Yesterday I damaged a MOSFET I'm using in a simple control circuit, but it failed in a strange way and I'm trying to figure out what might have caused the failure and how I might prevent it in the future.

The MOSFET is controlling a 5v sensor by switching the low side of the power for the sensor. It's controlled by a 3.3v microcontroller. I was in the process of simulating a sensor failure by unplugging the sensor when the failure occurred. The failure either occurred when I unplugged the sensor, or when I plugged it back in, I'm not certain.

What I do know is that after the test, the MOSFET has a short between the gate and the drain (~100 Ohms), but high resistance between the gain and the source (~200 kOhms).

Most times when I've blown up a MOSFET, it's due to over current or over voltage and all three contacts are conductive -- or at least the drain is conductive to the source. This is the first time I've seen the gate conductive to the drain.

I know that a floating source or gate can cause issues in MOSFETs, but I didn't think a floating drain could be problematic. Am I mistaken? As I was disconnecting the sensor and causing the drain to float, I don't think this would be an over voltage or current situation, would it?

If the floating drain is the issue, I'm thinking of adding a pullup resistor between the drain and 5v on the board such that there's always a load for the MOSFET even if the sensor is disconnected. But as I'm not sure the actual failure mode, I don't know if that would actually help anything.

Any thoughts or advice would be greatly appreciated!

 

Hello,

Make the gate resistor much smaller.
Values between 22 and 100 ohms might work.

Bertus

 

Does the Gate resistor go directly to the Micro? or is there a connector. I got confused as to what gets plugged in and what is all the time connected. Gates do not like being plugged in.

 

Is the sensor inductive?

 

The sensor is not inductive, it’s a low current digital sensor. Only the sensor gets plugged in, the gate is connected directly to the microcontroller on the board.

@bertus, is the thinking that the large resistor is allowing a transient voltage spike at the gate or something?

thanks for the thoughts so far!

 

Static electricity from the open drain to gate? Many KV?

 

Usually symptomatic of a overcurrent large enough to melt the source bonding wire. The rest of the FET fails short then the source bonding wire melts due to the uncontrolled current through the short-circuit die.

 

I’m thinking ESD across the drain/gate is most likely. I don’t think it was over current as the source did not short through to either gate or drain.

if ESD on the floating drain was the cause, would a pull-up to 5v on the drain mitigation the issue in the future?