ATX power supply -> lab bench power supply

[colin55]

That means you can cut them and separate them. They don't have to be connected to earth.

 

[mneary]

I have a nice little "Power Tronic WK-6145DL3" power supply and all 14 black wires come from the same huge blob of solder trace on the board.

To separate these black wires you would have to reverse engineer the secondary circuits and build a new foil pattern.

 

[colin55]

Simply cut them and measure each with a multimeter set to low ohms.
Pair up each winding and turn the transformer on. You will then be able to measure the voltage of each winding.

 

[mneary]

Are we discussing the same ATX computer power supplies? All of the ones that I've disassembled have all of their black wires soldered together and they cannot be separated. The various voltages are not different "windings" per se they are rectified and filtered, often using shared components. Cutting wires or connectors outside of the power supply does not separate them.

 

[Cobalt60]

But, can I run a fan at 7V by connecting the fans pos to 12V, and the fans neg to 5V, with no adverse effects on the rest of the system?

 

[mneary]

You need to have a load on the 5V at least equal to the fan's current consumption. This is in addition to whatever the minimum load that the power supply otherwise requires (Typically the 5V needs a load of a few amps for proper overall functioning of the supply).

So, if your fan uses 100 mA, and your 5V needs a minimum of 2A, then you need a load of 2.1A on 5V to run the power supply with the fan.

 

[Willbe]

could I simply connect the pos of the fan to 12V, and the neg of the fan to 5V, without any adverse effects, to run the fan at 7V?

Not to beat a semi-comatose horse, I don't believe the 5v supply will "start" properly if it's asked to sink current. Its internal diodes will be biased the wrong way.
At the same time I don't believe trying it would harm the 5v supply (maybe with a very light load).

 

[Shax]

It's obvious that the OP hasn't a clue about ATX or any other switch mode power supplies. As soon as you open one up, you expose lethal voltages and currents all over the pcb. They don't use regular transformers and linear regulators. They DO kill the irrsponsible or inexperienced electro-geek.

Do us all a favour, and BUY a decent bench supply.

 

[Cobalt60]

Im bumping up this post for a couple reasons. My main questions were all answered, but I have some more more questions for the final refinement of the project. First, Id like to respond to Shax.

For a living I fix digital cameras. Back when I was less experienced, I got shocked by the flash capacitors somewhat regularly. A typical flash cap is rated around 300V and 300uF. I even got shocked by 2 once. Anyway, to put things into perspective, how do the caps in ATX power supplies compare?

So to try to put some of your worries to rest, Id like to say a few things. First is, I know how to discharge capacitors using a discharger. Second is, I did not specify that I have to be working inside the casing. Although that was the original plan, I now am considering having all of the re-wiring done externally. Third, I am under the impression that turning on the power supply (shorting green to GND) when its not plugged in discharges at least most of whats in the capacitors. And lastly, all of the power supplies I have have been sitting around for quite a while. BUT, if you want to share any further safety procedures they would be very welcome.

OK now onto the further questions I have for the refinements of this project. First, I would like to know a source for some digital meters for this. Most important are that I can get a matching set for various voltages and amps, and a relatively cheap price. Also important would be ease of including these in the project (ive seen some that seem to have way more pins than I need).

Mainly the other thing I would appreciate some help on choosing is some sort of enclosure. It would have to be larger than the power supply, and relatively easy to work with. So it would ideally not be metal, or at least the front and back should be something easier to work with.

Thanks for any suggestions.

 

[mneary]

Probably the biggest "new" danger of the ATX power supplies is the two-hand danger. The flash cap of a digital camera would be small enough that you were unlikely to have a discharge across your body. With the the bigger the circuitry in the ATX supply you can get two hands in there. Through a few fingers, it hurts, but across your body it might stop your heart. The one-hand rule becomes the one-hand law!

 

[Spuzzum]

Ok, I know this post is old, but...

I've already read how to modify a psu into a lab power supply, and plan on using for a high powered led panel. I can use the +12v lines at 16A, and the +5v lines at 32A. I've already done the 10W 10ohm power restistor mod, and an spdt connected to the sense wires. It works .

Question I have though... read-me's say that the +5v and +12v make 7v, but to be careful if using for a "heavy load". .. What's considered a heavy load? Anything above the 16A available on the +12v line? How would you calculate what's available at 7v? Running these lights off of 7v would be more efficient than at 5v or 12v.


Thanks in advance