IGBT fail E C short

What's the IGBT's transfer function?

Thanks, Rich

Shoot thru? What is the Vce of the device and what is your supply voltage.

Cheers

The LC part makes this a non-simple pass regulator - component values?.

IGBTs have SOA limits similar to bipolars with large emitter-base shunt resistor values.

RL

This is a switching regulator, right? IGBTs can not be EVER used in the linear region, their thermal characteristics make them "hog" current to the center of the device, cooking the hottest few cells in the chip, then the failure spreads.

So, if you can't let them run linear, you have to drive the gate HARD, to get them from the "off" state to the "on" state abruptly, and vice versa. A weak driver will allow the device to dwell in the linear region long enough to overheat one spot on the die, leading to failure. Since the "hot spot" can be arbitrarily small, even very low currents can have this effect.

And, you have to take the Miller effect into account, where the swinging collector voltage couples to the gate, and attempts to defeat the gate driver, right at the worst region for that to happen.

Jon

Hmm, I've never heard that. Say, wouldn't the same physics apply equally to BJTs, or linear MOSFETs?

IGBTs have been used in a few audiophile amplifiers (read: probably poorly engineered, with bad results, but it worked nonetheless, and well enough for its creator) and I don't know of them exploding. I don't know what ratings they used, probably something heavily overrated, since you can't find IGBTs that do exactly 100V and 10A and dissipate 300W after all.

Tim