Hello. I took apart an old welder awhile back and kept a lot of the parts, but not the driver circuit. I basically have 2 brick IGBTs (each with two transistors so 4 transistors total), two large inductive chokes, a full bridge rectifier made with 4 high current diodes, some AC-DC power supplies (0 to +15, -15 to +15), a single loop current feedback device thing (looks like its an IC cause its got 4 small pins) and one of those gold, metal resistors of low value but high power rating. I think thats all of that I have. Oh and a big ass electrolytic capacitor (150,000uF).

Anyways. I wired up this circuit with just the DC components (not using the IGBTs) during my masters to make TiO2 using Ti electrodes and the resulting thermal plasma.

So now what I want to do is use this power supply to make a mean Tesla coil. I want to use an H-bride configuration, but im not familiar with driver circuits for such a low voltage, high current operation. This thing pulls like 200 amps at like 18 volts (when pulling an arc). But anyways I think its totally doable but im not sure where to start so I was hoping someone could give me some direction.

Also the driver circuit in the original welding circuit has a bunch of ICs on it (about 10-12), low voltage transformers, pretty sure a gate drive transformer, and plenty of capacitors and resistors and transistors. Also had 3 pots which were the only external controls to the circuit aside from the power switch. I know this is vague but oh well.

 

I see one big problem which is the frequency response (switching time) of the power devices.
Tesla coils depend on resonance that is usually at a fairly high frequency. Using mosfets ate frequencies that will keep them in their linear mode for any period of time will lead to a lot of heating.

 

So you're saying the IGBTs will not turn off fast enough and thus the next cycle will come and the IGBTs will still have partial saturation from the previous cycle? Im a civil engineer by trade so go easy on me haha.

And the resonant frequency of the coil ill be using is 339kHz and this frequency drops when the primary coil is introduced.

The IGBTs im using are 2MBI50L-120 and 2MBI75L-060. I attached the datasheets for them. The switching time is between 0.1us and 1us. It should have way enough time to switch for the coil im using then right?? Or am i reading it wrong.

 

The switching times in the data sheet are the fastest possible times with a driver able to remove the charge rapidly enough. a slower drive signal will result in a longer switching time. That applies to both inserting the charge to switch on, and removing the charge to switch off.

 

a slower drive signal meaning, for example, a 555 timer used as an interrupter?

im confused because Ive seen brick IGBTs used in other peoples Tesla coils circuits.

Are they using IGBTs that have fasters switching speeds?

 

What frequencies did those Tesla Coils operate at? A lower frequency could work better with a slower switcher, sort of.