Primary transistors failed open in generic power supply

[Th3_uN1Qu3]

Okay, i finally got to the bottom of something that was puzzling me for quite a while. On many of the dead PSUs i have scavenged for parts, the primary switches tested open on c-e. As i didn't have some good ones to compare at that time, i simply threw them in the "doubtful" bin.

About last week i blew one of the generic PSUs i had recapped, a "Smart Tech" LC-8460BTX, by trying to draw LOTS of current from 12v for an extended amount of time. I think i was drawing close to 400W and we all know those generic PSUs can barely do 2/3 of their ratings. There was a muffled pop and the primary heatsink was BURNING hot. The fuse did not blow, and 5vsb was still present. Primary transistors tested open.

I swapped them today with known good ones, and it fired right up. Running at 300W now no problem and sitting within ATX spec. So there you go - next time you find a unit that doesn't want to power up yet the fuse isn't blown, you have 5vsb and the diodes are good, open primary transistors are a common failure mode. Something to keep in mind.

[goodpsusearch]

Re: Primary transistors failed open in generic power supply

Sorry for the stupid question but how you test transistors with a multimeter?

[c_hegge]

Re: Primary transistors failed open in generic power supply

Set the MM to resistance and put the probes between the legs. If it shows open all the time, then the transistor is open. If it shows 0, then it's shorted.

[goodpsusearch]

Re: Primary transistors failed open in generic power supply

Until now, I thought when they fail, the transistors are shorted. Interesting...

[Wizard]

Re: Primary transistors failed open in generic power supply

I had some that failed with some resistance, not quite shorted, not quite diode voltage, simply resistance.

Cheers, Wizard

[cadiman]

Re: Primary transistors failed open in generic power supply

I think it depends on the type of transister and current draw. I have always seen mosfets and scr`s fail "on" ( power supplies and motor drives) while other types of transisters failed as Wizard talks about. I have never had an open transister on a power board, but you see and learn new things everyday. I never had to replace capaciters either until a couple of years ago.

[Th3_uN1Qu3]

Re: Primary transistors failed open in generic power supply

Set the MM to resistance and put the probes between the legs. If it shows open all the time, then the transistor is open. If it shows 0, then it's shorted.

Close but not quite. You use the diode test mode. In my experience transistors that experience a high current spike such as a short on the output on a psu thats not protected, they fail short.

The ones in my PSU failed due to excess heat, heatsinks too small. I think the bond wire failed, not the silicon.

[ben7]

Re: Primary transistors failed open in generic power supply

I have had transistors and MOSFETs fail open circuit, after running hot all day long for two years.

I think maybe the transistor might actually burn open inside especially if there is a large 7A+ fuse or a slow-blow fuse.

-Ben

[Badcappy]

Re: Primary transistors failed open in generic power supply

Close but not quite. You use the diode test mode. In my experience transistors that experience a high current spike such as a short on the output on a psu thats not protected, they fail short.

The ones in my PSU failed due to excess heat, heatsinks too small. I think the bond wire failed, not the silicon.

That "muffled pop" is a small explosion that probably blew apart the die, hence the fail open.