Why do my MOSFETs keep failing?

[dr_p]

Hi to all!
First of all, I read 4-5 forum threads about mosfets failing, but I still don't get it.


I want to add a 1K resistor parallel to a 30-300R car fuel sensor. The scope is to temporarily modify the fuel level seen by the computer. I won't go further into this, it doesn't matter since the MOSFET keeps failing on my bench, nevermind the car.


So I have a 555 wired as a monostable, outputting 500ms HIGH, then LOW all the time. I built the circuits shown in the pictures, all powered from 9-14V.
The 555 works as intended all the time, the problem is that the FET (2N7002) fails after some use, with R(ds) = 3...5K when OFF. So after failing, with the gate grounded, there are 3 to 5 K between Drain and Source. This kept happening 6-7 times, but I have no ideea why. The Drain was connected to the 1K resistor, and sometimes scoped (1M) with a reference to V+, sometimes left floating, sometimes added a LED to V+.
First I had the 555's output straight in the FET, then after 2 deaths I added a series 100R, then after 2 more deaths I added a zener to GND. I know it should have been next to the FET, not before the series 100R. I then exchanged the Zener for a 10k to GND. In all of these, the FET keeps dieing apparently at random.


Since there is virtually no current through it, I can only think of too high Vgs or Vds (maybe from floating Drain).
My power supply is floating, the scope is portable (and not charging) and I don't have an ESD mat.


My next thought is to add a 1M resistor between Drain and Source.

[Kjelt]

Put the 10k resistor directly across the gate and the source and from the gate a 10k to the driver ic works for me.
Still problems: is your solderiron referenced to ground or also floating? Which temperature do you solder and how long. Discharge yourself before soldering, keep the moistlevel in the room at >50%. Those small fets are real PITA sometimes.

[Niklas]

Is the sensor connected to some ECU or only to your PCB? Not all ECUs have low side (grounded) sensors. Reverse bias of gate-source?

[mij59]

Drain source break down maybe caused by exceeding the max. drain source voltage.
This could be due to switching a ( parasitic) inductance.

Try a diode freewheel diode, connect the  anode of the diode to the drain of the mosfet, and the cathode to +14V.

[dr_p]

I will reposition the 10k directly across the gate-source.
Soldering station is floating, a Weller set to 300 C, ~2 sec for each pin. I have no ideea of the moisture level.
I'll use a lower temp, around 250 C and try to ground the tip (and myself) via a 1M resistor.


The mosfet is not switching anything, actually the PCB is not connected to anything except maybe a scope or an LED. I'd like to stop it popping mosfets before I go any further. It's a grounded sensor, I measured. As a matter of fact, the device works, I have one installed in my car, but I want to build more of them and I'd like to learn why the mosfets keep dying on me.

[dr_p]

I think I got it now. I assembled 30 boards yesterday and no FET was harmed in the process. What I did in lower the temp to 250, ground myself an the tip (via 1M) and last but not least, I installed and tested everything, then added the MOSFETS as a very last step.


All the boards work fine  .


Thank you all for your advices!

[Zucca]

...the tip (via 1M)...

Wasn't the tip already connected to ground earth?

[dr_p]

no, it was floating