Help with converting pc power supply.

Joined
May 13, 2013
Messages
93
Location
NJ
I have a newer 450 watt ATX with the 24 pin connector...it only has one voltage signal wire that the newer ones have, a brown one shared by another brown that i soldered together......I took apart one of my old chargers to get at the resistors, 6 10 watt 4ohms. I tried all sorts of different combinations of resistors on the 12v, 5v and 3.3v, it refuses to kick on. I'm not getting dc voltage anywhere..grounding green wire does nothing. Maybe it needs different resistors, i don't know. A lot of the conversions cover older ones it's hard to get info for newer ones.
 
All ATX PSU's operate the same way, otherwise they aren't ATX. All you should need to know can be found here.
https://sites.google.com/site/tjinguytech/my-projects/convert-pc-ps
 
An ATX power supply should kick on if you ground the green wire (normally pin 14, I believe).

If it doesn't, then something else is wrong with it.
 
Mine is atx .2.3, specifically a thermaltake tr2 450w not sure if they added more safety switches and cutoffs in the design. As it sits now the charger is plugged in to the 12v, the green wire is jumped, I'm getting 5v at the purple and 17v at the gray, .4v between red and black before the resistor that's grounded. The .4v stays when the power is off and nothing at the yellow, all the brown wires are tied together. I switch it on and nothing.

I just added a 10w 4 ohm resistors to yellow red and brown wires to ground with the charger plugged in and switched it on...nothing.
 
Obvious test: Plug inot a bare motherboard and momentarily jumper the power on header pin pair of the motherboard. If the PSU powers up, then its' probably ok. If it doesn't, it's bad and you can either fix it or get a different one, but not worry about it anymore. ;)


But ATX any version still uses the same power on method and pins for all of them. Some manufacturers don't use the same color wires there (I've seen all black on two "generic" ones over the years), but the pins themselves will still be the same ones.


I *have* seen quite a few PSUs that won't turn on without a sufficient resistive load on 12V, sometimes also on 3.3V and/or 5V. That's probably why many of the "PSU tester" units contain a largish power resistor on each of those lines, aside from the load test they sometimes claim to be able to do.
 
I just edited the post above to say that i had resistors on all the plugs. I'm not a computer guy so i don't have parts around. Do you feel the resistors have would be sufficient on all the loads?....I'm just trying to get it to do something that resembles any hope in getting it to work. Full AC power through when turned on. It was new untested and i cut all the wires before testing anything which was probably a bad idea.

I've been through lots of DOA ebike parts in the past few months i figured i'd take it into my own hands, so it's pretty frustrating to have this fail. Oh well at least i have some extra wires and a big cooling fan to play with.
 
I connected all the colors together on the 24 pin connector, will this achieve the amp's as rated on the case, or do I have to do the same with the Molex round-pin connectors?

I can measure voltages and they seem lower then they should be. I cant measure amps. The fan turns on with the PS-ON wire to ground. I think I need to use the 10ohm 10W power resistor from my bundle of 5V soldered together to my bundle of GND wire soldered together. Maybe then I can measure amperages.

I am missing a -5V wire

I got,
4 wires of +3.33V @ 16A
5 wires of +5.08V @ 19A
2 wires of +12V @ 11A
1 wire of -12V @0.15A

I just read online that you can go +12V and -12V for +24VV then +5V and -12V for 17V then +12V and 0 for +12V then +5V and -5V for 10V then +12V and +5V for 7V. How would I then calculate the amps?

If I get this working, I am going to buy a few Dell 12V 65A server PSU's.
 
They are not seperate power supplies, its just one PSU that they were joining different voltages. Some have installed a timer for variable voltages.

Connect PS-ON to gnd. If PSU needs a load, some say use 10ohm 10watt power resistor on 5V to gnd. 5VSB is always on even when the PSU is off but still plugged in for the wake-on feature of computers. PG I think measures 5V if everything is good.

Maybe I need to solder every single same color wire together to obtain the rated capacities.
 
markz said:
They are not seperate power supplies, its just one PSU that they were joining different voltages.

Yes; that's clear from your original post. ;) I stated it the way I did because each different voltage is effecively a separate supply, sharing a common ground.
 
markz said:
I just read online that you can go +12V and -12V for +24VV then +5V and -12V for 17V then +12V and 0 for +12V then +5V and -5V for 10V then +12V and +5V for 7V. How would I then calculate the amps?

I think what AW's saying is any combination using the -12V rail will limit you to 0.15A at the combined voltage.
e.g Combining the 3.3V and the 12V will be limited to 11A (from the 12V rail).
 
Back
Top