I have a circuit using an IRF540 well within its current and voltage ratings.
The MOSFET failed. 54Ω drain-source and 5.1Ω drain-gate.

Would overheating result in a drain-gate short or was there some other problem that caused this?

 

Would overheating result in a drain-gate short or was there some other problem that caused this?

I think it's unlikely that overheating would cause that. My guess would be some sort of transient voltage spike that punched through the gate oxide.

 

I second OBW0549's analysis.

 

+1

 

That's what I was afraid of. I can increase cooling without too much effort but I don't know where any spike could have come from. The circuit is below. I could put a 15V zener directly between the gate and source. Any other suggestions?

 

FWIW, Googling on "failure mechanisms in power mosfets" produced a wealth of information. These are the first three items (there are many more, I just didn't explore further):

http://www.ti.com/lit/an/slyt502/slyt502.pdf
https://www.vishay.com/docs/69087/an911.pdf
http://www.4qd.co.uk/docs/mosfet-failure-mechanisms/

The last document provides a definition of MOSFET: Magically Obliterated Smoke and Fire Emitting Transistor. Heh.

 

Having read the the third link above, I will be adding a capacitor and ferrites at the motor. This is fairly easy to do.

 

I have added more heatsinking and capacitors (10n & 100n) across the motor and added ferrite rings on the motor wires.
A few, perhaps a dozen, on-off cycles and the MOSFET is short again (all three terminals).
I am going to use a relay instead.

 

Hi Albert,

Just in case you intended to add a diagram in post #5, it's not visible for me. Not that I could help but I thought I'd point that out.

 

Yes, it seems I did forget to include the diagram but it's too late now. It's been overwritten by the relay version.

 

Hi,

I would recommend you to use IRF540. IRF540 is an NPN MOSFET, which gives excellent results, when it comes to fast switching. It works on HEXFET technology and is capable of perforing under critical temperatures. You should have a look at IRF540 Pinout and laso downlaod its Proteus simulation. It will help you in its better understanding.

Thanks.

Moderators note : removed hidden link which is site promotion

 

Zombie thread, but ...

The IRF540 is, by modern standards, a very mediocre MOSFET. For reasons I don't understand, possibly that a few million got dumped on the hobby market at low price, it pops up all over the place in the hobby world. If you don't get them at a deep discount there a many much better FETs that cost less.

Note that #11 is yet another occasion of the same person resurrecting a zombie thread in order to promote the webiste at the link.

 

Note that #11 is yet another occasion of the same person resurrecting a zombie thread in order to promote the webiste at the link.

Did you report that post? I usually do when I see them, even if they are borderline. This one seems blatant.

 

BTW the relay switch is still going strong

 

BTW the relay switch is still going strong

Because of the popularity of the 540, I think I've heard of counterfeits being out there. Any chance you ran into some?

 

I don't remember where they came from.
It is switching the spindle (48V 6A) on my CNC so it's only one on/off cycle per tool and the relay seems to be able to cope with that.