expect to have more in a couple weeks, so hopefully they come back in stock soon. You can try look for alternatives if you wish.

I got a response from the supplier to the effect that connecting in serial will subject some components to stresses outside of their specification, and would likely cause catastrophic component failure. If that's not enough, the other problem I've seen in the research I've done is that unless you can engineer all supplies to power up simultaneously, there is a likelihood of individual supplies driving others in reverse polarity. I can solve that one by powering up all supplies with no load connected, and then having one switch device (mechanical or electronic) which switches in the load. Anyway I've decided not to pursue this route and I'll use a modded MOT (or two) with a bridge rectifier and smoothing caps. Unless I can find something better and accessible.

@paul K Thanks for all the the great advice Paul, much appreciated. When I get some time and a little more knowledge I will eventually test it but not expecting any miracles, and won't even go near it until positive I'm safe even though equipment may not be. What sucks is I have one of these cheap 2.5kw china ZVS melters and can't use it properly wihout a strong PS. Will be interesting to see which blows up first - the series wired PS, or the barebones chinese ZVS melter.

@Mike Smith All of the cases must remain connected to the AC earth for safety which implies that the cases are all grounded together regardless of whether there is direct contact between power supply cases. As the first power supply in the series should be unmodified, all of the cases will be tied to the DC ground. If two 12 V supplies were placed in series, then the DC output of the second supply would measure 24 V when connecting a meter to either case. I am unsure what you mean regarding measuring the same voltage from "any of the rails" to ground as the individual output terminals along the series will provide different voltages relative to the first supply's DC ground or any of the cases.

Yes, here in the U.K. the supply voltage is 230V and plug outlets on a ring main are capable of delivering 13 amps. Using an MOT is empirical as I don't know the number of turns on the primary. As its a 3kVA power requirement and assuming no losses (!) between primary and secondary, V*A Primary = V* A secondary. At the secondary I'd have 48V @ 60 amps which gives a primary current of 13 amps - the limit in the U.K. I'll probably have to tone it down a bit by adjusting the number of secondary windings - the thing I don't know until I try is if I can get sufficient number of turns of suitable gauge wire to physically fit into the secondary windings space. I'll give it a try and report back.

Yes, here in the U.K. the supply voltage is 230V and plug outlets on a ring main are capable of delivering 13 amps. Using an MOT is empirical as I don't know the number of turns on the primary. As its a 3kVA power requirement and assuming no losses (!) between primary and secondary, V*A Primary = V* A secondary. At the secondary I'd have 48V @ 60 amps which gives a primary current of 13 amps - the limit in the U.K. I'll probably have to tone it down a bit by adjusting the number of secondary windings - the thing I don't know until I try is if I can get sufficient number of turns of suitable gauge wire to physically fit into the secondary windings space. I'll give it a try and report back.

@A Newcomer Not what I wanted to hear regarding saturating transformer cores obviosly but good to know. How about falling back to 2 MOTs drawing less than 3kw(heavan forbid), or maybe some sort of series or series/parallel configuration?

To bad as well on the 120V primary insulation rating. Much appreciated!

Maybe......

Now one has the guts to go big - show us how it's done.

- - - How much Aluminum can you melt with that crucible?

Heat times depend on many factors. So there is no one-size-fits all answer sorry.

The great thing about this design is the L-C is self tuning (aka self resonating) since the MOSFET gates are directly coupled to the L-C circuit. The overall length & number of turns does play a major role in workcoil design. This is an area where I'm still learning. So I'm not sure my advise would be all that useful

@Schematix Thanks, I'll probably build a few coils, following the advice on the Ambrell link about coil size matching workpiece size. I'll experiment a bit and report back. The MOSFETS are currently unavailable with long lead times, so I'll look around for equivalent specs. They're a bit more meaty than the ones I use for microcontroller stepper motor projects!