repair pc power supply

[pharaon]

i hope someone can help me fixing this power supply,
when i try to turn it on the fan just move less than second then stop
when i measure the voltage on the fan and try to start the power it show 15 V then fade to 0 right away.

i tested some smd component , found short on the red circle capacitor , i moved it with hot air but the short remain on board and the capacitor where not shorted out of the board so i put it back on>

what else should i examine ?

 

[Bluejets]

Cannot see the output plugs but it looks like an older one given the power output.........
So......may be nothing wrong with it, it might just be waiting for the PG signal (power good)
That usually entails bridging the green in the plug to black common.
Yours might be different though.

 

[pharaon]

That usually entails bridging the green in the plug to black common.
Yours might be different though

that's what i mentioned earlier, when i turn it on by bridging the green in the plug to black common

when i try to turn it on the fan just move less than second then stop

 

[Bluejets]

Well, to be fair , you didn't say that.
Look for any dodgy or leaking electrolytics.

 

[pharaon]

Well, to be fair , you didn't say that.
Look for any dodgy or leaking electrolytics.

well i'm trying so far, all capacitors looking fine to me just by look didn't test any yet.

i also found this site for such power supply schematic , not sure which could be similar to mine
http://danyk.cz/s_atx_en.html

 

[pharaon]

found this diode on high voltage section of the board with number AT02P-200 which is open circuit not shorted nor drop voltage
is that normal? couldn't find datasheet for it

 

[Bluejets]

Have no idea which diode you are referring to as the photo is not specific.

However, normally the characteristic of the supply shutting down would be some kind of short.
If testing a diode, lift at least one leg to test otherwise you will get inaccurate results.

How did you test the diode...??(exactly)

Edit;....ah, see now ..D4.
Possibly in the start up circuit given the size of the adjacent small transformer.

When you take photos like this with the top in one direction and the bottom view swivelled 90 degrees around, makes getting a lock on things difficult.

 

[pharaon]

How did you test the diode...??(exactly)

Put multimeter on diode mode, negative prope on negative leg of the diode and positive on positive leg of the diode and it's open circuit ( show 0L on multimeter)
Then reverse prope and still same results,
Although on board where the diode connecting it's continuity ( shorted)



Took off the board two resistors and capacitor in red circle and tested they are fine.

Although the tow pins pointed at with blue arrows are shorted, guess it's the secondary coil design. not sure


also took off the 6 diodes and mosfet in the next picture and tested they are fine

[url=https://ibb.co/SJnDwJ4]
[/URL]

 

[Bluejets]

Diode legs are usually referred to as anode and cathode, that way avoids any confusion.

Photos have the flash turned on by the look of it, difficult to make out any chip markings.

Underside view with arrows and red circles, something looks to have blown up there.
Also above right hand side same photo.

Probably not necessary to lift resistors, especially surface mount as you can easily damage them.
Better to trace back and see if any other circuit section has any influence on the readings.

Cannot see reference point to where you are now as compared to the original diode.
Can see a couple of optos there in the next photo where you have the heatsinks red rectangle, this would be on the other side of the isolation section it appears, from where the diode was removed.

You do realise you'd probably pick up a working 500w supply second hand for $5.?
Many tutorials out there on pc supply repairs which may help as getting an actual schematic would be well nigh impossible.
Google " Coolmaster 500 pc repair" on Youtube.
Here is one on a 400 model for example.......

Note you can watch it on youtube even though it says unavailable here.

 

[pharaon]

You do realise you'd probably pick up a working 500w supply second hand for $5.?

well i'd like to fix it for learning purpose


when i connect the PSU to AC source i got 5V on PS_on pin
now when i connect PS_ON pin to ground i got 5V on 5VSB pin only and non of the rest pins.

when i connect the multi-meter to the fan wires and try to start the PSU by shorting green and ground pins i get reads 0.9 V that fade to 0.0 V right away.

none of the both heat-sink on high voltage or low voltage sections get warm.

i've also tested the low side output voltage of the primary transformer and there's no present voltage, outsource i tested while the green wire is shorted to ground in the connector pin.on the next picture i'm showing how i tested it with multimeter props



is that some sort of IC failure? or is it something else

 

[Bluejets]

One can also learn by reading as we used to in the old days, it's called study.
Pick a subject and find as much info on it as you can and go from there.

The 5VSB I imagine would be the standby voltage generated by one section of the supply, usually low current.
The main supply needs it to "start".
There are many sections to the unit but by following a set routine you should be able to narrow it down at least.
Start by learning how, in theory the unit works.
You'll learn much more that way than by poking around willy nilly hoping to fall upon a fault.

For example,........... the mains input is fused and feeds via a filter into the main rectifier.
High voltage capacitors smooth the input which is then switched usually by a dedicated (usually small) ic to the input of a transformer at a much higher frequency than the mains.(so smaller trannys can be used)
One can see the definite isolation line between mains input section and the transformer output on the underside of the pc board.
A small amount of output is then fed back to the switcher via opto isolators to regulate the output.
Different tapping on the transformer give different voltages as well as linear regulators for dedicated sections.
Your "fan kicking" and then turning off, point to the unit shutting down due to overload somewhere in the form of a short or excess load.

Google "pc power supply schematics", you'll get plenty. Then watch some youtube videos on the subject.

 

[pharaon]

Then watch some youtube videos on the subjec

I've been doing that and got pretty much the information you've just explain,
Although an input DC voltage on the primary transformer and no output voltage on low voltage side is that normal or cause of the IC switching failure.

Your "fan kicking" and then turning off, point to the unit shutting down due to overload somewhere in the form of a short or excess load.

Well I'm trying to find out the shorted component.

 

[Bluejets]

Although an input DC voltage on the primary transformer and no output voltage on low voltage side is that normal or cause of the IC switching failure.

Technically it's a high frequency ( maybe 400hz) switched dc on the primary input to the transformer.
Can be the chip as they get quite a hard life, but as in many things, not always.
Sometimes it will be the associated switching mosfet (if used) or failure of one of the feedback items, rarely the opto though.

 

[pharaon]

or failure of one of the feedback items

Such as what?
I've tested the optocoupler while green lime is shorted to the ground as a start signal for psua t low voltage side, one gives 0 V, the other one give 0.99V, the middle one kept fluctuaute at 1 and 0.6 too fast. I've tested by multimeter by putting two propes on the octocupler two legs at low voltage side.
Any suggestions

 

[mikey5791]

Have you tested for shorts on those 3 leg mosfets you took off from the two separate heatsinks?

 

[pharaon]

Have you tested for shorts on those 3 leg mosfets you took off from the two separate heatsinks?

i done fixing it long time ago thanks

 

[Martaine2005]

i done fixing it long time ago thanks

What was the problem?
The forum would like to know the fix!
It may help others in the future.

Martin