Completed Project Output voltage from 180W ATX PSU on +12V rail only 9.6V to 9.7V?

Thread Starter

Rahulk70

Joined Dec 16, 2016
536
Hello everyone,

I recently got three ATX PSUs which I plan to use as bench supplies. I've successfully managed to get one of them (250W) converted and running as bench supply. But the other two a 150W and 180W. The former only has a click when I power it up and has no output voltage nor does the fan spin.

Now the third 180W PSU is the one my question is related to. I shorted the GREEN and BLACK wires and the PSU fan is spinning and the out put on +5V and 3.3V rails are 4.9V and 3.4V respectively. But on the +12V rail the output is only 9.7V and when I connect a load (CPU fan) on the +12V rail it did not run and the PSU shut after like 10 sec. I even tried connecting it to my UPS and it is when I noticed that the "SITE FAULT WIRING" LED came ON. Anyone has any idea what could be the issue here.I know there is high voltage and precaution needs to be taken so if its a small issue that I can rectify I'm planning to rectify it by myself. Thanks in advance guys.
 

Dodgydave

Joined Jun 22, 2012
11,285
Have you put a load on the 5 or 12V rails, like a 10 ohms resistor 5W or an indicator bulb, as some psus will shutdown with no load?
 

Thread Starter

Rahulk70

Joined Dec 16, 2016
536
Have you put a load on the 5 or 12V rails, like a 10 ohms resistor 5W or an indicator bulb, as some psus will shutdown with no load?
Yes that is the very first thing I did coz some old PSU tend to blow up without dummy loads and since this is a 2004 PSU I didn't take a chance. But with dummy load it shuts down after 3-5 secs. Without any dummy load the PSU fan runs.
 

Dodgydave

Joined Jun 22, 2012
11,285
Then its going into overcurrent shutdown, it can be bypassed, or altered, you need to know which chip it uses for the pwm,

but if you don't have the skills to repair atx or switchmode psus, it's better to bin it.
 

Thread Starter

Rahulk70

Joined Dec 16, 2016
536
Then its going into overcurrent shutdown, it can be bypassed, or altered, you need to know which chip it uses for the pwm,

but if you don't have the skills to repair atx or switchmode psus, it's better to bin it.
I'll try to find the PWM IC and see how its functioning. Yes I've preformed some high voltage repairs. So, I'll give it a shot if its a small issue. What are some possible components of failures? I was thinking of checking the caps first.
 

Dodgydave

Joined Jun 22, 2012
11,285
Can't see it being a capacitor, it most likely is a resistor or transformer, output diodes, atx psus monitor the current in the pulse transformer, its converted into a voltage and its fed back to one of the comparator pins on the pwm chip, take the back off and post pictures of all the chips.

Just out of interest, does it do the same on the other rails, 3,5v??
 

Thread Starter

Rahulk70

Joined Dec 16, 2016
536
Can't see it being a capacitor, it most likely is a resistor or transformer, output diodes, atx psus monitor the current in the pulse transformer, its converted into a voltage and its fed back to one of the comparator pins on the pwm chip, take the back off and post pictures of all the chips.

Just out of interest, does it do the same on the other rails, 3,5v??
Hi,
Nope the voltage reading from the 5V and 3V rails are reading correct. I opened the PSU yesterday. I didn't see any visible damage like a blown component or anything but when I tried to do continuity test on the output side Transistor/Mosfet it was like dead shorts aka beep. Looks like its damaged and anyway I removed the board and kept it to use some components from it for future projects as a backup for parts maybe. The other PSU that didn't power ON, I opened up that too and in that it was clearly visible "a diode, a resistor, cap and transistor where all black sooty and blown up". So it is also waiting to be salavaged for parts.:)

So, I guess its kinda true that most people say. Only if the caps are bad, then by recapping the ATX can be restored else it ain't worth throw it in the bin.
 
Last edited by a moderator:

Dodgydave

Joined Jun 22, 2012
11,285
Hi,
Nope the voltage reading from the 5V and 3V rails are reading correct. I opened the PSU yesterday. I didn't see any visible damage like a blown component or anything but when I tried to do continuity test on the output side Transistor/Mosfet it was like dead shorts aka beep. Looks like its damaged and anyway I removed the board and kept it to use some components from it for future projects as a backup for parts maybe. The other PSU that didn't power ON, I opened up that too and in that it was clearly visible "a diode, a resistor, cap and transistor where all black sooty and blown up". So it is also waiting to be salavaged for parts.:)
If it's from the output of the transformer It will be a twin diode usually common Cathode, like a U30D40A series...

http://www.profusionplc.com/products/u30d40a
 

Thread Starter

Rahulk70

Joined Dec 16, 2016
536
If it's from the output of the transformer It will be a twin diode usually common Cathode, like a U30D40A series...

http://www.profusionplc.com/products/u30d40a
Yep. It looked something like that. I didn't carefully look at the product name. I'll check and let you know. So, if I change that part it will fix the +12V rail? I think I have a similar part from another 450W SMPS I had salvaged long back.
 

Dodgydave

Joined Jun 22, 2012
11,285
Easy to work out, the two outer legs will go the the transformer windings, chances are if its blown thats the reason it's going into overcurrent shutdown, .
 

Thread Starter

Rahulk70

Joined Dec 16, 2016
536
Easy to work out, the two outer legs will go the the transformer windings, chances are if its blown thats the reason it's going into overcurrent shutdown, .
I'll try swapping that part with another one and see how it goes. But if its something to do with the transformer, then its a dead end right?
 

Thread Starter

Rahulk70

Joined Dec 16, 2016
536
Yes, it's cheaper to bin it than repair, hopefully it could cure it.
Hi Dave,
Woohoo!! :) Got some good news. I fixed the PSU. Was just like u told. It was the high speed twin diode that screwed it up. Replaced it with a BYW51-100_to_BYW51-200 which I had salvaged from an old 450W PSU long back. Glad I didn't bin it and salvaged most parts. Its up and running now. Added a dummy load to +5V rail and connected my 12V peltier cooler to the +12V rail and curently under test run to see how it performs for half an hour. Cheers!!

My over excitement at seeing it work made me forget safety. I had powered the PSU from a spare UPS that I usually use for testing mains type circuits. I had run the PSU with the covers open and after unplugging forgot to discharge the huge caps(450V 200uF). Accidentally touched the leads under the board. Nasty shock and burned my finger slightly.:eek:
 

Thread Starter

Rahulk70

Joined Dec 16, 2016
536
Its been more than 20 mins that I'm running the PSU and at the 12V rail its current draw is around 2.8A. I've two questions:

1. When I opened the SMPS I saw a pot. The current output is 11.62V. But my other Dell 250W PSU bench supply has around 12.2V. Is it okay to vary the pot to get 12V exactly. I tried changing and was able to get 12v but reverted it back to the initial 11.62V. Thought I did ask first and then change. I've heard of PSUs blowing up.
2. The input 120V side of the PSU got really warm after 20 mins with 12V rail having 2.8A current draw.Is it normal that PSU to get warm? Also what is the max continious Amp that can be drawn from 12V rail?

Thanks
 

Dodgydave

Joined Jun 22, 2012
11,285
If its designed for variable output, then alter it to 12V it won't damage it. There should be a chart on the case of the atx, giving current limits,..
 
Top