What's new
Carbonite

South Africa's Top Online Tech Classifieds!
Register a free account today to become a member! (No Under 18's)
Home of C.U.D.

Fixing a PSU

Violater

Senior Member
Rating - 100%
32   0   0
Joined
Apr 22, 2010
Messages
433
Reaction score
8
Points
2,665
Age
38
My SO's PSU blew thanks to a faulty multiplug.
I've already replaced it, but I thought I might try fixing the old one as it was much quieter than the new one.
I've opened it up and the only thing that I can find wrong with it is is the blown voltage regulator.
To fix it would it just be a case of soldering in a new one? If so, how do I find out which replacement I need?
o3t9b.jpg
 
Last edited:
The component in the picture which appears to have shorted is a transistor.
Fixing it is relatively east provided you can get a similar one from the local electronics shops in your area. They are usually quite cheap.

Your biggest problem is going to be to get the right one, considering this one has melted and you might not be able to see the code on the component. :)

Be that as it may remove it and check if any of the code on the transistor is still visible... if not try to Google the Circuit Diagram for your particular PSU.
Then you will need to read the diagram to see if you can establish which transistor has blown.
From there you just need to get a transistor and solder the new one to the board.

A word of warning:
Transistors are used to smooth supply to IC's. That means if a Transistor has blow it could have resulted in a surge over the IC... that means the IC could be fried as well which is another story all together. (not that it would not be fixable)
But this is a worst case scenario and by no means a reason not to try and fix the PSU... just bear it in mind if the PSU does not operate as it did before.

Anyway drop me a PM if you need more advise and GLWT, always rewarding to fix something like this and have it work afterwards... Makes you feel like a real Tony Stark! :D

Cheers
D
 
Yes, but also bear in mind that the transistor blew for a reason. There may be more underlying issues that aren't apparently visible.
 
Agree with Damin, that there is a power transistor (one of many used to supply the large currents in modern PSUs). I'm willing to bet it would be the same model as the others in the picture that are all connected to the heatsink(s). Check the others and the PCB markings to confirm, although this isn't foolproof. If you can get your hands on the circuit diagram as suggested, that would be best.

However, chances are good that that's not the only thing that's wrong, and I'd think twice before testing it out on delicate/sensitive PC components... it can easily go wrong. Oh, and just another word of warning, just stay clear of the large caps in the PSU, they can hold charge for a long time and can give you a good zap if you're not careful. Anyhow, have fun :cool:
 
Aah, awesome, thanks
I've removed it, but as expected there aren't any legible marking left.
The PSU is a gigabyte ge-p470a-c2,
Searching now for a diagram.


The component in the picture which appears to have shorted is a transistor.
Fixing it is relatively east provided you can get a similar one from the local electronics shops in your area. They are usually quite cheap.

Your biggest problem is going to be to get the right one, considering this one has melted and you might not be able to see the code on the component. :)

Be that as it may remove it and check if any of the code on the transistor is still visible... if not try to Google the Circuit Diagram for your particular PSU.
Then you will need to read the diagram to see if you can establish which transistor has blown.
From there you just need to get a transistor and solder the new one to the board.

A word of warning:
Transistors are used to smooth supply to IC's. That means if a Transistor has blow it could have resulted in a surge over the IC... that means the IC could be fried as well which is another story all together. (not that it would not be fixable)
But this is a worst case scenario and by no means a reason not to try and fix the PSU... just bear it in mind if the PSU does not operate as it did before.

Anyway drop me a PM if you need more advise and GLWT, always rewarding to fix something like this and have it work afterwards... Makes you feel like a real Tony Stark! :D

Cheers
D
 
Aah, awesome, thanks
I've removed it, but as expected there aren't any legible marking left.
The PSU is a gigabyte ge-p470a-c2,
Searching now for a diagram.
is your psu a odin 470w? if so i might be able to check the no on the part you need
 
is your psu a odin 470w? if so i might be able to check the no on the part you need
Awesome, it looks like it's a unbranded odin that came with a case, but that's definitely the model no.

If you could check that would be great!
 
Yes, but also bear in mind that the transistor blew for a reason. There may be more underlying issues that aren't apparently visible.

This is true, but he did say in his original post that the PSU was damaged by a faulty multiplug...

I do however agree with you Oj0 there could be something else wrong with the PSU. That being said a blown power transistor is usually due to an overload from the mains.

But as P said above, I would not test the PSU on computer components. Rather get a multi meter and check the outputs before plugging it in to anything expensive. :)

OP drop me a pm if you need an explanation on how to test.
Cheers
D

Sent from my HTC One X using Tapatalk 2
 
Can't see the pic but from what the others say, it sounds like a power FET. In which case your PSU is probably fine, these things blow fast and can OC before anything else gets damaged.

I'd expect to see 4 of them arranged in a line on an aluminium heat sink with screws holding them flush? You could look at the part numbers of the rest and of these and see what they look like. From there it should be pretty easy to ascertain what the missing one should be.
 
This is true, but he did say in his original post that the PSU was damaged by a faulty multiplug...

I do however agree with you Oj0 there could be something else wrong with the PSU. That being said a blown power transistor is usually due to an overload from the mains.

But as P said above, I would not test the PSU on computer components. Rather get a multi meter and check the outputs before plugging it in to anything expensive. :)

OP drop me a pm if you need an explanation on how to test.
Cheers
D

Sent from my HTC One X using Tapatalk 2
Great - will do once I have a replacement part
send me a pm with your email and i will email you the pic i took, the text is as follows
b k28
FQPF
9N50C
Awesome, thank you very much!
 
Yo bud what your looking for is these Unfortunately it looks like RS no longer stocks them, but they should have a suitable replacement. If you go to the website during office hours and search around a bit a chat window pops up, let them find you a alternate (rs are in the kyalami office park)

EDIT:
like a muppet I didnt look further down the page.. this is an alternative
 
Last edited:
Just in case there isn't enough info there, from looking a goliath's posts, you want a 500V N-channel FET... Any electronics store (eg. Mantech) will have them for under R10.
 
Hey Guys - thanks for all the help, I'm hoping to go to A1radio over the weekend, they should stock this. I'll update once I've replaced it.

29f3276a-7c0b-b0ce.jpg

I have emailed it just dont know if it got to you
Sent from my GT-I9100 using Tapatalk 2
Hey bud, didn't get your email, but thanks for the pic - appreciate it!
 
From a quick comparison between the electrical characteristics of the FQPF9N50C and IRF840, I'd say yes.
 
From a quick comparison between the electrical characteristics of the FQPF9N50C and IRF840, I'd say yes.

While it *should* work, you'd normally want to meet or exceed the original with the replacement part's spec; whereas, in this case, the IRF840 is the underspec'd (8A) part. You want to at least meet the drain current and come in under the drain source resistance with the new bit. Then look that the threshold voltages (gate source) are roughly the same and make sure you exceed the power dissipation of the original. Wouldn't worry overly much about switching times and all that jazz...

Don't quite have time right now to confirm but have a look at an IRFP450 for a possible replacement...
 
While it *should* work, you'd normally want to meet or exceed the original with the replacement part's spec; whereas, in this case, the IRF840 is the underspec'd (8A) part. You want to at least meet the drain current and come in under the drain source resistance with the new bit. Then look that the threshold voltages (gate source) are roughly the same and make sure you exceed the power dissipation of the original. Wouldn't worry overly much about switching times and all that jazz...

Don't quite have time right now to confirm but have a look at an IRFP450 for a possible replacement...

Read here:

FQPF9N50C - FQPF9N50C datasheet(1/10 Pages) FAIRCHILD | 500V N-Channel MOSFET
IRFP450 - http://www.datasheetcatalog.org/datasheet/SGSThomsonMicroelectronics/mXrutyw.pdf
IRF840 - http://www.datasheetcatalog.org/datasheet/fairchild/IRF840.pdf

But I agree with Silvertongue... IRFP450 is better replacement than IRF840.
 
Last edited:
I got the required transistor and only then removed the PCB from its enclosure.

Unfortunately, the other side of the PCB had quite a few burnt resistors and other darkened bits.
I wasn't too confident after that so I just left it.

Thanks for the help though everyone
 

Users who are viewing this thread

Back
Top Bottom