[Moved] Troubleshooting PC ATX SMPS

Fixing a PC ATX SMPS

Hi all,

I have a 400W ATX smps lying around which I'm trying to repair to run an old motherboard. Got it from my cousin with the fuse blown off. On checking, I found that both 13007 switching transistors had blown out of which one was shorting the circuit. The bridge, capacitors and rest were fine, so replaced both 13007's along with the fuse and tested all caps on the whole board and all components on the primary side. Connected a 35w automotive bulb as load onto the +5v and on switching it on, the supply was fine and gives steady +5 on the VSB line. But when I short the PS_on to GND to switch the supply on, the smps starts making a arcing like sound as if something had a loose connection. It was clearly from the transformer, so disconnected and de-soldered it out and checked, but it seems to be fine. Tried resoldering everything and powered it up again but the same arcing sound. The smps fan tries to start, but moves a bit and stops. Checked the voltage on the lines, everything was low at around 2v on the 5v lines.. After 10 secs, the NTC blew up and the supply died again. The opto seems fine when tested out of the circuit.

I am sure of a shorted semiconductor somewhere on the secondary but could anyone point me where to start? The circuit has SB1040, STPS201 rectifiers and something called STPR10 on the secondary. The PWM is a TL494 and controller is a PJ339 but I have no Idea how to check these. No other diodes or resistors seems to be blown visually. I hope the 35W bulb wont be a problem.

Thanks a lot!

Therre is a specific category for repai tips. Moved to Service Manuals, Requests, Repair Tips [alexan_e]

Click to expand...

text is in french but you can feed it to google translator.
hope this helps...

That looks nice, but pdf is secured and so unable to copy paste it to translator Thanks anyway!

I checked the smps again, all zenners are fine, all rectifiers are good, primary side is good, and all electrolytic caps and diodes are fine.. tested fusible resistors and all seems good, now what remains is the ordinary resistors and ceramic caps which seems crazy to solder and pull out one leg out of the circuit for all to check them.. somebody please help me! :-|

oups.. sorry - here is the unlocked pdf.

That was a nice article, but its more focused on the primary side and I had fixed my primary side of the supply. The problem now is a short somewhere on the secondary, but wait let me check the 494 as stated in the articles by checking for +5v on pin 14.. thanks!

a cheep-cheep-cheep sound coming from the smps often leads to a cap with a bad esr on the secondary side...
did you replace the caps and/or do you have an esr meter to check them ?

other valuable source for repairing smps :
Notes on the Troubleshooting and Repair of Small Switchmode Power Supplies

to build or buy an esr meter :
**broken link removed**

Now thats a good point! everyone says that sound is due to to caps on primary, but ur point makes more sense since I have checked the primary. For the caps on the secondary, I removed each cap out and charged them with a 9v battery, waited for 10 secs and connected to a dvm and it discharges slowly from ~9v over some time, so I think they should be ok. I dont have an ESR meter, but might build one in the future if it works with dvm as well. Can I replace these caps with good quality ordinary caps, as I have no idea where to find these 'low ESR' types here.. I should check the caps again..

you won't test a cap with bad esr with this method ...

basically an esr meter is like an ohm meter, but the esr meter injects an ac voltage at 100khz while an ohmmeter injects a dc voltage...
look at my esr repository above link. you will have plenty of easy to build esr meters (and also more difficult ones...)
there are esr meters that can display the results on a dvm like you want ...

you will also be able to test the primary cap that often fails too due to a bad esr ...
if you dont want to build an easy esr meter then you have to change all the chemical caps

I will be building one, but not now, as it will be usefull only for this one at the moment, so I'll try replacing the caps, but some caps have odd ratings like 1500uF, can I replace them with 2200uF's as thats the closest higher I can get?

yes you can replace with 2200uF, the higher the better, same voltage rating of better
but if your cap has a bad esr, it will heat and soon break...
I use my esr meter almost everytime I have something with caps to fix ... at least one a week !
I think it is something almost more useful than a dmm as todays all electronics that fails there are caps, and bad caps are 80% responsible for the failure ...

Hmm.. seems good.. will try for one soon.. thanks!! also another help.. I have another smps which stopped working after running without a load, what could be the prob? all it gives is a 5v vsb.

Hi,
If you tell me the supply model number, may be I can help you on component level. Some times the electrolyte capacitors seem to b ok on the low energy circuit but when fixed in the proper circuit are shown themselves leaky (my experience) and in your case it seem strongly some capacitor is leaky (not short circuited) which is causing the supply overloading.

Its a Frontech ATX 400W SMPS with model no JIL-2411.. Im replacing the caps today with higher valued ordinary ones as the caps on board have odd values like 1500uF 10V, so will be replacing with 2200uF 16/25V and low esr caps arent available here, so hope the good quality ordinary ones will work.. or will it be a problem?

Hello,

Could somebody please help me, as this thing is getting on my nerves. I replaced all electrolytic capacitors on board, replaced the TL494 and LM339 IC's, replaced 13007 switching transistors, tested each and every diode and rectifier and zenner on it, tested all fusible resistors and ceramics on the primary, and now when I connect to a motherboard + processor + hard disk, and power on, the smps starts making the arcing sound again and the processor fan tries to spin, but moves a few mm and stops. The 5v vsb is fine, but nothing on all other lines. I wasted a lot of time and money on this thing, could somebody please help me?

may be the overvoltage part of the supply that shuts down the atx ?
200W ATX PC POWER SUPPLY
regards,

Hi friends,
now kripton arrived at correct view, there is some overvoltge/over current feedback making the smps turn on and off or making wide variation in switching duty cycle which will result in audible noise from tranformer core. Normally two voltage feedbacks (12V & 5V) summed and single reference votage is take as feedback. you can identify them by their blue colour stands for metal film resistors. regarding feedback pin details i will reply later

krypton and murali, thanks a lot for that help.. And yes, as you said, the supply is cycling.. Since i had replaced all ic's, i will have to check the zenner's caps and resistors.. There is only one metal film blue resistor, it seems ok, let me check the rest.. Any help regarding any details wud be a lot helpful! Thanks a lot!

I have fixed a lot of smps, but never an atx smps as I work on mac os systems...
I fix 80-90% of the dead smps I try to fix
sometimes it's good to give up and buy a new one, after too much time on it...
btw I've never seen a dead ic in a smps... always a cap, a resistor, or a semiconductor like diode or mosfet...
good luck in your fixing !

I normally design and make electronic projects but im into microcontrollers more.. I can fix almost any prob in a low voltage dc circuit, but inductors and high voltage transformers are totally new to me, and since this was my first smps repair attempt, i was totally lost as i didnt have an idea of its working.. Now i have a better view.. Thanks to you all.. Havent checked the board since, will check and let you know.. Thanks again its working.. Now i have a better view.. Thanks to you all.. Havent checked the board since, will check and let you know.. Thanks again