Modifying a Power Supply

[dmw16]

Here's the story...

I want to modify an Apple G5 case to house a PC. I have a pretty good idea how to tackle most of the problems, but there is one that has me a little bit puzzled and I am hoping someone here might be able to help.

The corners of the Apple case have radii and as a result a standard PC power supply won't fit. So what I'd like to do is remove the guts from a regular PC power supply and drop them into a Mac power supply chassis. (Very similar to what's seen in THIS thread, post #7.)

I know there is some serious voltage stored in the capacitor(s) inside the power supply, but I am wondering if anyone with more knowledge of these things could tell me if I can do this transplant without having to discharge the cap(s)? Or if I do, what is the PROPER (ie, no screw driver ) method of discharging it?

This may be beyond my skill level as I don't know a ton about these things, but I do know enough to know there are some dangerous components inside.

Any advice would be appreciated, thanks.

 

[thaltek]

if your your looking to be safe unplug said powersupply and let her sit out for an hour.... during this time any charge stored up in the capacitors "should" be discharged through a discharge resistor...... should you go the more adventurous route and want to play right away then if you do the following you should be safe.... first find a pair of "dish-washing gloves" heavy chemical gloves should also work just so long as they are plastic or rubber and don't have any holes.....
1. open the psu AFTER ITS UNPLUGED and be careful not to touch internal components.
2. unless its a reallly expensive power supply (in which case your better off leaving it sit out) then with care (only touching plastic parts) unscrew the main board of the psu from the case...... carefully set the whole thing upside down....
3. now with a suitable resistor locate where the capacitor is soldered onto the board and take the resistor and short out the terminals.....

a quick warning about shorting out a capacitor.... its really not a good idea to use a screwdriver or metal implement as it shortens the life of the capacitor and can also in rare situations cause the capacitor to explode..... please only use a suitable resistor to discharge in a controlled manner.....


but really i take psu's apart quite often..... so long as you remember to not touch metal bits your fine....

 

[Elledan]

Proper way to discharge a PSU is with a big resistor (around a MOhm or so) in between the capacitor and ground. As for the danger level, the voltage of a capacitor isn't high enough to penetrate the skin, so you'll just get shocked really well (been there, done that, got shocked). Still very uncomfortable, though.

PCBs are best handled around the edges, i.e. avoid touching ICs and other sensitive components. Use a grounded metal part near you to lower your potential if possible. Using a ground plane on the PSU PCB itself works too.

 

[preyash]

I want to see pics when your are done. I would practice witha non working PSU first if you are doing it first time.

 

[keenan]

I'd use something considerably smaller than 1MR. That's going to take quite a while to discharge a power supply bulk capacitor - they store a lot of energy and the 1M will burn it off quite slowly, several minutes at least for a typical capacitor. I'd use a 47K 1W resistor instead. Use your DMM to measure the voltage across the capacitor until it's at a safe value. A good way to do this is to alligator clip one end of the resistor to the PSU case or some other easy ground, and alligator clip your DMM ground to same. Then tape the resistor body to a plastic pen tube or some other insulating material and clip the positive lead to the other resistor end (also sticking out the bottom of the pen). Then you can probe around and ground the caps safely. A similar approach would use a short alligator clip lead to your DMM's positive probe instead of the pen, just using the probe as the contact.

Most PSUs should have self-discharge circuits for safety. They'll make it slightly less efficienct, but there should be enough leakage that the caps are discharged within a couple minutes after power is removed. Double check with your DMM, but letting it sit for a while is probably the easiest and safest way to go, just make sure you test all the capacitors before handling.

Many power supplies will have upwards of 300V appearing internally. Be very careful. Despite the fact that you'll find many people who've shocked themselves and lived, this voltage is high enough and the energy stored more than large enough to kill you. Be extremely cautious and for safety always work with one hand behind your back (this way you won't accidentally create a circuit between your hands and thus across your heart).

 

[Format _C:]

Here is a tip If the PSU is still connected to a Motherboard just unplug it from the wall outlet and press the power button and you should see the Leds flash and the Caps should be discharged

 

[thaltek]

or just to what format c just said........

Many power supplies will have upwards of 300V appearing internally. Be very careful. Despite the fact that you'll find many people who've shocked themselves and lived, this voltage is high enough and the energy stored more than large enough to kill you. Be extremely cautious and for safety always work with one hand behind your back (this way you won't accidentally create a circuit between your hands and thus across your heart).

ah, yes..... but its worded a little wrong.... normal static electricity runs about 3kV.... and last i checked my grandpa is still running around on carpet in wool socks in the winter.....>_>

its not the voltage that kills its the current..... but the tip of having one hand behind your back is right on and is common practice in industry......

 

[keenan]

ah, yes..... but its worded a little wrong.... normal static electricity runs about 3kV.... and last i checked my grandpa is still running around on carpet in wool socks in the winter.....>_>

It's not wrong at all. Please point out the error. Capacitors don't store current, they store energy. Current is just a measure of the movement of charge, it's meaningless until a circuit is created and the current actually flows. It would be inaccurate to describe it any other way than I did. A large quantity of stored energy plus a high voltage is a deadly situation - it's these two things that combine to generate the deadly current that kills, as you rightly point out, but you need both high voltage and sufficient energy and the current is the result, not the other way around. Your grandpa makes a very poor capacitor and stores so little charge that virtually no current flows when a circuit is created. If he stored more energy, it could be dangerous.

Why do people feel the need to 'correct' when there's nothing wrong with the original statement...

 

[jeremyshaw]

it's e-peen, of course. I feel (felt) the need to bash anybody who slighted me in the least.

It's easier to 'look good' online, and this has been happening a lot to [H] recently. Too much thread crap.

Hope that answers it!