Mosfet gate pulsed high Voltage survival??

[keshab.patel]

Hi everyone. I am working in a automotive project. The circuit consists of a mosfet with simple 1 K resistor gate drive. So it needs to be tested with ISO 7637-2 pulses.

The absolute maximum gate voltage  is rated at +-20V. Is it dc +-20V ?

Can it survive if a 1 micro second pulse of 100V is applied through a 1k resistor to the gate. 

[amspire]

Basically, you must not let the gate voltage ever exceed the limits. If the limits are exceeded, the mosfet may easily keep working, but may start to degrade over time leading to failure.

Now will 100V/1us cause the gate voltage to go over 20V? Depends on the gate capacitance and the normal maximum gate voltage. It probably will. Adding zeners to the gate will protect it.

[Marco]

Can it survive if a 1 micro second pulse of 100V is applied through a 1k resistor to the gate.

Protected gate MOSFET yes, unprotected gate MOSFET only if the gate capacitance doesn't charge up too high in that microsecond.

AEC-Q101 automotive rated MOSFETs will have ESD grades in their datasheet. I'd still pass on any MOSFET which doesn't mention gate protection though.

[T3sl4co1l]

Why would you ever consider putting a naked, unconditioned, unprotected, unfiltered automotive wire, up to a transistor?  Any terminal, let alone the gate?

This sounds like the beginnings of a terrible design, top to bottom...

Tim

[keshab.patel]

 I have seen a design like this implemented. So wondered...how!

[T3sl4co1l]

Hrm..

The rating is peak, unless otherwise stated.

Sometimes a peak rating is given, e.g., 20V DC, 30V peak.  In that case, the pulse duration will also be given.

If the gate charge is on the order of 10nC or more, then the 1k resistor will provide a filtering effect, significant for a pulse of 1us length.

In practice, an average power MOSFET fails around Vgs = 60 to 80V, though it could be as low as 30 or 40V, hence the conservative rating of 20 or 30V.

As mentioned, a MOSFET equipped with ESD diodes may also survive such treatment.

And now that you know... you know why not to do it.

Tim

[danadak]

The issue is Gate Ox rupture of the MOSFET, and that can happen
for very fast pulses. Generally catastrophic.

http://ambientelectrons.org/wp-content/uploads/2012/02/presentation.pdf


Regards, Dana.

[RoGeorge]

It can survive, but don't count on it. Protect the gate.

Here is an experiment where the transistor is permanently altered by a Vgs overvoltage: https://hackaday.io/project/16616-your-mosfet-is-not-good-enough-then-modify-it

[amspire]

I once worked at a factory building mosfet mains inverters that had poor ESD procedures. A few devices failed in the factory on test, but the incidence of blown mosfets in the field was horrific.

Full ESD procedures were implemented in the storeroom and manufacturing. The return rate dropped to near zero.

Basically, if the gate gets a microscopic puncture, the mosfet often keeps working, but there is a flaw that gradually deteriorates until failure.

[CraigHB]

I think everyone who builds their own circuits has blown out a MOSFET gate at one point.  They can do some funny stuff, like keep working with a current drain through the gate.  You have really have to protect them if any transient voltage can make its way there.  The circuit should be designed to to survive a lot more than the minimum allowance.  If nothing else use a protected one.  You see them with built in clamp diodes pretty commonly.  For any automotive application those should be used at the least.  Automotive stuff has to be pretty tough, big transients in the electrical system are commonplace.

[Nerull]

I once worked at a factory building mosfet mains inverters that had poor ESD procedures. A few devices failed in the factory on test, but the incidence of blown mosfets in the field was horrific.

Full ESD procedures were implemented in the storeroom and manufacturing. The return rate dropped to near zero.

Basically, if the gate gets a microscopic puncture, the mosfet often keeps working, but there is a flaw that gradually deteriorates until failure.

One company I worked for built fluid warmers that used PWMed mosfets to drive a heating element in a control loop. These had the bad habit of self-immolating during testing, until the non-ESD safe plastic sheet someone had hung up around the makeshift spray-on conformal coating booth that the operator had to push open to get inside was replaced with a setup that didn't make the operators hair stand up.

Management complained that ESD wasn't that big of a deal for this circuit until the failure rate went to near zero after I took a pair of sheers to the plastic.