What destroyed this MOSFET?

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
could have two reasons
a:coil has no diode negative pulse could destroy s d junction
b:gate is blown up voltage to high also missing a small resistor to gate >=100Ohm)
The input cap gate is at high freq a low impedance absorbing current resistor will limit this.
D1/C1 absorb the positive-going flyback from the inductor - in fact, that's fundamental to the operation of the circuit.
It's not shown in the schematic for simplicity, but in fact I have a 2 ohm resistor in series with the gate. The driving circuit is an MIC4452YM, theoretically capable of supplying 12 amps peak.
 

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
The inductive kick from the coil blew the gate oxide while the gate and source were interchanged. The max voltage from Drain to Gate is like 20V. YOur circuit produces hundreds of volts.
In order for there to be an inductive kick, current would somehow have to flow through L1 in the first place. With the incorrect circuit on the right, I still can't figure out how that could happen.
 

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
Thanks everyone for your thoughts. It occurs to me that many people are looking at the circuit on the left and trying to treat this as a design error. What I'm hoping is that someone more insightful than myself can look at the assembly error, the circuit on the right, and explain to me the sequence of events that led to the failure.

I will drop the issue if need be, but I have a monkey-like curiosity for more details of what happened. At my previous job, I was paid handsomely to do forensic analysis on this sort of thing. Without a theoretical model of what happened, we would not know whether something else was damaged or weakened in the process of failure.
 

crutschow

Joined Mar 14, 2008
34,281
I did an LTspice simulation of your incorrect circuit but none of the voltages or currents showed anything that would cause your observed failure.

What is the source of the 10V, 10us pulse?
 

hp1729

Joined Nov 23, 2015
2,304
View attachment 98111
First of all, I fully acknowledge that I did something stupid, so I'm not looking for "user error" as the root cause.;)

I was bringing up the circuit on the left. It's part of a buck-boost converter, but I've simplified the schematic for illustration. Q2/D2 (and Q1/D1) are actually one part, an FDPF085N10, an N-channel MOSFET with body diode. Vdd is 24 volts. When a button is pressed, a single 10-volt, 10 uS positive-going pulse is driven into the MOSFET by a driver (not shown).

I tested the MOSFET before using it, and it was good. Then I soldered it onto the PC board, turned on the power, and hit the button a few times, expecting the voltage to jump at least a little. Nothing happened. In fact, I found that my 24-volt supply had gone into shutdown.

Ok, my fault, I had inserted the MOSFET backwards, and in fact, gate and source were swapped, resulting in the circuit on the right. I removed the MOSFET and found gate, source, and drain fused together. I replaced the MOSFET with one oriented in the correct direction, and everything worked fine.

So the question: in technical terms, what was the mechanism of the failure?
When you put the MOSFET in backwards that built in diode was forward biased. Almost a dead short from ground through the coil to +24 V.
 

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
I did an LTspice simulation of your incorrect circuit but none of the voltages or currents showed anything that would cause your observed failure.

What is the source of the 10V, 10us pulse?
The drive originates in a PIC processor (PICAXE, actually). The single pulse is driven through an MIC4452YM mosfet driver into the gate of the main switch.

It gives me some consolation that no one so far has found a satisfactory explanation. I did test the MOSFET in before soldering into the circuit, but maybe I damaged it during soldering? When I put in the replacement part the right way, I was careful to put a jumper while between gate and source for the whole time the thing was being soldered in.
 

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
When you put the MOSFET in backwards that built in diode was forward biased. Almost a dead short from ground through the coil to +24 V.
Yes, that was my first thought, too, but I realized that by putting the part in "backwards," I was swapping gate and source, not source and drain.
 

sailorjoe

Joined Jun 4, 2013
364
Yes, that was my first thought, too, but I realized that by putting the part in "backwards," I was swapping gate and source, not source and drain.
Look at figure 16 of the datasheet. It indicates that the falling edge of the input pulse may have caused the transistor to forward bias, including the intrinsic diode, and without any current limit, the whole thing avalanched to destruction. Of course, it's just a guess.
 
Last edited:

ScottWang

Joined Aug 23, 2012
7,397
If the coil is the load, but it has no any protection for mosfet from emf, when the power is turn off, the coil reverse the polarity, there is a positive on Vs and the current via Vs → Vd → Gnd → 24V, the current loop destroyed the mosfet.
 

DerStrom8

Joined Feb 20, 2011
2,390
Excessive voltage will punch through the gate-source oxide layer and result in permanent damage. This seems obvious enough, but it is not so obvious that transient gate-to-source overvoltages can be generated that are quite unrelated to, and well in excess of, the amplitude of the applied drive signal. The problem is illustrated by reference to Figure 2. If we assume that the impedance, Z, of the drive source is high, then any positive-going change of voltage applied across the drain and source terminals (caused, for example, by the switching of another device in the circuit) will be reflected as a positivegoing voltage transient across the source and the drain terminals, in the approximate ratio of:

upload_2016-1-8_9-19-24.png

The above ratio is typically about 1 to 6.
 

crutschow

Joined Mar 14, 2008
34,281
.......................
It gives me some consolation that no one so far has found a satisfactory explanation. I did test the MOSFET in before soldering into the circuit, but maybe I damaged it during soldering? When I put in the replacement part the right way, I was careful to put a jumper while between gate and source for the whole time the thing was being soldered in.
Given the lack of a good theory for how the incorrectly wired circuit as shown could generate currents or voltages large enough to damage the MOSFET, and given that my simulation also showed no such voltage or currents, then perhaps we use Occam's razor and conclude that it was damaged during the initial soldering process.
Wrapping a small bare wire around the gate and source during installation of a MOSFET is always a good procedure.
 
Last edited:

crutschow

Joined Mar 14, 2008
34,281
A soldering iron, if not properly grounded to the workbench antistatic mat (which we all have, right?) can zap a MOSFET during the soldering process.
Perhaps that's what happened.
 

Thread Starter

Roderick Young

Joined Feb 22, 2015
408
Look at figure 16 of the datasheet. It indicates that the falling edge of the input pulse may have caused the transistor to forward bias, including the intrinsic diode, and without any current limit, the whole thing avalanched to destruction. Of course, it's just a guess.
Excessive voltage will punch through the gate-source oxide layer and result in permanent damage. This seems obvious enough, but it is not so obvious that transient gate-to-source overvoltages can be generated that are quite unrelated to, and well in excess of, the amplitude of the applied drive signal. The problem is illustrated by reference to Figure 2. If we assume that the impedance, Z, of the drive source is high, then any positive-going change of voltage applied across the drain and source terminals (caused, for example, by the switching of another device in the circuit) will be reflected as a positivegoing voltage transient across the source and the drain terminals, in the approximate ratio of:

View attachment 98192

The above ratio is typically about 1 to 6.
Interesting. Possibly ringing caused the transistor to turn on for a bit. Could be.

If the coil is the load, but it has no any protection for mosfet from emf, when the power is turn off, the coil reverse the polarity, there is a positive on Vs and the current via Vs → Vd → Gnd → 24V, the current loop destroyed the mosfet.
This is interesting, too. Perhaps some kind of transient condition either at power-up or power-down. The problem with either of these is that current somehow needs to be flowing in L1 at some time. There could very well be current flowing, but unfortunately, I've been unable to convince myself how that could happen with the incorrectly wired circuit.

One experiment that I can think of is to get yet another part, be more rigorous in anti-static precautions, intentionally solder it in backwards, then power up and down to see if anything dies. If it doesn't, I could try pulsing the gate again to see if it dies. If I was at work, I would have already done something like that. These mosfets are perhaps $2 a piece, and I do have a bunch of them. But since I'm at home, those are MY two dollars, and moreover, nobody is paying me to find the answer. I'm also loath to keep soldering and unsoldering parts from the board - I might ruin it. The board has another $250 worth of stuff on it. If it was work, I wouldn't blink, but when it's my money, it's a different story...
 
Top