Exploded transistor

[jasonbay13]

i have a 500w psu that died on me a few years ago. still had it around so i decided to open it up and see if it could be fixed. 1 bulged cap, others way high on esr. and i've never seen this before; an exploded transistor.
http://pontoppidan.info/lars/repairs...ed_2sc1318.jpg

like the one in the picture, but top half. if i can find a replacement transistor and caps think it will start working again? and have you ever had an exploded transistor?

[stj]

Re: Exploded transistor

foto's of the board?
someone may have one to get the part number.

[eccerr0r]

Re: Exploded transistor

Usually transistors don't die/explode unless something else caused it to explode. Quite possibly the bulged caps caused its demise. My guess is that this transistor was part of the 5VSB circuitry, but yes having a picture of the board would be helpful to debug.

Of the overstressed transistors I've had, most just burned a hole through the case but with the heat involved, cracking like that isn't out of the realm of possibility at all.

[SteveNielsen]

Re: Exploded transistor

I've seen a couple of them go like that. I've seen an IC with a chunk blown out too. It's kind of rare.

[jasonbay13]

Re: Exploded transistor

well, it gets stranger. i had 2 psus that did nothing at all when plugged in. one had a blown transistor and another (one of the bigger ones) which i couldnt get number from.

the other has only 1 exploded transistor which i'm gonna replace, along with the bad caps and hope it works.

its a c945, any cross-reference on this i might be able to salvage from my other boards?

[SteveNielsen]

Re: Exploded transistor

General purpose NPN. Here's the basic data sheet for it.

I'd check the other parts around/connected to it.

[stj]

Re: Exploded transistor

2sc945

[TechGeek]

Re: Exploded transistor

Did that thing get 400VDC to it? Something rated for 5.2V or 6.3V gets 400VDC to it, it's probably going to explode, like that transistor.

[SteveNielsen]

Re: Exploded transistor

2sc945 is not the same spec as c945.

[tom66]

Re: Exploded transistor

2sc945 is not the same spec as c945.

Generally when a transistor part is <letter> + <3 digits> it is a Japanese-style part number beginning with "2S".

[SteveNielsen]

Re: Exploded transistor

That may well be true, however in this case it isn't. Look at the data sheets for the two. They are not the same transistor. That doesn't mean a 2sc945 can't be substituted for a c945 in some circuits, but the 2sc has a significantly lower power dissipation limit and a greater current gain. 2sc is also rated for 50v, the c is 60v.

[stj]

Re: Exploded transistor

it's also one of the most common japanese transistors on the planet.

[ben7]

Re: Exploded transistor

That may well be true, however in this case it isn't. Look at the data sheets for the two. They are not the same transistor. That doesn't mean a 2sc945 can't be substituted for a c945 in some circuits, but the 2sc has a significantly lower power dissipation limit and a greater current gain. 2sc is also rated for 50v, the c is 60v.

According to a datasheet I found, the C945 has a breakdown of 50V. The 60V rating is the collector to base voltage rating.

Also, they have different Hfe classifications. (O, Y, GR, BL)
I did find that the C945 is rated for 400mW, and the 2SC945 is 250mW.
The ratings might differ from manufacturer to manufacturer. Either way, 400mW seems a little much for a heatsinkless TO-92 package!

[stj]

Re: Exploded transistor

you can do over 1w in that type of package.
take a look at ZTX753!!

[ben7]

Re: Exploded transistor

you can do over 1w in that type of package.
take a look at ZTX753!!

It's rated for 200C vs the 2SC945's rating of 125C. Also, once it goes above 130C, the maximum allowable power dissipation decreases.

I'd never run parts over about 100C, unless it is a resistor. The hotter parts run, the less reliable they will be. (Though most people don't seem to care about that...)
(Also, reminds me of the minimum load resistors in PSUs, right up against cheap capacitors ... the extra heat helps kill the caps faster!!!)

That's as far as I will argue.
-Ben

[budm]

Re: Exploded transistor

It looks like the Chinese specmanship.
This C945 by STANTON is rated at 250mW.
So if you look around it will come up with lots of numbers.
I would trust the Japanese spec of the (2S)C945.

[stj]

Re: Exploded transistor

the blown one has a logo that looks like mitsubishi

[tom66]

Re: Exploded transistor

you can do over 1w in that type of package.
take a look at ZTX753!!

1W in a TO-92? Good luck on that. The standard thermal resistance for that package is about 150C/W. Sure, if your ambient is 0C you might be able to do it. But for all other reasonable applications, stay around 0.5W maximum.

There are extended-height TO-92 packages that may also have additional copper inside of the package; this may increase the power dissipation rating. But a standard TO-92 cannot dissipate a watt. The ZTX75x specify a rating of 175C/W, so even that cannot dissipate a watt unless you are at 25C and can tolerate a 200C junction... that's extremely optimistic... If you heatsink it sure - but heatsinking a TO-92 is silly, you should use a TO-220 or a surface mount package (D2PAK, etc) if you want a heatsunk device.

[canadaboy25]

Re: Exploded transistor

Take apart the base of one of those curly CFL light bulbs. Those usually have a bad cap or an exploded transistor in them.

[ben7]

Re: Exploded transistor

Take apart the base of one of those curly CFL light bulbs. Those usually have a bad cap or an exploded transistor in them.

+1

I've also seen popped/overheated film capacitors and resistors that have holes in them (burnt out).

One had a transistor that had popped in half :P

Also, I hate fusible resistors, I think they are a fire hazard when overloaded, but not severely overloaded. They are flameproof, but it doesn't stop nearby items from catching fire. Even the plastic cfl lamp base WILL burn if given a heat source. It just won't (well, shouldn't) continue burning on it's own.