ATX psu +12v rail reading very low

[polashd]

I'm trying to convert my old ATX pc psu to bench power supply.

I put a dummy load on +5v rail.
the psu works.
+5v rail shows 7 - 8v
but +12v shows less then 1v (0.4 -0.5v) & the Fan also doesn't spin . I can't understand why?

can anyone help me please?

( the controller is KA7500B) (& nothing is converted yet)

 

[alec_t]

The PSU may be 'working' but, if the load is not excessive, it is faulty. Neither the 5V nor 12V rail is correct. Obviously the fan won't spin if its supply is only 1V (it would normally be powered via the 12V supply).

 

[MrAl]

I'm trying to convert my old ATX pc psu to bench power supply.

I put a dummy load on +5v rail.
the psu works.
+5v rail shows 7 - 8v
but +12v shows less then 1v (0.4 -0.5v) & the Fan also doesn't spin . I can't understand why?

can anyone help me please?

( the controller is KA7500B) (& nothing is converted yet)

Hello there,

Do you have pin 16 (the green wire) jumpered to one of the black wires? You need to jump pin 16 the green wire to say pin 17 a black wire to get the whole supply to turn on. Some power supplies will output +5v at maybe 3 amps even before the whole supply comes on, and that will be seen at pin 9 (violet).
If on the other hand you are seeing +5 out of one of the red wires then something might be wrong with the power supply.
Also, i have always used a dummy load on the +3.3v line (orange) not the +5v line (one of the reds) but that might also work too. 5 or 10 ohms should do it. The dummy load connects from an orange wire to a black wire.
This is for a ATX power supply with a standard 24 pin connector.

 

[polashd]

Do you have pin 16 (the green wire) jumpered to one of the black wires?

Yes, I shorted the green wire with one black wire & then only I get +7- +8v on 5v rail. and it runs a motor well.

If on the other hand you are seeing +5 out of one of the red wires then something might be wrong with the power supply.

what maybe the reason that +5v is working but +12v not.
can you suggest me how to look for the cause of the problem?

 

[polashd]

The PSU may be 'working' but, if the load is not excessive, it is faulty. Neither the 5V nor 12V rail is correct. Obviously the fan won't spin if its supply is only 1V (it would normally be powered via the 12V supply).

5v rail runs a motor well (with about 8v 800mA).

 

[ericgibbs]

hi,
Have you seen this link?

https://www.wikihow.com/Convert-a-Computer-ATX-Power-Supply-to-a-Lab-Power-Supply

E.

 

[alec_t]

5v rail runs a motor well (with about 8v 800mA).

I don't doubt that you can run a 12V fan motor from 8V: the point is that the 5V rail shouldn't be at 8V. You need to solve that problem before trying to build your bench supply.

 

[absf]

( the controller is KA7500B) (& nothing is converted yet)

I think the first thing to do is to get hold of the datasheet of KA7500B. From my database, KA7500 is an Korean equivalent of the TL494. Next google for an ATX schematic and start from the +12V supply outputs.

Is there a LM339 and TL431 chips inside your switcher? The working of ATX is quite complicated and can be tricky to repair. I normally use an isolation transformer when repairing this type of mains transformerless circuits.

Good luck with your troubleshooting.

Allen

 

[polashd]

Is there a LM339

Yes there is a LM339. what it does in the circuit

 

[absf]

I attach 2 ATX PSU for your reference. One with KA7500 but without the LM339. The other one using TL494 but with a LM339.

Allen

 

[MrAl]

Yes, I shorted the green wire with one black wire & then only I get +7- +8v on 5v rail. and it runs a motor well.




what maybe the reason that +5v is working but +12v not.
can you suggest me how to look for the cause of the problem?

Hello again,


One of the reasons power supplies like these fail is because the electrolytic caps dry out or for some other reason they fail. So one thing you could do is look for high ESR (effective series resistance) in all the electrolytic caps and see if any have very high resistance. I think there are testers for this, but what i did was i pulled all mine out of the circuit board and tested them with a frequency generator and scope. When the ESR goes high it is also probably that the capacitance goes down quite a bit, so you could test for that too.
Alternately if you dont have the test equipment, you can try replacing all the electrolytic caps, especially all the ones that handle significant power like the output caps and the two large converter caps. The output caps are the worst ones. It could be that changing them alone will fix the problem. This kind of problem came about even more when some cheap low quality caps crept into the market some time ago and ended up in some power supplies.

Because you are getting +7v or more out of the +5v line it sounds like the converter part of the circuit is working, but the regulation part is not. This could mean a lot of things like the 12v line is shorted by a bad component (like output cap). The LM339 is often used to detect overvoltage and undervoltage and some other things.

You should note here though that without having the power supply right in front of me to do tests on i can not be sure what is wrong with it. If you do own an oscilloscope however then we can do some tests if you like, but you have to be warned that there are very dangerous voltages inside the power supply case in certain places so there is always risk of a bad shock which could lead to death. There's always that risk so if you've never worked on anything like this it is better that you do not attempt it.

How many watts is the power supply rated for? Maybe you can get a new one for cheap on the web somewhere. Various places sell these things quite cheap these days if they are not too high in wattage. You might find one as low as 15 dollars (USD) somewhere for example.

Do you intend to convert this into a hobby power supply too?