If there is voltage there, the device is either going to draw current, or it is not going to. The power supply is fulfilling its end of the obligation if you are seeing voltage. Some USB devices - actually, I'd wager, a lot of USB devices, are not going to draw current blindly if you leave D+ and D- floating. It's some Chinese standard (I can't remember which one) whereby tying D+ and D- together will let the device know that it's at least attached to *a* charger, even if the device may not know *what* charger exactly. Try doing this.
Working or not, that power supply is really not going to be all that useful to you as a bench supply. Those power supplies date back to a day when onboard regulation was pretty much not a thing on motherboards, and the architecture actually dated back to the AT days, and ATX was kind of tacked on as a cheap addition to bring their old builds up to the modern standard. Why I believe this is the case, i that you have mega 3.3v and mega 5v but almost nothing on the 12v. 8A isn't *useless* - but for a notionally 300 watt power supply, it may as well be.
I'd recommend using a real ATX power supply that has a decent 12v output if you want to be pulling meaningful amounts of power for stuff. Granted, for anything but really power electronics, 12v 8A is plenty - before conversion, that is. The second you need to step that voltage up, useful power drops at a rapid rate of knots.
Hope this advice that no one asked for is of any help.
EDIT: I feel like an idiot for not seeing the resistor, for some reason I saw differently. I want to leave the above alone, in case someone else comes across this. Now the problem you may be having, is that the device is seeing some weird resistance across the output terminals and is going "wtf is wrong with my port, I am not charging from here". This can particularly be the case if, for example, a PS3 controller usually authenticates with some sort of controller upstream of the port itself, and then negotiates charge, or, the PS3 has some obtuse, bespoke current-setting resistor. Being Sony, I'd imagine that any combination of any of the above is true. I'm out at the moment so I can't be arsed checking, but try removing the resistor and just tying D+ and D- together. If nothing else, this should make most any other device that was built to work with the Chinese standard (which is a whole metric buggerton of things made this side of the publishing of the Magna Carta).
Hope this helps the even less folks who will find it handy!
EDIT 2/P.S.
Few things:
- You might want to reconsider using actually not worthless electronic devices as load resistors, while unlikely with ATX supply, you probably don't want to accidentally burn it.
- Please do put some kind of fuses on the lines. PC PSUs can put out tens of amps on the 5V/12V line and you probably do not want to weld something accidentally with it.
- There is no real standard on how much/little load you need to put on ATX supply to be stable so that's basically trial and error part, but I'd at least double whatever I found to be a minimum just in case
- Have you seen ? This + ATX power supply is easy way to get features useful for lab power supply, like the current limiting.
Throw that antiquated AT garbage out and buy one of these. Seriously. I have the higher power model of this (the fixed version of the one that Dave had drop its guts on him) and in addition to being a fun little project, it's a great switching supply. If you are going to use a switching supply at all for a bench supply, it might as well be a quarter-arsed decent one. You'll need to BYO power supply, er, power supply though. For 20v 3A or whatever the little self-contained module is, this isn't a huge issue. For 60v 20A... yeah I'm still trying to work that one out, myself. The smaller module requires almost no assembly work at all, or precisely zero assembly if you don't mind using it as just a bare module. For something that is so cheap, and gives you a safe, stable, and programmable, supply, you can't go wrong.