I'm performing the basic modifications to a used computer power supply so it will power my iCharger 106B+. Based on my PC building experience, the PSU seems generally well-constructed and fits into what I consider the "solid" category for a unit in its class (250W, +12V /16A)--should be more than enough to do what I need it to do.

A quick question, however, as my last dose of freshman physics/electronics was a long time ago: for an ATX power supply like this, is there any benefit to terminating multiple +12V (yellow) leads to a single terminal? Would this stack the amperage? I'm not sure if the PSU's rating stamp means that the total amperage across all 12V rails is 16A or if 16A is for each individual +12V rail. I have the ultimate goal of cleaning this thing up so there are only two nicely bound terminals. If I don't need to tie the other +12V rails together, I'll desolder the extra wires.

Thanks in advance.

for an ATX power supply like this, is there any benefit to terminating multiple +12V (yellow) leads to a single terminal?

The benefit using multiple wires is that a single yellow (or black) wire can not handle the whole 16A. Each is about 22-20awg and could probably handle 5-8A each. As such I would use 3-5 yellow (same with black) per output port.

The ratings on the power supply are global. If it says it can handle 16A on the 12V rail, that is the total output. Here is my how-to if you want to read some more.
http://sites.google.com/site/tjinguy.../convert-pc-ps

Originally Posted by tjinguy

The benefit using multiple wires is that a single yellow (or black) wire can not handle the whole 16A. Each is about 22-20awg and could probably handle 5-8A each. As such I would use 3-5 yellow (same with black) per output port.

The ratings on the power supply are global. If it says it can handle 16A on the 12V rail, that is the total output. Here is my how-to if you want to read some more.
http://sites.google.com/site/tjinguy.../convert-pc-ps

Perfect. That's exactly what I wanted to know and is what I was thinking. You have a nice tutorial set up there. Thank you very much!

[Edit] These results aren't entirely accurate. Please read post #45 to see why. [End edit]

Here is some load testing that I did with a very cheap computer PSU. The point being that they are great for powering 50 watt chargers, but may be lacking when it comes to providing power for more powerful chargers. You usually can't complain about the price though.

To perform these tests I used a notebook computer running on battery to collect data from a CBA battery discharger that was used to load the PSU, in order to isolate it from shore power that was powering the PSU. I used test jumpers from the PSU to the CBA discharger that were probably smaller that the 12V power input leads to a 50 Watt LiPo battery charger. So, to be fair, I am listing the voltage drop. Measuring the voltage at the power supply with a digital multimeter I got 12.61V and the CBA discharger measured 12.68V both with no load. Then, under a 10A load, 10.97V at the PSU terminals and 10.44 at the CBA discharger. So, at 10A I got a .46V drop (taking into account .07V difference in the no load reading.)

Also, inside the power supply there are four black wires for negative going to the black banana jack on the converted PSU, and four yellow going to the red banana jack. So there shouldn't be any gread voltage drop on the internal wires at these current levels.

On the graph the black line represents a 1 ampere load for thirty seconds, then no load for thirty seconds. The red line 5 amps, and the green line 7 amps. Not quite the 18 amps listed on the label.

So, a converted computer power supply unit should make a decent power supply for a 50 Watt LiPo battery charger. But, don't expect to be able to use anything close to the power figures on the label.

Assuming no other funny business with an odd ball pc power supply, a properly converted one holds voltage great all the way past its rates output. I know this because I have been using them for years and have made many for friends. For example I have one at work for charging packs while I work. It is older Antec model and is rated for 18A on the 12V rail. Using a 20A static load the voltage drops from 12.41V unloaded to 11.97V loaded. I have charged many times with a calculated input of 21A and it never misses a beat.

The point is that any decent quality, properly converted pc PS should work great for any charger from a 50W unit to 400W unit. Just out of curiosity, what exactly were you using to load the 5V bus and did you use multiple wires to feed each banana jack?

I don't remember what resistor value that I used for the 5v rail, but I'll check. I used four of the yellow wires to the positive terminal and four black to the negative, so there shouldn't be too much resistance there.

It is in fact a very cheap power supply so I wouldn't be surprised that it regulates poorly. I find this much voltage drop typical, but I've only convertered very cheap power supplies like the one in the photos.

The good power supplies that I have such as PC power and cooling and Antec I never convert because I use them in computers. They are good power supplies.

I'll dig out the power supply in the photo, check the resistor and post back.

I too have converted some bargain basement power supplies (sub-$25 models) and I have seen more voltage variance in these, but usually adding more load on the 5V bus and firmed it right up. Actually the more I do this, the more I feel that 1A is insufficient for many models. I am now leaning on a 2-3 10W 1ohm resisters in series as a good choice. That will draw 1.67-2.50A and that seems to stabilize the 12V bus better, especially in cheaper models.

Looks like I used a 10Ω 10 watt resistor to load the 5v rail. Maybe I need a little more load than a half ampere on the 5v rail?

Or one can do some EBaying and for under $25 come up with a 12V 55 amp. Server PS which is very easy to rig up.

https://www.rcgroups.com/forums/show...1#post14004361

Originally Posted by u.rusty

Looks like I used a 10Ω 10 watt resistor to load the 5v rail. Maybe I need a little more load than a half ampere on the 5v rail?

Yes more load on the 5V rail will help. Try putting another 10ohm in parallel with that one, or even 2 more, and see how it changes things.

Originally Posted by tjinguy

Yes more load on the 5V rail will help. Try putting another 10ohm in parallel with that one, or even 2 more, and see how it changes things.

5Ω or less...I'll have to see what resistors I have laying around of adequate wattage and try it.

can the positive and negative run on the same wire my charger has a connector pin and looking at the wire there is only one ?

I am not sure what you are asking. Could you explain further?

well by using the yellow + and black - wires from the power supply my charger does not have a positive and negative input for 2 wires. it only allows for one single wire to go in to the charger to power the charger . so do i make one single wire with the 2 yellow and black wires ?

Your charger must have two wires to power it, positive and negative. Assuming it is powered by direct current. Is power connector for your charger a coaxial plug? Coaxial power plugs have one conductor in the center and another on the outside of the plug, like the Radio Shack power plug at the coaxial plug link.