old ATX power supply needed for science project

This pertains to generic ATX.

shows that pins 14 and 15 should be shorted. The colors correspond to an ATX power supply I have in a "clone" computer.

Dell is different, though.

looks right, based on wire colors.

I presume I short the grey wire to a black wire, to turn the thing on. Any advice appreciated.

Thanks! I just found the same website. The Dell ATX disagrees with the generic ATX, though. Dell has proprietary pinouts. The grey wire must go low.

that

Actually, if I am to interpret the diagram for the Dell pinouts literally, the PS_On (grey wire) should be tied high (+5?) to turn the power supply on. Yeah, but that doesn't make a lot of sense, unless that very line is considered to be "low" when open... Well, I suppose anything is possible, but I will be that a tie to ground will turn the thing on. That would imply that a correction is in order for that diagram.

The same website made a point if declaring the green wire to be "true low" by naming it "/PS_ON." So I guess I'll have to experiment.

"John B" wrote Dell has proprietary pinouts.

I shorted grey to black, and nothing happened. I was hoping the cooling fan would turn on. I then shorted grey to other colored wires, and still no action. The computer from which this old p/s was pulled was not dead...just obsolete. So the p/s is supposed to be OK. Tomorrow my son will pull the p/s out of another old Dell. He'll try the same thing. But this will have to wait until tomorrow. The p/s we tried has a fuse deep inside, and not the kind you can see through. So I unplugged the power cord from the wall receptacle, shorted the power prongs to dissipate any residual charge, and rang the 117VAC prongs with an ohmmeter. There was some conductance...implying that the fuse is OK. Mañana.

to

There is some more info here

that corroborates the theory that the grey wire should be shorted to ground. So I shorted the grey wire to the chassis, just in case that's different from the black wire. The fan did not start up. I put a DC voltmeter between red and black, hoping to see 5 VDC. When I shorted the grey wire to the can, I noticed a very brief blip between red and black. That implies that the p/s is folding back, after briefly supplying +5, perhaps as a consequence of the zero load at the various voltage supply lines. There is no motherboard here. I shorted grey to chassis again, this time monitoring for +12 VDC. Same thing; momentary blip on the voltmeter. So I think there is some foldback issue, here, in the p/s design.

"Jonathan Westhues" wrote

power

if

Bring nPS_ON low. In other words, tie the green wire to one of the black wires.

See

for the pinout.

Jonathan

How strange; I thought that was standard. But I would say yes then, grey to black.

Jonathan

Yeah. There was nothing at +5VSB. I like you're idea of powering an old disc drive. I've got lots of junk like that. I just don't want to connect the motherboard from the cannibalized computer, if I can help it. So you agree...that "no load" equals "no go"? I wonder what the reason for that would be? I suppose the gods of mobo design have some reason for this. Like Tevya said in "Fiddler on the Roof," would it upset some great eternal plan, if I were to know... WHY?!

red

regulation)

drive

powered

fan

dead...just

You might try measuring the voltage between the +5VSB line (pin 6 with the strange Dell layout) and ground. It should be about 5 V, even when the power supply is "off".

Jonathan

This PSU should not even be designated ATX. It doesn't conform to ATX pinouts. While I've never purchased an assembled system you can sure bet I'll never purcahse a Dell after seeing this.

--
Abit KT7-Raid (KT133) Tbred B core CPU @2400MHz (24x100FSB)
My server http://wesnewell.no-ip.com/cpu.php
Verizon server http://mysite.verizon.net/res0exft/cpu.htm

Wait, you don't have any load? A lot of them will run (with bad regulation) with no load but they aren't designed to. Can you connect an old disk drive or something, or a light bulb or a power resistor across the 5 V rail?

Did you measure the +5VSB line to make sure that the power supply is powered and kind of working?

Jonathan

It might work with no load or it might not, but if it works at all then it will work with a load. Hook up an old hard drive or something and you will be sure.

If you check to make sure that the +5VSB voltage is present before you short any wires together then you can at least be sure that you have the pinout right.

eternal

Computer power supplies are switchmode power supplies. They transfer energy from a high voltage DC rail (generated directly off the mains voltage) into the load voltages in short pulses. The power supply can only put energy into the load; with no load there is nothing consuming energy, and it is difficult to control the output voltage so it shuts down.

Jonathan

The black wires are Ground anything Yellow, Red or Orange will be supply, Any other colors may be for switching.

No-load operation is an added expense, not justified in the original application. These units were/are extremely cost-critical.

Some units may even overshoot - triggering an overvoltage protection circuit. Reseting this may require removal of input power for some minutes.

RL

I understand your indignation, where Dell has "hijacked" the term ATX while imposing a proprietary wiring specification. HOWEVER. Dell has been making fabulous computers for many years, offers good warranties, and has many trampled competitors, not to mention loyal repeat customers, to prove its suporiority in the marketplace.

Although I build my own, as do you, most people don't any do that more than they build their own cars. And I have been in too many noname computer "factories" where simple anti-static rules are ignored. That compares to medical practices that ignore anti-biotic precautions. Of course, testing was virtually non-existent.

Years ago, I compared Dell to Gateway. Gateway was cheaper, but admitted to testing only one in five computers coming off the line. Dell tested every one.

"Wes Newell" wrote >

The power supply works. It needs a load, such as an old floppy drive. I used an old 5 1/4" floppy drive. Interestingly, the load must exist BEFORE the grey wire is shorted to ground AND (in the Boolean sense) the 117VAC line is plugged in. IOW, either of the two latter conditions can be used to turn the p/s on or off. Furthermore, the load status is latched by the p/s, at the moment the p/s is turned on. IOW, I can disconnect the load after powering up the p/s, and voltages remain on the colored wires. I notice, however, that the +12 (yellow) degrades on my DVM when the load is disconnected. Simply stated, load status is latched at the moment the grey wire is grounded and the 117VAC is present.

It was not sufficient to place a mere 150 ohm resistor between +5V and GND. The floppy drive sufficed.

The power OK line was +5 only when voltages are available at p/s outputs. Furthermore, the cooling fan turns on under the same condition. So the power OK line and cooling fan are equal indicators.

Thanks for your help. My son will be delighted.

"Jonathan Westhues" wrote

short

energy

into

into

From the way I see it, Gateway committed suicide by opening "dedicated" showroom stores. Dell lives because they provide a competitive product (nothing special) and don't have showrooms or distributor shelves to stock with equipment that is soon antiquated - at least price wise. With this model, Dell has much lower business cost overhead.

Forrest

Motherboard Help By HAL web site:

< snip > < snip >

do not switch them on without a load or dummy load.

if

I have an idea that Dell was better capitalized, too. IOW, investors were more willing to plunge dollars into Dell than Gateway. So how did Dell earn that enviable status, many years ago? By relentlessly earning customer loyalty. By leading the field in "mail order" marketing. Dell has always had a sensational catalog, and a fabulous website. And though Dell was typically a bit more expensive than Gateway, its customers swore the difference was worth the cost.

to

every