Mazilli ZVS Induction Heater (With IGBT's)

[Dylan Burger]

Good day all.

Im writing this so i may share knowledge i have gained while constructing a IGBT driven mazilli induction heater.
This entire post will also apply to MOSFET driven versions of this driver.

I have been told there are issues when trying to use IGBT's with this circuit , but I am not sure if these people were referring to the "brick" type IGBT's or TO-220 style ones. I am using generic Field Stop IGBT's that are purposed for use in inverters and mid frequency range stuff. They cant reliably run my SSTC though. ~400Khz is a bit harsh at mains voltage...

There will be troubleshooting help at the bottom of post to assist with general issues that may arise as well as operating tips.

First of all, the circuit diagram : [From Kaizer Power Electronics]




I have only changed the value of a few components, however it should not be critical to the functioning of this circuit. The components specified and their values are a great starting point for most induction heaters.

I used 18v 2w Zenner diodes instead of 12v [for components D1 and D2] (Make sure that your zenner voltage does not exceed your MOSFET's or IGBT's max gate  rating. Most are safe at ~18v. Always check your datasheet !)

For R3 and R4 i used 500 ohm resistors instead of 470.

For D3 and D4 i used BYV-26-200 diodes.

And of course, the transistors i used are 40N65FL2 IGBT's.

Also, i have a slightly different additional inductor placed on the positive rail (L1) - see my schematic based on Mads' schematic. ( below ) 



And here is a view of the inductors :



PART 1  : The Build

*Ensure that all connections are very solid and that you use a heavy gauge wire / high current rated wiring for anything coming from the work coil, tank capacitor and pins 2 + 3 of your  MOSFET/ IGBT's. having thin or low current cables here will heat your MOSFET/ IGBT's and the cable itself will more than likely fail!

My tank cap is just a bunch of 0.68uF 300V MKP capacitors put in parallel. I have ~5uF tank capacitance. Even so , it still heats up pretty easily. It would be a good idea to have large MMC type capacitor to dissipate heat and handle more current.

Here are some photos taken of the project  :




The IGBT's are mounted on computer heat sinks that have been joined to each other by two pieces of acrylic sheet that was screwed onto the mounting points on the heat sink.

The work coil is 18 turns of trifilar wound 1mm magnet wire. I had to wind it trifilar due to the coil starting to burn from the current in it. I plan to upgrade it as soon as i can find some 3mm or 5mm copper tubing.

PART 2: Testing

Now comes the part where you power it on for the first time. This can be a very critical point. Why ? Most people dont want to blow anything up if they hooked it up wrong, so they start with low power and in turn this causes the Mosfets to not oscillate. You get a short and either something blows up or you get a spike in current draw. Its also a good idea to have a beefy power supply, one that can (at minimum) supply 12 Volts @ 10 amperes.

That said, the following is merely the way i go about testing a newly built Mazilli heater, and this serves as a guideline - NEVER FORGET TO APPLY YOUR OWN LOGIC. 

-- (1) Make sure you have something that can indicate how much current is being drawn, and something connected in parallel with the tank capacitor so you can measure or monitor the resonant frequency and tank voltage.

-- (2) Have a trip switch set in place, mainly so that if there is a short, it wont destroy anything if you cannot power it off in time. At which point it trips depends greatly on what your power supply can deliver. my supply can push up to  ~100A @ 50V DC, so i chose a 20A breaker. If anything goes bad , it will trip, and the worst that could happen is a busted MOSFET/IGBT.

-- (3) Go over your circuit and ensure everything is soldered correctly and that you do not have any solder blobs that might cause a short or unwanted connection. Also ensure that it is soldered according to the schematic and that you did not make any logical errors.

-- (4) Have all your test equipment ( amp meters and volt/frequency meters ) hooked up. Beware that your ammeter will need to be able to measure more than what your power supply can deliver so that you know you wont damage the meter. I use a 600A clamp meter to measure anything above 10A.

-- (5) Power it on at a minimum of 12V. If your current draw spikes dramatically (above 10A ) AND you have no oscillation (eg 0Hz frequency) and no tank voltage, there is a short circuit and something is wrong.

-- (6) Assuming that everything is going well, you will have anyting from 15Khz to 200Khz resonant frequency across the tank cap, and you will have about 3 times input voltage across the tank (with no load in the work coil). Current draw will vary from a small amount to a huge amount, approaching the max your power supply can muster - depending on if there is a object in your work coil or not.

*Note :: Never , ever , let your work piece short out a part of the work coil. Current draw will spike and you may damage your transistors.


PART 3 : Tweaking and troubleshooting

Do you think your induction heater can perform better ? Does your induction heater not work at all? Well, this section is here to assist you in finding issues.

This part will be done in a Q&A style , for ease of reading and so that you dont have to read through paragraph upon paragraph to find a solution. (hint - use your browsers search function to look for specific issues such as "high current draw" by using CTRL+F)

-Q : My MOSFET's / IGBT's are heating up and there is very little current draw, or no object in the work coil !
-A : Check your resonant frequency, chances are your cap is too small, the work coil is too small or your MOSFET/IGBT's just dont enjoy the frequency you are running them at. However , if your frequency is below 60Khz, it is not a frequency issue and this answer might not help. Try a bigger cap anyway and see what it does.


-Q : My MOSFET's / IGBT's are heating up and there is very little current draw, or no object in the work coil, PLUS the frequency is within the operating range of my MOSTFET's/IGBT's!
-A: From my experience, this is caused by leakage inductance from the wires connecting to the work coil, tank cap, and transistors. Make them beefier and keep them as short as possible to stop the heating.


-Q : There is a audible sound coming from my heater, and the transistors are  getting really hot.
-A : It would be a good idea to check your resonant frequency -  if its below 20kHz youll be able to hear it. Try less turns on the work coil or decrease the tank capacitance.


-Q :I can still hear a audible sound coming from my heater, and my transistors are still getting hot !
-A :Chances are, there is a short over one of your transistors, or your work coil has become detached. power it off and make sure both transistors still work before you power it on again. If you still hear it check your work coil connections.


-Q :It draws a ton of current ,heats up the transistors , but it wont heat anything !
-A : Your work coil has too little turns, or your tank cap is too small. Check your resonant frequency, if its very high ( around or above 100Khz) then you should make your tank cap bigger and add more turns a little at a time until the current draw becomes reasonable.


-Q :Nope, still tons of current and heating of transistors - frequency is 0Hz!
-A : It is not oscillating. There is a problem somewhere, test your transistors and make sure all diodes are still working.



Thats all I can think of right now, ill post updates at a later stage as i think of improvements to this post.

[Mads Barnkob]

That is a great guide on testing and trouble shooting, thank you for taking the time to share that.

Not the most low inductance design, but I must also say that I am amazed of the discoloured alligator clips, you have without a doubt pushed around 50-100 Ampere through them!
You need some much better connection to the MMC than that, since you have the capacitors mounted in a ring, use at least 4 wires from the large wire, spread out evenly on the ring, to distribute current sharing on the capacitors as even as possible. Scroll down to current sharing between parallel capacitors on this part of my DRSSTC guide: http://kaizerpowerelectronics.dk/tesla-coils/drsstc-design-guide/dc-bus-capacitor/

Do you have any video of heating/melting some metal?

[Dylan Burger]

It was fun to do, and I enjoy sharing information I've learned along the way.

Yes, the clips have even started to glow at one point. I was planning to solder a coin type piece of metal to either side and then to that I'd solder the blue cables. I'll definitely take a look at your guide, and thank you for the advice.

Unfortunately no. I was planning to take a video this past weekend and ended up using the IGBT's in my sstc after I rebuilt it and added some modifications to the GDT. Unfortunately, the GDT was not the issue and I ended up destroying the IGBT's (after a successful 40v run, I switched to 110 to tune it further. After that it only made 3cm arcs.) I promptly decided I should reduce primary turns by tapping it off. I slowly decreased turns, still only 3cm arcs, untill I switched it on and a loud pop indicated that an IGBT  died.  Still no decent output and I am mildly disappointed in it XD.

[Mads Barnkob]

Unfortunately no. I was planning to take a video this past weekend and ended up using the IGBT's in my sstc after I rebuilt it and added some modifications to the GDT. Unfortunately, the GDT was not the issue and I ended up destroying the IGBT's (after a successful 40v run, I switched to 110 to tune it further. After that it only made 3cm arcs.) I promptly decided I should reduce primary turns by tapping it off. I slowly decreased turns, still only 3cm arcs, untill I switched it on and a loud pop indicated that an IGBT  died.  Still no decent output and I am mildly disappointed in it XD.

You should write a thread about the SSTC, let us see if we can troubleshoot it and get it working properly

[Dylan Burger]

You should write a thread about the SSTC, let us see if we can troubleshoot it and get it working properly

Thank you for the suggestion, I'll see if I can get a thread up today. I'll try and be as detailled as possible with all the specifications.  I appreciate the help