Author Topic: Blown mosfet : I don't know why  (Read 4565 times)

0 Members and 1 Guest are viewing this topic.

Offline ratataxTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: fr
Blown mosfet : I don't know why
« on: October 02, 2019, 10:17:19 pm »
Hi

I use a P-channel mosfet for polarity protection in my circuit (an AO3407 http://www.aosmd.com/pdfs/datasheet/ao3407.pdf)  it's super simple :



Works great, I draw about 1A at 12V and it doesn't get hot.

However when i was doing some tests it suddently smoked. It still allows current to flow, but has a small burn mark close to the Gate pin. I was just probing some things on a totally unrelated part of the circuit, and I'm pretty sure I didn't do any short circuits, in fact my circuit kept running after that....

Bad part ? Is it possible that ESD causes this ? Is there a way to make it more sturdy, for safe use in a consumer product ? I'd like to understand what happened  :o
 

Offline ratataxTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: fr
Re: Blown mosfet : I don't know why
« Reply #1 on: October 02, 2019, 10:46:21 pm »
From https://www.4qd.co.uk/docs/mosfet-failure-mechanisms/ :

Quote
The cause of this failure is a very high voltage, very fast transient spike (which may be positive or negative going). If such a spike gets onto the drain of a MOSFET, it gets coupled through the MOSFETs internal capacitance to the gate. If enough energy gets coupled, the voltage on the gate rises above the maximum allowable level – and the MOSFET dies instantaneously. The process takes less than a nano-second! The initial spike destroys the gate-body insulation, so that the gate is connected to the body. Once that has happened, the MOSFET explodes in a cloud of flame and black smoke. We have one documented case where the battery wire worked loose, causing a spark. It must have been this that caused the gate breakdown for the explosion of flame and smoke did not happen until the battery wire was re-connected some time later! Which demonstrates how very difficult cause and effect can be to connect!

I think that's was happened here, I often plug/unplug the power supply wires to do some tests, I noticed it had a small spark the last time I plugged it in, then the next time i powered it on it smoked.
But I'm not entirely sure how I prevent that. By adding a resistor on the gate ?
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1408
  • Country: us
Re: Blown mosfet : I don't know why
« Reply #2 on: October 02, 2019, 10:51:59 pm »
Deleted because wrong. Oops. That'll teach me to dash off a quick reply.
« Last Edit: October 03, 2019, 09:20:05 am by MagicSmoker »
 

Offline ratataxTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: fr
Re: Blown mosfet : I don't know why
« Reply #3 on: October 02, 2019, 10:59:52 pm »
I'm confused, most websites show the load connected to the Source and it works fine as i tested it.

But some adds a resistor to the gate, or even a zener sometimes to prevent overvoltages
 

Offline tboy

  • Contributor
  • Posts: 14
  • Country: ca
Re: Blown mosfet : I don't know why
« Reply #4 on: October 03, 2019, 12:38:07 am »
According to:
https://www.infineon.com/dgdl/Reverse-Batery-Protection-Rev2.pdf?fileId=db3a304412b407950112b41887722615
You have it wired correctly.  A better MOSFET symbol in your schematic would be a plus.  :)
 
The following users thanked this post: Mechatrommer

Offline amspire

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Blown mosfet : I don't know why
« Reply #5 on: October 03, 2019, 01:17:58 am »
Mosfets are pretty tough, but more then 20V source-gate will kill them instantly. It takes very little energy - microamps can blow a tiny hole in a mosfet gate. What is the nature of the load and 12V source. Is the load inductive - such as motors?

In particular, if you have an inductive load and the 12V is suddenly removed, you have a 100% guarantied blown mosfet.

You probably want to at least protect the gate. Put a 100ohm resistor in series with the gate and put a 18V zener from source to gate.
« Last Edit: October 03, 2019, 01:30:06 am by amspire »
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Blown mosfet : I don't know why
« Reply #6 on: October 03, 2019, 06:19:16 am »
Limited number of possible reasons. Rule them out one by one.

1) D-S overvoltage, including any short spike
   - Stray inductance in the system, causing ringing. Ringing caused by stray inductance in input wires can easily go 2x the supply. A lot of inductance in the output (motor, relay) can generate a very high voltage, easily 10x the supply, in which case a properly placed freewheeling diode is needed (to form a tightly laid out half bridge).
   - Switching too fast increases the spiking caused by stray inductance. Gate resistor can slow it down.

2) Overcurrent, including any short spike
  - Capacitors in the output? They are short circuits and take "infinite" current at turn-on. Look at the pulse safe operating area graphs, taking the capacitor ESR and charging time (depending on amount of C) into account

3) Gate overvoltage
  - including ESD shock, especially for small MOSFETs with too little protective gate capacitance

4) Excess thermal dissipation (I^2 * Rds(on))
  4b) Excess thermal dissipation in case the device is not conducting as well as you think it is (desaturation):
       - your gate voltage is too low, all the time or during some edge case you didn't think of
       - you are switching slowly and spending time in high-loss state
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 7454
  • Country: pl
Re: Blown mosfet : I don't know why
« Reply #7 on: October 03, 2019, 06:56:02 am »
MOSFETs like to fail short on all 3 pins if overheated. If it's burnt near the gate pin, it may have shorted out completely and vaporized its gate bonding wire due to the resulting short circuit to ground. If you have a capacitance meter you would see zero capacitance from gate to the other terminals. And a short between drain and source in both directions.

To add to all the possible reasons given so far: you may have shorted the load or shorted the power input terminals, allowing load capacitors to discharge through them. This configuration doesn't actually prevent reverse current flow, it only prevent reverse current flow when load-side voltage is less than a few V.
 

Offline ratataxTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: fr
Re: Blown mosfet : I don't know why
« Reply #8 on: October 03, 2019, 09:34:33 am »
Thank you for the answers !

Load is the DC/DC converter then a raspberry Pi and some other digital circuitry. Nothing inductive, but a small capacitance since I have a few (3x22uf) MLCC capacitors at the output of the DC/DC converter

I'd like to avoid any abrupt current spike at start, currently there is a small spark.
 
Would something like this be a good start for protection ?



(I'm trying to reuse values I already use in my circuit -- hence the 75K resistor, and zener would clamp at 3.6V.. I guess it would work OK since gate threshold is around 2V according to the datasheet)
« Last Edit: October 03, 2019, 09:39:28 am by ratatax »
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
Re: Blown mosfet : I don't know why
« Reply #9 on: October 04, 2019, 08:27:20 am »
MLCCs used in decoupling applications can cause quite large voltage spikes in some circumstances. If you have any significant inductance in your supply wiring (quite likely by the sound of it) then have a read of this app note
 
The following users thanked this post: nuclearcat

Offline ratataxTopic starter

  • Regular Contributor
  • *
  • Posts: 134
  • Country: fr
Re: Blown mosfet : I don't know why
« Reply #10 on: October 05, 2019, 05:01:43 pm »
This is super interesting. I measured my 12V in when switching on and it has a peak to about 20V. It shoudn't be what destroyed the mosfet, but I'll try to damped it a bit just to be safer on the long run !
My custom DC/DC converter has a 10 uF + 100nF mlcc input cap.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Blown mosfet : I don't know why
« Reply #11 on: October 05, 2019, 05:38:34 pm »
Add a plain old cheap electrolytic capacitor with C many times the C of  the MLCCs. For your case, say anything over 47uF does the trick.

Having high-ESR capacitance on the DC bus dampens oscillations, but also limits the voltage rise in case you have an inductive load discharging back to the bus.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf