Why you shouldn't open power supplies (w/video)

[Marinetuck]

If the Cap is large it can kill you or really F ing hurt... of course most people have no idea what a cap is or what it does lol any one want to school these people up lol

Drain the caps before playing with them and use a multi meter to make sure.

 

[larrymoencurly]

If I'm scraping computer parts and take apart a PSU...
how long do I have to let the PSU sit before taking it apart. I don't want to get zapped. I saw that the APFC section can take a day to drain, so is one day good?

If the PSU is working OK, then the fastest way to discharge the capacitors is by leaving it turned on while disconnecting the AC from it. That should drain the capacitors completely in 2 seconds. If that isn't done (or the PSU doesn't work), then most likely the bleeder resistor across each high voltage capacitor will discharge it in under 20 minutes. But you can't put all your faith in bleeder resistors because sometimes they've run hot for so long that they or their solder joints crack, or the PSU may not have any bleeders, as was the case with my 350W Enermax. I learned that when I left the AC disconnected from it overnight but its fan blipped anyway when I pressed the computer's front panel power button the next day before turning on the AC. Regardless, always take the precaution of unplugging the AC, waiting 20 minutes, and then shorting each big heatsink to the case ground before you touch anything inside. You should be able to short the heatsinks to the case before removing the case top, by poking a screwdriver with a plastic handle through the vent holes or fan grill (be sure you hit only the heatsink, not other components inside), but make sure you get a good ground connection because paint can prevent that. If you connect the screwdriver shaft to the ground prong, do it before poking the screwdriver into the PSU.

 

[eskimochaos]

That was extremely unexciting.

 

[micjmac]

hahahaha ive seen this happen a time or two, actuallly happened to me once... i had charged it, i use a test light instead of a resistor so its easier to tell when its charged, no more light but then i took it out to do some work and appearently i left it out too long (few days) and didnt recharge it hooked up the positive and pshew!!!!!! sparks everwhere and melted half the eye connector off some 4 gauge i think it was a 2 farad as well stopped using them caps once i learned moar about systems though. now i have a battery bank

You guys had to say it. Now I want to see a video of a screw driver getting blown up hahaha

 

[jazon1]

you want to see some big arks check this one out lol FF to about 1:15 for the fun to start make sure you have your sound down a little lol.

 

[Ivy]

940 uF @ 450V from the Corsair 1200i. That should be the highest possible value, the EVGA Supernova got same value on the caps. So we can expect close to 1000 uF on the biggest supplys.

Someone wanna touch it?

Anyway, the stuff Bobnova used on the experiment was only about 1/3 of the charge of the most huge PSUs and it was charged to 130 V only, so it was a baby cap at a weakened condition. Its usualy a cap used inside 500 W PSUs i guess. So imagine what can happen when the most huge cap is charged at max... i dont like to know, its just not fun to watch, to much toast.

 

[PolePosition]

It was always understood to me that its the amps that will kill you, not the voltage.

 

[Bobnova]

It takes volts to get the amps places. You can expose your skin to a 1v supply that has the capability of 1000000 amps behind it and nothing will happen, your skins resistance is too high to flow any meaningful amperage at 1v.

Now if you reduce the skins resistance, say by smearing it with ECG probe paste, you'll get some current flow.
Or if you raise the voltage, 100v is enough to get some current through the skin. More voltage = more current.
Ohms law.

 

[wagex]

It takes volts to get the amps places. You can expose your skin to a 1v supply that has the capability of 1000000 amps behind it and nothing will happen, your skins resistance is too high to flow any meaningful amperage at 1v.

Now if you reduce the skins resistance, say by smearing it with ECG probe paste, you'll get some current flow.
Or if you raise the voltage, 100v is enough to get some current through the skin. More voltage = more current.
Ohms law.

thank you for the explanation bob i never did understand how that worked because i have been sweaty and been zapped by car batteries and i know they have a ton of amps lol thank you for the explanation, i figured it was something along those lines but i never knew for sure

 

[PolePosition]

Still doesn't dispute the facts.
http://electrical.about.com/od/electricalsafety/a/amperagekills.htm

 

[Bobnova]

There are no facts on that page.
For starters, about.com is a horrendous source.
For second, it doesn't mention voltage anywhere.
For third, it doesn't mention the fact that where you get the current flow makes a huge difference. You can flow amps through your toe and live. You can flow 40ma through your heart and die.
For fourth nowhere does it talk about how the amperage is getting there.
For fifth, please go read up on ohms law. You need the background of knowing why current flows to talk about current flowing.

Go grab both posts of a car battery with dry hands. That source can cough up 600-1000 amps. You won't even get a tingle. That is a fact.

 

[wagex]

Go grab both posts of a car battery with dry hands. That source can cough up 600-1000 amps. You won't even get a tingle. That is a fact.

if your drippign sweat though and touch the positive, the ground will give you a bite on the forearm but thats also a very sensitive area

 

[Bobnova]

True, heh. Note that you aren't dead, despite it being a source that can cough up hundreds of thousands of milliamps!

I assume you're not dead anyway.

 

[wagex]

idk im not all here sometimes

oh noes i cant find my pulse :O

 

[PolePosition]

There are no facts on that page.
For starters, about.com is a horrendous source.
For second, it doesn't mention voltage anywhere.
For third, it doesn't mention the fact that where you get the current flow makes a huge difference. You can flow amps through your toe and live. You can flow 40ma through your heart and die.
For fourth nowhere does it talk about how the amperage is getting there.
For fifth, please go read up on ohms law. You need the background of knowing why current flows to talk about current flowing.

Go grab both posts of a car battery with dry hands. That source can cough up 600-1000 amps. You won't even get a tingle. That is a fact.

I'll have to keep that in mind the next time About.com gets referenced here (which has happened) with links to other facts irrelevant to this thread since you're saying everything there is false. Perhaps you will believe this site as a fact, where current is equivalent to amps;
http://www.physics.ohio-state.edu/~p616/safety/fatal_current.html
My point was about electrical basics, and if you don't believe it, ask any licensed electrician about what actually kills you. Of course you do need voltage to carry current. I'm not disputing that fact, but its not the voltage that kills you, its the amps, which are delivered by the voltage.
A comparable analogy would be to stand in front of a vehicle at idle. The driver steps on the gas and runs you over. The gas did not kill you, the vehicle did.
Its true that grabbing a battery's terminals with both hands won't shock you. That is because the circuit is not grounded to you, it is grounded to the body of the vehicle, insulated by rubber tires to actual ground. Connect two leads to the + and - terminals and stick both to your tongue and see what happens.

 

[Bobnova]

The only difference between your scenario and mine is the contact location. You can touch 'em while hooked up to the car. You can touch 'em disconnected. It does not matter.
"Ground" is an incorrect concept in either situation, too.
You are a complete circuit from + to -, which is all that matters.
You can hook that battery up any way you'd like to, and it still won't shock you if you have dry hands. Measure your resistance sometime with a multimeter and then play with ohms law and voltages for a while.

Yes, current kills you. This does not mean that you should not be afraid of voltages in capacitors.
Nor does it mean that you should be afraid of current sources. That is my major issue with the "Current kills not voltage!" statement. Without voltage, current does not flow.
With insufficient voltage, insufficient current flows.
This is especially relevant to capacitors as they are capable of supplying tremendous current. The current in the video above is many amps, for instance. The capacitors on the 12 V rail of a PSU are capable of supplying even more amperage. Far more, and for a longer duration of time as well (which is important and often ignored in the "OMG CURRENT KILLS" links).
Are they dangerous? No, not really. They lack enough voltage to get meaningful current through your skin.

The APFC caps at 380 V however have plenty of voltage to get meaningful current through your skin, and will do so!
Also, while 100 V and 100ma can kill you, 10,000 V and 100ma will do it rather faster.


As a final note, you do realize that outward appearances are that you're arguing that APFC capacitors are safe to play with, right?

 

[PolePosition]

I've changed capacitors in AC condensers too many times to count, and never got the slightest shock, and they are 220v. I just flipped the breaker to cut off the current and immediately began my work. I've also wired whole houses and replaced hundreds of service panels which previously had buss fuses and knob and tube wiring, but all that is a bit different from computer PSUs.
I tore the old PSU apart right after it had been connected to a power outlet to see what was wrong or causing the problem and never thought once about getting shocked. Now I won't say APFC capacitors are safe to play with and not one of my posts has made that statement whatsoever. It is always better to be safe than sorry when fooling with ANY electrical components. Always unplug the power source. I only made a simple statement about how it was "my understanding that it is the amps(current) that kill a person, not the voltage".

The things is, if a car battery is 12v and can't kill you, and a PSU only delivers 12v at most on any rail, then I guess it too can't kill you. LOL it might shock you like static electricity when running your feet across carpet and touching a metal object, which is all I saw in the video. Now if you had the PSU plugged into the wall outlet while connecting that alligator clip.......That would make sense how it got welded, and certainly give one a jolt. I'm assuming that clip had an insulated end where you grabbed it.

1.5 amps is the most you will find on any PSU 12v rail
12 amps is the most you will find on any PSU 5v rail.
Other rails measure less than 0.?? amps. I can't recall actual figures.
These are figures on a 130w PSU, I'm sure they could be more with higher powered PSUs. Something I need to read up on and learn more about. I never bothered to check the specs that deeply with my 750w PSU. I just made the connections and moved on.

Can you name the amps of the PSU in the video of that particular rail where the spark occured? I'm just curious, thats all. They should be printed on the UL label on the enclosure. It'd be interesting to know just how much current was passing through.

Finally, I don't see many people really playing around with PSUs. When its determined one has failed, the user usually just unplugs, unmounts, tosses, and installs a new one.
LOL, they're not like us guys who get into splicing wires to make em work.

 

[Bobnova]

There was no PSU in the video. Not a computer PSU anyway. Just the cap out of the APFC bits with a homebuilt buck regulator to charge it. The alligator clip is connected to cap -, it touches a lead connected to cap +.
The cap was charged to the voltage shown on the multimeter.

You can find rather more than 1.5 amps on a 12v rail, I'm not sure what you're talking about there.

The sparks you get from shorting a PSU 12v rail can be pretty solid too, I've taken to testing SCP lately, on a 40a rail that makes for a nice spark before the SCP kicks in! Less of a POW than the cap gives though.

 

[BeepBeep2]

This should be completely safe after a few hours to let it discharge, correct?

 

[Bobnova]

In theory? Yes.
All the PSUs I've opened so far have enough self-discharge to make it safe, and the caps that don't self-discharge (the X caps in the transient filter) either have a bleed resistor or an active bleed IC to discharge them.

I haven't been opening cheap PSUs though, there may be a point where that becomes too expensive!