MOSFET failure mode: short between gate and drain?

Thread Starter

alexandicity

Joined Jun 28, 2013
9
Hello!

I have a PCB with ten n-channel enhancement FETs (specifically, IRF6662s). Three of these have failed, apparently at the same time, and now don't ever turn off. Investigating, I see that there is a ~150 ohm path between the drain and gate, which could cause this stuck-on state (and that is not present on other, working FETs).

Is a short between a gate and drain a common failure mode of MOSFETs? Is there a cause that explains why all three are suffering the same effects with about the same magnitude? They all share the same drain net, but are isolated elsewhere. They did all get heated quite a lot when I tried to do some rework on them with my iron, which may have caused some damage...

Thanks!
 

Thread Starter

alexandicity

Joined Jun 28, 2013
9
Yeah - drain and source shorts I've seen and understand OK. I was just perplexed to see this gate-drain short.

Relevant schematic area:

upload_2018-1-20_15-52-20.png
The drain is connected to the left (yellow) lead which is a 7.2V supply., which is shared between all the FETs. The right lead out of the source goes to a load, and then to ground. The gate network is driven by this surge suppressor IC which drives the gate to a set voltage above the source. This works fine on most of the FETs except these three that have failed..
 

dl324

Joined Mar 30, 2015
17,131
Gate oxide wearout can occur normally or breakdown can caused by Electrical Overstress (EOS) or Electrostatic Discharge (ESD). Determining source of failure requires tearing down the device.

Could the gates of any of the failing MOSFETs been subjected to ESD or EOS?
 

Thread Starter

alexandicity

Joined Jun 28, 2013
9
I am working an an ESD controlled area, and that 10Ω I would have thought would have protected it against the worst discharges, but an ESD is always possible. I struggle to think how it would affect three devices in such a similar way (unless a discharge across drain-source, which is partially common).

For EOS, I don't think so. The voltages in the system are at most 10% of the FETs' rated capabilities - and I don't see any other sources of potential...

Can thermal damage cause this gate oxide wearout?
 

Attachments

Thread Starter

alexandicity

Joined Jun 28, 2013
9
Hi Albert - yes, that seems to be a possibility. Certainly a cause that affects the drain but damages the gate area is something that fits here. Seems like a rare failure, but certainly a possible...
 

djsfantasi

Joined Apr 11, 2010
9,188
Today I had a curious problem, a power mosfet that was good, after mounting it had a resistance gate source of few ohms. Source drain was ok, I have supply a large current gate source until it has opened. Then the mosfet worked correctly.
This post is considered “necroposting”, which is bringing a dead post back to life. You, the TS, should start a new post for your question. This is over two years old and should be left to rest peacefully.
 
Top