Mosfet failure mechanism / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Mosfet failure mechanism

1 2 3 next

Move Thread

LAN_403

Dr. Dark Current

Wed Apr 18 2007, 05:09PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

I was playing with a simple mosfet flyback driver, using the well-known irfp250. It was working great until the mosfet suddenly popped, without any heating. How can a "fully avalanche rated" mosfet die without any warning (heating)?? (if it was overvoltage the mosfet would avalanche and start to get hot before the failure right?)

HV Enthusiast

Wed Apr 18 2007, 07:11PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

you can pop a MOSFET with overcurrent and not have the MOSFET actually feel warm.

Dr. Dark Current

Wed Apr 18 2007, 07:16PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

EastVoltResearch wrote ...

you can pop a MOSFET with overcurrent and not have the MOSFET actually feel warm.

I think it was not overcurrent (irfp250 is rated 30A and the driver draws ~500mA@24V). the driver was working for many minutes without any mosfet heating, then when drawing an arc the fet suddenly popped. The reason why I am asking is because I want to build the whole adjustable flyback psu into a box, so it needs to be reliable.

thedatastream

Wed Apr 18 2007, 07:36PM

Registered Member #505 Joined: Sun Nov 19 2006, 06:42PM
Location: Yorkshire!
Posts: 329

Can you give details of the failure? All pins short together, short d-s, open d-s, other?

I would imagine that a MOSFET could fail from a sufficiently large over voltage condition of the avalanche rating was exceeded by a sufficient amount.

Marko

Wed Apr 18 2007, 08:11PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Do you have a schematic of the circuit you are using currently?

That mosfet is extremely rugged regarding avalanche; Datasheet claims that it can avalanche as much current as it switches wich would equate 6,6kW of dissipation; so only way avalanching can kill the mosfet is overheating.

Electroholic

Wed Apr 18 2007, 08:23PM

Registered Member #191 Joined: Fri Feb 17 2006, 02:01AM
Location: Esbjerg Denmark
Posts: 720

do you have like a cap across the supply rails and the cap just discharge into it when it shorts out?

Dr. Dark Current

Wed Apr 18 2007, 08:31PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Firkragg wrote ...

Do you have a schematic of the circuit you are using currently?

That mosfet is extremely rugged regarding avalanche; Datasheet claims that it can avalanche as much current as it switches wich would equate 6,6kW of dissipation; so only way avalanching can kill the mosfet is overheating.

The schematic is a standard single-ended mosfet driver (the fet is driven by the great pwm ic sg3525, and has a 100n cap across d-s). The failure was all pins shorted together. It is maybe possible that the fet was slightly damaged before. i think i will buy some new ones and try again
edit>

Electroholic wrote ...

do you have like a cap across the supply rails and the cap just discharge into it when it shorts out?

yes, i actually heard pop sound as the cap discharged into flyback's primary.

Marko

Wed Apr 18 2007, 08:34PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Are you using schottky diodes or a zener on SG3525 output to clamp off the miller charge?

PS. i'm not thinking rationally.. I really don't believe this can be cause.. there must just be some more major flaw somewhere.

If you are shopping, try using a higher voltage mosfet like IRFP450.
It allows for higher primary voltage wich seems to outweight the con of higher on-resistance.

Steve Conner

Wed Apr 18 2007, 10:12PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

If the power dissipation suddenly increases, the die can melt before the heat has time to spread out of the package. Richie Burnett's old SSTC used to self-destruct within a few seconds of turn-on that way. He fixed it by getting rid of those silicone thermal washers and bolting the devices directly to four isolated heatsinks.

Firkragg: That avalanche current spec will be a peak rating, of course

The continuous avalanche rating will be probably several tens of watts. I was once repairing a PFC front end, and forgot to put the boost diode back in when putting it back together, so it cranked itself to full duty cycle and about 250W of input power went into avalanching the switch (an IRFP460) It actually lasted for a good few seconds, and got burning hot before dying.

Marko

Wed Apr 18 2007, 10:56PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Firkragg: That avalanche current spec will be a peak rating, of course

Surely, what I wanted to point out is that mosfet simply can't irreversibly break down while avalanching, it can just die from overheating. (pulse with limited by maximum junction temperautre).

Surely if you put 33A trough it in avalanche you get 6,6kW of dissipation, so at that point mosfet wouldn't last very long, but practically there's little chance for direct irreversible breakdown with such a robust rating.

I was once repairing a PFC front end, and forgot to put the boost diode back in when putting it back together, so it cranked itself to full duty cycle and about 250W of input power went into avalanching the switch (an IRFP460)

I'm sorry about that..

But still, I never seen any of those newer mosfets to die from avlanching without getting hot.

SUrge current can kill the mosfet in some occasions, but gate breakdown is still most probable cause of 'cold death'.

1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob