Setting OCD without blowing up transistors

[profdc9]

My tesla coil has a full bridge of FGH60N60SMD IGBTs.  I am trying to get a handle on how to set the overcurrent without blowing up the transistors.

From what I have been told, these can handle about 300 A peak current.  My current sense transformer is 625:1 consisting of two 25:1 cascaded ferrite transformers.  The burden resistor I am using is 4.7 ohms.

For 300 A, assuming a 625:1 ratio, the current through the transformer on the secondary is 0.48 A.  This means I should set the voltage on the comparator to 2.25 V, correct?  Is there any correction for the forward drop of the Schottky diodes?   I am using a 625:1 transformer because I hope to add a second set of IGBTs to make the peak current 600 A.

Dan

[oneKone]

I think Steve ward wrote in the drsstc log book about the voltage drop.

What I've been doing to set the ocd is set a voltage (let's say I want 300a ocd) at the comparator for a 200a limit. I then use a seperate ct connected to an oscilloscope and run the coil, slowly adjusting the voltage between runs until a 300a limit is achieved.

I've only done this method twice, and I'm also confident there's a better way of doing it. But it works for me.

[Hydron]

You can ignore the forward drop of the diodes, they do not affect the current passed through the burden resistor (and thus voltage across it). The critical things are the CT ratio and the burden resistor value, and it sounds like you have the right idea about setting the voltage based on this.

I have also in the past set OCD with a measure & adjust strategy using a second CT, but either method should work. If you're aiming for a particular current then be a little conservative with the OCD setting, as the current can still rise for up to another half cycle after a trip. However the 300A goal you're after isn't likely critical, so it's reasonable to just set the correct voltage for a 300A trip and leave it be (unless the coil has a particularly low impedance primary tank, which I assume is not the case here).