Power question

[TheStu]

I have a 360 power brick that I'm trying to adapt to power a GPU, and I've run into a snag, probably because I haven't worked with this sort of thing in about 10 years.

Anyway, the power brick is 203w and has 8 pins:

5v
Ground

12v
Ground
12v
Ground
12v
Ground

The 5v can be shorted to turn on the brick, so that leaves me with the 3 12v lines, each at 5.5A, yes?

If I wanted to get 2 6pin 12v connectors out of that, can I just splice and split the lines to add and divide the amperage as needed? The voltage should stay constant, right?

So if I splice all 3 lines, I should have 12v, 16.5A on the new, single line? Then if I split it into 2, I'd have 12v 8.25A? Or am I doing this completely wrong?

 

[uclabachelor]

I have a 360 power brick that I'm trying to adapt to power a GPU, and I've run into a snag, probably because I haven't worked with this sort of thing in about 10 years.

Anyway, the power brick is 203w and has 8 pins:

5v
Ground

12v
Ground
12v
Ground
12v
Ground

The 5v can be shorted to turn on the brick, so that leaves me with the 3 12v lines, each at 5.5A, yes?

If I wanted to get 2 6pin 12v connectors out of that, can I just splice and split the lines to add and divide the amperage as needed? The voltage should stay constant, right?

So if I splice all 3 lines, I should have 12v, 16.5A on the new, single line? Then if I split it into 2, I'd have 12v 8.25A? Or am I doing this completely wrong?

Yes, buy a proper PSU to power the GPU, unless you want to risk frying your GPU.

 

[TheStu]

Yes, buy a proper PSU to power the GPU, unless you want to risk frying your GPU.

Electrically, am I doing something wrong electrically? It's a side project, I'm just trying a thing.

 

[TuxDave]

Electrically, am I doing something wrong electrically? It's a side project, I'm just trying a thing.

I'm just throwing out ideas. The power lines are probably not perfectly symmetrical and so if you pull 16.5A on a 3x power line, I'm not sure if you can count on it being 5.5A distributed evenly.

 

[PottedMeat]

On my 175W 360 power supply there is only one 12V output that can supply 14A, the three 12V terminals have one wire each tied together at the supply.

You need to make sure the total load is less than what the 360 can supply.

To turn it on you can use a switch to connect the PS_ON wire to the +5V line.

 

[TheStu]

I'm just throwing out ideas. The power lines are probably not perfectly symmetrical and so if you pull 16.5A on a 3x power line, I'm not sure if you can count on it being 5.5A distributed evenly.

On my 175W 360 power supply there is only one 12V output that can supply 14A, the three 12V terminals have one wire each tied together at the supply.

You need to make sure the total load is less than what the 360 can supply.

To turn it on you can use a switch to connect the PS_ON wire to the +5V line.

Thanks for the replies. My multimeter ate the dust, and the new one hasn't arrived yet, so I can't double check the amperage, but after wiring it up today, it seems to be working.

I tied all 3 lines together, though I think PottedMeat is right, it's just one 12V rail off the power brick.

As for the load, according to the PCIe spec, the 6pin connector is only rated for 75w, and I'm only hooking up a gtx650 which is just 1x 6pin so i'm in the clear.

 

[imagoon]

I would be fairly worried about ground float that. 12V is 12V referenced to the ground of the supply. There is basically no reason why the ground from the 360 supply and the computer supply can't be 0V to ~170V apart from each other. Tying the the grounds together could possibly cause a fairly high current load in the ground connections. Also supply noise that could produce harmonics and the like could induce interesting currents.

 

[TheStu]

I would be fairly worried about ground float that. 12V is 12V referenced to the ground of the supply. There is basically no reason why the ground from the 360 supply and the computer supply can't be 0V to ~170V apart from each other. Tying the the grounds together could possibly cause a fairly high current load in the ground connections. Also supply noise that could produce harmonics and the like could induce interesting currents.

I think it's all 1 ground source in the brick, it was just split into 3 lines to be run down the cable to the connector (probably so that it can then be split inside the 360).

I don't fully understand the issues you raised, so i'm going to blithely ignore them until I do.

 

[Tuna-Fish]

I think it's all 1 ground source in the brick, it was just split into 3 lines to be run down the cable to the connector (probably so that it can then be split inside the 360).

I don't fully understand the issues you raised, so i'm going to blithely ignore them until I do.

The ground of the 3 lines from the power brick are the same. The problem is that the ground of the computer that the GPU is attached to is not necessarily the same as the ground the power brick produces. This can cause disastrous problems, up to and including fireworks. For example, signal lines that are only supposed to have a few volts and essentially no amps flowing through can end up burning as the difference between the grounds flows through them. This could be caused by high transient variation in one of the grounds -- say, when the dryer comes off.

The very least you need to take care of when using that system is that the power brick and the computer psu are plugged into the same mains phase and the same ground. Depending on how a house is wired, two adjancent power outlets could have very different grounds/power. For more robustness, you could tie the grounds of the two power supplies together with something heavy duty.

 

[imagoon]

I think it's all 1 ground source in the brick, it was just split into 3 lines to be run down the cable to the connector (probably so that it can then be split inside the 360).

I don't fully understand the issues you raised, so i'm going to blithely ignore them until I do.

No, the power supply in the case vs the 360 power supply. Attaching the (-) of the 2 supplies could induce a huge current possibly and the video card would eat that.

--edit--
Basically what Tuna-Fish said. Both the grounds and the (-) feeds could be different.

 

[Micrornd]

No, the power supply in the case vs the 360 power supply. Attaching the (-) of the 2 supplies could induce a huge current possibly and the video card would eat that.

--edit--
Basically what Tuna-Fish said. Both the grounds and the (-) feeds could be different.

I don't think either of you understand the power circuitry of a PCI-e video card. :'(

The 6 or 8 pin power connector grounds are already tied to the ground of the PSU through the ground pins of the PCI-e connector.

Ground potential difference isn't an issue here.

Auxiliary video card power supplies have been around for a long time.
example -
http://www.newegg.com/Product/Produc...82E16817104054

There's no real difference between what he wants to do and running dual or triple PSUs in the same machine

His only real issues are making sure it turns on with his PSU, wiring the connectors properly, and that it will really supply the power needed.
Turning on the brick is best accomplished with a relay

 

[imagoon]

I don't think either of you understand the power circuitry of a PCI-e video card. :'(

The 6 or 8 pin power connector grounds are already tied to the ground of the PSU through the ground pins of the PCI-e connector.

Ground potential difference isn't an issue here.

Auxiliary video card power supplies have been around for a long time.
example -
http://www.newegg.com/Product/Produc...82E16817104054

There's no real difference between what he wants to do and running dual or triple PSUs in the same machine

His only real issues are making sure it turns on with his PSU, wiring the connectors properly, and that it will really supply the power needed.
Turning on the brick is best accomplished with a relay

There is a significant difference. Grounds on the 6pin connector (-) is tied to PCI-e. The ATX standards define how grounds should be derived. The 360 power supply is not an ATX supply. IE there is no guarantee that grounds are at the state. Since ATX has that defined, you can do multiPSU (fairly) safely.

So yes the aux supplies exists. They how ever have already done all the "hard" work for you, like designing it so it doesn't fry your system.

 

[Micrornd]

There is a significant difference. Grounds on the 6pin connector (-) is tied to PCI-e. The ATX standards define how grounds should be derived. The 360 power supply is not an ATX supply. IE there is no guarantee that grounds are at the state. Since ATX has that defined, you can do multiPSU (fairly) safely.

Yes, the grounds on the PCI-e busses are tied to the 6 and 8 pin Aux video card plugs, and they are also tied to the common ground in the PSU and the MB grounds per ATX specs, they are not in any way floating.
That is exactly why when you plug in an auxiliary PSU or secondary PSU the ground potential is eliminated because they both become/are on 1 common ground.

They are both switching power supplies, so their voltages are regulated to the "ground" reference by design, whether that be a true "earth" ground or a "floating" ground (as is commonly done in creating a regulated 7v supply by using the 5v supply as the floating ground for the 12v supply, without disturbing the actual 5v or 12v supply referenced to the "earth" ground).

The only remaining voltage potential problem that might exist, is faulty regulation by one or the other PSU.
The only other problem could be "line noise", but since this is a PC power brick to start with, it is as adequately filtered as the ATX PSU.
Since neither of these are homemade units, but rather documented mass produced PSUs, those problems do not exist to any greater extent than that of any PSU would.

So yes the aux supplies exists. They how ever have already done all the "hard" work for you, like designing it so it doesn't fry your system.

If you truly believe this, please document it, so we can see these extra added precautions.
All the ones I have opened to date (6) were nothing more that "power bricks" with 6, 8 and sometimes 4pin connectors.

 

[imagoon]

Yes, the grounds on the PCI-e busses are tied to the 6 and 8 pin Aux video card plugs, and they are also tied to the common ground in the PSU and the MB grounds per ATX specs, they are not in any way floating.
That is exactly why when you plug in an auxiliary PSU or secondary PSU the ground potential is eliminated because they both become/are on 1 common ground.

They are both switching power supplies, so their voltages are regulated to the "ground" reference by design, whether that be a true "earth" ground or a "floating" ground (as is commonly done in creating a regulated 7v supply by using the 5v supply as the floating ground for the 12v supply, without disturbing the actual 5v or 12v supply referenced to the "earth" ground).

The only remaining voltage potential problem that might exist, is faulty regulation by one or the other PSU.
The only other problem could be "line noise", but since this is a PC power brick to start with, it is as adequately filtered as the ATX PSU.
Since neither of these are homemade units, but rather documented mass produced PSUs, those problems do not exist to any greater extent than that of any PSU would.



If you truly believe this, please document it, so we can see these extra added precautions.
All the ones I have opened to date (6) were nothing more that "power bricks" with 6, 8 and sometimes 4pin connectors.

You make quite a few assumptions there. (-) isn't always earth ground. In DC "ground" is commonly called (-) and it not the same thing as earth ground. It also can be incorrect as (+) ground systems also exist. There simply is no way guarantee the 360 supply uses the same potential as an ATX supply. It might but there is no guarantee.

#1 earth ground should have no potential. Earth ground can be (+) to one supply and (-) to another.

So as switching power supplies, voltages is the delta between the negative and positive leads. Two power supplies are not required to derive (-) at the same potential. Two switching supplies can derive 12 volts between their own negative and positive lead and there is no requirement for the delta between both negatives to be 0V. The exception to this is ATX which specification wise states that (-) should be 0V in relation to the earth ground. Since the 360 PSU is not ATX, it could develop (-) at delta 1000V from earth and 1012V from earth and run a 12V 360 fine, esp since the supply itself has no reference to earth ground as it doesn't even have a ground pole on it.

You might want to go read up on power supply theory. ATX = someone else has already done all the hard work for you. Same with that aux supply you linked. Someone already did all the hard work for you to make it safe for computer use. Since you just called it a "power brick" that should tell you how much you don't know about PSU design.

 

[PottedMeat]

looks like someone already did it - https://www.youtube.com/watch?v=0TOhvrsE93A

360 203W Power Supply -> 2x6 Pin PCIe -> GTX275

they even used the PS_ON sense tied to the PC +5V to wake up the 360 when necessary

 

[imagoon]

Guess that answers that. Seems quite a few people on the net have done it.

 

[Micrornd]

You make quite a few assumptions there. (-) isn't always earth ground. In DC "ground" is commonly called (-) and it not the same thing as earth ground. It also can be incorrect as (+) ground systems also exist. There simply is no way guarantee the 360 supply uses the same potential as an ATX supply. It might but there is no guarantee.

#1 earth ground should have no potential. Earth ground can be (+) to one supply and (-) to another.

So as switching power supplies, voltages is the delta between the negative and positive leads. Two power supplies are not required to derive (-) at the same potential. Two switching supplies can derive 12 volts between their own negative and positive lead and there is no requirement for the delta between both negatives to be 0V. The exception to this is ATX which specification wise states that (-) should be 0V in relation to the earth ground. Since the 360 PSU is not ATX, it could develop (-) at delta 1000V from earth and 1012V from earth and run a 12V 360 fine, esp since the supply itself has no reference to earth ground as it doesn't even have a ground pole on it.

That is exactly the point, I have been trying to get you to understand.
Once plugged into the video card the potential is eliminated because both switching supplies use and are tied to the same ground reference.
Neither will have a ground reference voltage higher than the other, once their respective "grounds" become tied together as happens when plugged into the video card, and both will produce their rated regulated output.
(And if I have to say it, I will, The 2 PSUs should not be plugged into the PC or video card when powered on, OK?)

BTW - why do you say the 360 PSU is not ATX spec., do you know this for a fact? And for that matter why do you think that really matters?
Cisco 48v PSUs, Dell 5v with no 3.3v PSUs, micro and nano "bricks", laptop "bricks", etc. none meet the full ATX spec. and neither do many of those sold for PC PSUs (depending on the revision) yet they all seem to power "PCs" without a problem.
Adapt and innovate, it is what separates use from the "animals" ^_^

You might want to go read up on power supply theory. ATX = someone else has already done all the hard work for you. Same with that aux supply you linked. Someone already did all the hard work for you to make it safe for computer use. Since you just called it a "power brick" that should tell you how much you don't know about PSU design.

Don't get out much, huh?
I referred to the aux. power supplies as a "power brick" because of it's shape, which is what that style of external power supply is known as by most of us in the industry that use externally located PSUs.
They can have single or multiple voltages, they can or cannot meet any number of specs..
Whether they look like a brick or a shoebox they are still referred to as "bricks", and in today's market 85% are switching power supplies.

I hope my rhetoric has not offended you, but I can't see any other way to get you to see the obvious.
There is diagram in the link I provided showing how the Aux. booster video power supply connects, that is identical to what the OP wants to do.
I just don't know how I can may it plainer that this works and is not dangerous (beyond any normal stds. of PSUs and PCs precautions) :'(

 

[TheStu]

looks like someone already did it - https://www.youtube.com/watch?v=0TOhvrsE93A

360 203W Power Supply -> 2x6 Pin PCIe -> GTX275

they even used the PS_ON sense tied to the PC +5V to wake up the 360 when necessary

First time I saw that. I had looked into this before I started, but most of what I found was the reverse, using a PC PSU to power a 360.

Thanks.

 

[imagoon]

That is exactly the point, I have been trying to get you to understand.
Once plugged into the video card the potential is eliminated because both switching supplies use and are tied to the same ground reference.

Untrue. The is the possibility of harmonics and competing load that can disrupt the reference in one or both supplies causing high current ground loops or disrupting voltage management. The point you keep ignoring is if the potential is different, the video card becomes the current path and likely will fry in the process.

But you seem to think you know more than people like TI engineers who actually design these supplies:

http://www.ti.com/lit/an/slyt298/slyt298.pdf

Pay attention to figure 6. The middle one is what you are building when you link the 2 switching supplies.

Due to potential loading differences there is a decent chance, without proper isolation, to generate large currents, harmonics and noise in the negative / grounding path.

Again read up on PSU design before you keep trying to correct people. There is more going on that you know in proper power supply. Doubly so in a computer / electronics power supply.

 

[TheStu]

Untrue. The is the possibility of harmonics and competing load that can disrupt the reference in one or both supplies causing high current ground loops or disrupting voltage management. The point you keep ignoring is if the potential is different, the video card becomes the current path and likely will fry in the process.

But you seem to think you know more than people like TI engineers who actually design these supplies:

http://www.ti.com/lit/an/slyt298/slyt298.pdf

Pay attention to figure 6. The middle one is what you are building when you link the 2 switching supplies.

Due to potential loading differences there is a decent chance, without proper isolation, to generate large currents, harmonics and noise in the negative / grounding path.

Again read up on PSU design before you keep trying to correct people. There is more going on that you know in proper power supply. Doubly so in a computer / electronics power supply.

So what steps should be taken to reduce this problem? Is there a way to peg the 360's ground to the PC's in some way?

 

[imagoon]

So what steps should be taken to reduce this problem? Is there a way to peg the 360's ground to the PC's in some way?

Typically you would use a couple of diodes, some load resistors and caps. However since the demos on the internet don't result in the supplies letting the magic smoke out, there is high odds that the PSU's they are using already had at least the load control and diodes inside the unit to prevent the back feeding power from each supply back tracking in to the secondary stage of the PSU units.

Anyway, this is a more generic response:

Use your meter and connect one lead to (-) on each supply, plug them in, power them and load them up a bit. If DC and AC volts = 0 then they are internally referenced to the same thing. This might be unlikely since the 360 PSU is not grounded to the wall as evidenced buy the 2 prong cord. Personally I would then use high impedance sandbar resistor to connect the grounds and see if that pulls the (-) down to 0V. If it does, you should do a basic current check (IE if the sand bar gets hot then you have a problem.) You would then use some high current diodes to prevent power from back feeding in to the supplies, wiring them per current flow.

The issue being that a supply that floats a little higher voltage (say 12.05V vs 12.0V) will generally provide most of the current in the circuit until the voltage droops below that of the other supply and the trend will reverse. You would have to build a droop circuit to try and deal with this issue so the supplies don't sway current loads.

more specific to you:

Do the (-) to (-) meter test I mentioned These supplies are likely are already dioded and internally loaded. So I would put a small capacitor on the 360 supplies 12V to 0V. A tiny guy like this:

http://www.parts-express.com/47uf-25v-radial-mini-electrolytic-capacitor--020-1082

should be enough. Make sure to get the polarity straight. Also there is pretty high odds the video card already has a pretty decent 12V isolation circuit but you might want to check that 12V on the 6 pin is not a direct short to 12v on the PCI-e bus side. Odds are it is pretty high already, but better safe than sorry. If both supplies are showing 0V you can also tie the supply (-) together.

 

[TheStu]

Typically you would use a couple of diodes, some load resistors and caps. However since the demos on the internet don't result in the supplies letting the magic smoke out, there is high odds that the PSU's they are using already had at least the load control and diodes inside the unit to prevent the back feeding power from each supply back tracking in to the secondary stage of the PSU units.

Anyway, this is a more generic response:

Use your meter and connect one lead to (-) on each supply, plug them in, power them and load them up a bit. If DC and AC volts = 0 then they are internally referenced to the same thing. This might be unlikely since the 360 PSU is not grounded to the wall as evidenced buy the 2 prong cord. Personally I would then use high impedance sandbar resistor to connect the grounds and see if that pulls the (-) down to 0V. If it does, you should do a basic current check (IE if the sand bar gets hot then you have a problem.) You would then use some high current diodes to prevent power from back feeding in to the supplies, wiring them per current flow.

The issue being that a supply that floats a little higher voltage (say 12.05V vs 12.0V) will generally provide most of the current in the circuit until the voltage droops below that of the other supply and the trend will reverse. You would have to build a droop circuit to try and deal with this issue so the supplies don't sway current loads.

more specific to you:

Do the (-) to (-) meter test I mentioned These supplies are likely are already dioded and internally loaded. So I would put a small capacitor on the 360 supplies 12V to 0V. A tiny guy like this:

http://www.parts-express.com/47uf-25v-radial-mini-electrolytic-capacitor--020-1082

should be enough. Make sure to get the polarity straight. Also there is pretty high odds the video card already has a pretty decent 12V isolation circuit but you might want to check that 12V on the 6 pin is not a direct short to 12v on the PCI-e bus side. Odds are it is pretty high already, but better safe than sorry. If both supplies are showing 0V you can also tie the supply (-) together.

The 360 PSU I have does take a 3 prong power connector. It's a key version of the same kind used on a PC PSU, but I'll do the multimeter test, and I'll try and rig up that cap.

Thanks

 

[Micrornd]

Untrue. The is the possibility of harmonics and competing load that can disrupt the reference in one or both supplies causing high current ground loops or disrupting voltage management. The point you keep ignoring is if the potential is different, the video card becomes the current path and likely will fry in the process.

You might want to read the article you quoted.
Or at least read the 1st paragraph, where Tom explains what the white paper is about.
While Tom does somewhat explain how a switched-mode PS works, he is only using it to illustrate how a ground potential can exist between a local and a remote site, especially if the structures' grounding systems are different and signal isolation is not used.

But you seem to think you know more than people like TI engineers who actually design these supplies:

http://www.ti.com/lit/an/slyt298/slyt298.pdf

Possibly, but I've only played poker with Tom once, although I did walk away with a good chunk of his stake. :whiste:

Tom's white paper doesn't apply to what is being discussed here.
Both PSUs are connected locally to the same PC, which in turn is connected to a single structure ground system, not two separate electrical devices separated by distance, communicating with each other with a ground potential being enacted across the ground or drain of the communications cable, caused by using separate (and possibly different style) structure grounding systems.

Pay attention to figure 6. The middle one is what you are building when you link the 2 switching supplies.

No, it isn't.
As I said above, if you read Tom's white paper it is about the dangers of ground potential in linking the grounds of local and remote sites and the additional pitfalls in noise and disruption of signal that can occur.
Tom specializes in signal management and transmission on the hardware level and predominantly in using those signals on the software level.

Due to potential loading differences there is a decent chance, without proper isolation, to generate large currents, harmonics and noise in the negative / grounding path.

No to most of that as it doesn't apply to what the OP proposed , and the level of noise is well below threshold.

Again read up on PSU design before you keep trying to correct people. There is more going on that you know in proper power supply. Doubly so in a computer / electronics power supply.

You're right, I haven't had to design a SWPS in probably 20 yrs., I either delegate or most of the time specify just what and where to buy.
But neither of those choices doesn't mean I don't know how or even more importantly doesn't mean I don't know what will or will not actually work

So, I guess what you are saying is multiple PSUs serving the same PC, redundant 1 of 2, 2 of 3, 3 of 4 etc. PSUs, or Auxillary Video PSUs can't work or are dangerous.
I'm sure that is news to the industry and all the rest of us that use them in their designs and/or products.

I'm done with this thread, feel free to believe what you want to, I'm sure the world was always flat until someone proved it round, but I'm done trying to prove that the OP's proposal is perfectly workable.

 

[imagoon]

You might want to read the article you quoted.
Or at least read the 1st paragraph, where Tom explains what the white paper is about.
While Tom does somewhat explain how a switched-mode PS works, he is only using it to illustrate how a ground potential can exist between a local and a remote site, especially if the structures' grounding systems are different and signal isolation is not used.


Possibly, but I've only played poker with Tom once, although I did walk away with a good chunk of his stake. :whiste:

Tom's white paper doesn't apply to what is being discussed here.
Both PSUs are connected locally to the same PC, which in turn is connected to a single structure ground system, not two separate electrical devices separated by distance, communicating with each other with a ground potential being enacted across the ground or drain of the communications cable, caused by using separate (and possibly different style) structure grounding systems.


No, it isn't.
As I said above, if you read Tom's white paper it is about the dangers of ground potential in linking the grounds of local and remote sites and the additional pitfalls in noise and disruption of signal that can occur.
Tom specializes in signal management and transmission on the hardware level and predominantly in using those signals on the software level.




No to most of that as it doesn't apply to what the OP proposed , and the level of noise is well below threshold.



You're right, I haven't had to design a SWPS in probably 20 yrs., I either delegate or most of the time specify just what and where to buy.
But neither of those choices doesn't mean I don't know how or even more importantly doesn't mean I don't know what will or will not actually work

So, I guess what you are saying is multiple PSUs serving the same PC, redundant 1 of 2, 2 of 3, 3 of 4 etc. PSUs, or Auxillary Video PSUs can't work or are dangerous.
I'm sure that is news to the industry and all the rest of us that use them in their designs and/or products.

I'm done with this thread, feel free to believe what you want to, I'm sure the world was always flat until someone proved it round, but I'm done trying to prove that the OP's proposal is perfectly workable.

Induced currents in the grounds / (-) feeds is power supplies 98 / high school level. All you are doing to me is proving you don't really know anything about it. Go ahead and pick random switch mode supplies together and tell us how it goes.

I also highly doubt your first person account.

It doesn't matter is the power supplies are 8 miles away or 3 inches. If they are not properly isolated, controlled and balanced it will cause damage to the supply and the devices it is running.

So, I guess what you are saying is multiple PSUs serving the same PC, redundant 1 of 2, 2 of 3, 3 of 4 etc. PSUs, or Auxillary Video PSUs can't work or are dangerous.
I'm sure that is news to the industry and all the rest of us that use them in their designs and/or products.

See here again you are making things up again. I never said this. However redundant power supplies in these devices are rarely a 360PSU spliced in to a video card. Redundant PSUs are typically attached together with a controller board the assists in load balancing when applicable and controlling fail over as needed.

I'm done with this thread, feel free to believe what you want to, I'm sure the world was always flat until someone proved it round, but I'm done trying to prove that the OP's proposal is perfectly workable.

Feel free to point out where I said this was impossible and unworkable. You won't find it. Like most of your points there were made up on the spot for this thread. I just said what amounts to: before you hook it all up, check it with a meter so you don't find yourself with a pile of burnt up electronics.

If you are going to keep making stuff up, then please, leave, you will be helping the OP.

 

[imagoon]

Stu, to show that I am not jerking you around:

http://www.servethehome.com/redundant-power-supply-units-psus-how-they-work-quick-overview/

As I mentioned above I don't want to see your next post as "Well it all smoked up." Hooking ATX supplies to each other isn't just an ATX "Y" cable. Typically there are some other guts going on with them and some extra stuff inside the supplies themselves. Same thing with just attaching a 360 supply or any supply to another.