I have a system I'm working on that has no standard 4 pin molex power connectors, but it does have a PCI-E style power connector available. Since it is a nonstandard PCI-E connector it has 3 ground lines, 1 5v and 1 12v, nothing else. I figured this would be easy enough to tap in to so I ordered up a female connector and was going to create a standard 4 pin molex converter for it. Before I got too far in to it though I did a voltage check just to make sure. One of the 3 ground pins will give me the correct readings for both 12v and 5v, meaning if I hook up the tester to that particular ground pin and test the 5v and 12v separately I get the correct voltages. However, if I use the other ground points I get 1.72v for the 5v and 8.88v for the 12v. So now I'm a bit confused. As far as I know I can't tie both the 12v and 5v in to the same ground correct? I know they all tie back to a common ground anyways as I've taken apart PSU's before and noticed this. There must be a reason why they separate them out in the lines, otherwise why bother running two ground lines?*EDIT* Ok so I ran across some more info that states SSD's do not draw from the 12v line, only mechanical drives do. This is good news as it sounds like I can use the 5v line and use my pigtail converter to power two SSD's with it. I'm still curious if there is a way to tie in to the 12 as well though, it would be nice to have it available if I ever need 12v for something. Thanks!
Sounds like the other two 'grounds' are in fact +3.3v. You should be able to use the same 0v ground for both +12v and the +5v provided the cable is thick enough to carry the current of attached devices. As you said its all a common connection anyway. Besides it sounds like the connector already uses one ground for 3 distinct positive voltages across 4 pins and that was deemed ok by whatever electronics engineer designed the solution.
Wow, I never gave that any thought but it totally makes sense. When I was getting the readings I thought something was strange with the numbers; there seemed to be a correlation between the deltas. I've had food poisoning for the last few days so my mind isn't very clear! I will be in front of the box again tomorrow so I will test that theory about the 3.3v lines. So we've got one vote for not being able to combine them and one vote for being able to combine them. I think you might be right Tealc, it seems whoever designed this did so knowing that one ground would be OK. I think this connector was designed to be plugged in to a proprietary backplane to power either 2 3.5 or 2 2.5 hard drives according to some obscure references in the documentation. Unless there was some voodoo taking place on the backplane that is above my very basic understanding of electricity, it sounds like this will work as expected. *EDIT* I tested out the 3.3v theory early, they are actually ground lines and not 3.3v. Pretty odd, I wonder why the voltage differences. Any ideas? It doesn't really matter of course but I'm still interested on what is going on with them.
It's too convenient that 12v less 3.3v happens to be 8.7v and 5v less 3.3v happens to be 1.7v. Either you've measured wrong or those two cables have a potential, with reference to the other ground line of +3.3v. They could form some sort of virtual ground I suppose but the voltages it's then going to supply are a bit useless for anything inside a PC. How did you ascertain they are ground lines? From the circuit or something?
At first I simply saw the cables were black and assumed that they were ground For testing though I hooked up the volt meter just like I did for the 12v and 5v, connecting to the known ground line I got a reading of 0.00 for both of the suspected 3.3 lines.
