Opening A Power Supply [Safety Question]

Chocolinx · Post by **Chocolinx** » Sun Apr 29, 2007 8:29 am

I keep on hearing around different forums that I need to be extremely careful when opening power supplies; even when they aren't plugged in. Can someone explain to me the dangers of opening a power supply and why it's so dangerous?

angelkiller · Post by **angelkiller** » Sun Apr 29, 2007 8:56 am

I'm no expert, but I've heard that capacitors store charge even when the power is unplugged. I've also heard that some heatsinks are "live" so you don't want to touch them. But I've never seen this personally though.

AuraAllan · Post by **AuraAllan** » Sun Apr 29, 2007 9:18 am

I think you'll be safe if you unplug the powercord and flip the power switch a few times. That should do it.

SilentKev · Post by **SilentKev** » Sun Apr 29, 2007 10:44 am

Flipping the power switch on the PS will not drain it.

Unplug the PS from the wall outlet and leave its own switch in the ON position. Next, press the computer's power switch in and hold it in for 10 seconds. This will drain most of the power from the PS. Your fans or lights may come on very briefly.

To completely drain the power a PS, a testing device will work best because it bridges all of the voltage rails.

EndoSteel · Post by **EndoSteel** » Sun Apr 29, 2007 12:02 pm

The capacitors discharge by themselves in 10-20 seconds after the unit is plugged off.

Mr Evil · Post by **Mr Evil** » Sun Apr 29, 2007 12:43 pm

EndoSteel wrote:The capacitors discharge by themselves in 10-20 seconds after the unit is plugged off.

Not a safe assumption. Depending on the design of the circuitry, capacitors may remain charged for months after power has been removed. SilentKev's suggestion is the best course of action.

SaxicolousOne · Post by **SaxicolousOne** » Sun Apr 29, 2007 2:06 pm

Yes, capacitors can most definitely remain charged for a while. I was doing a little something inside a device (a strobe tuner, actually) that had been unplugged for a few minutes, and it shocked the BEJEEZUS out of me. It really hurt. Then, as much as that reminded me to be careful, it happened again moments later! I wasn't even sure what I had touched.

Drain the device if possible, for sure. Then, a great rule I have heard is to not touch anything that you can't ordinarily touch when the unit is fully assembled. Use little pliers and screwdrivers with non-conducting handles and that sort of thing to do your work. On top of that, I wear gloves now. I know I probably look like a sissy, but so be it. I have a low pain threshold.

Chocolinx · Post by **Chocolinx** » Sun Apr 29, 2007 5:45 pm

SaxicolousOne wrote:Yes, capacitors can most definitely remain charged for a while. I was doing a little something inside a device (a strobe tuner, actually) that had been unplugged for a few minutes, and it shocked the BEJEEZUS out of me. It really hurt. Then, as much as that reminded me to be careful, it happened again moments later! I wasn't even sure what I had touched.

Drain the device if possible, for sure. Then, a great rule I have heard is to not touch anything that you can't ordinarily touch when the unit is fully assembled. Use little pliers and screwdrivers with non-conducting handles and that sort of thing to do your work. On top of that, I wear gloves now. I know I probably look like a sissy, but so be it. I have a low pain threshold.

This is what I'm most worried about. I heard that getting shocked by a toaster can kill you becuase of only 3A going straight to the heart. Even after a shock you can still die a few days later because you're heart is going through an irregular beat.

So, so far from reading the thread I got, do not touch anything but the metal from the enclosure. And unplug and hold power button for 10 seconds. Any other tips to avoid a charge going through my body? And any other scientific info on this matter?

Basically what I'm doing is leaving a PS open, only thing left will be the grill with the power cord slot and the circuit board with all the power thingies. The power supply that I'll be doing this to is a SS-300SFD it's a small form factor psu by Seasonic, SPCR has a review on it.

Mr Evil · Post by **Mr Evil** » Sun Apr 29, 2007 6:04 pm

Chocolinx wrote:Any other tips to avoid a charge going through my body?

Hold one hand behind your back when working on high voltages. That way you can't get current flowing up one arm, through your heart, and back down the other arm.

Don't worry about it too much though: a shock from a charged PSU capacitor can be painful (did it to myself once and had a nice pair of burn marks on one finger to show for it), but it probably won't kill you unless you have a heart condition.

mr. poopyhead · Post by **mr. poopyhead** » Sun Apr 29, 2007 9:31 pm

there's an entire thread about it here:
viewtopic.php?t=6123

but even with 2 pages of posts, there doesn't seems to be a definitive answer to this question...

can someone settle this once and for all?

jaganath · Post by **jaganath** » Mon Apr 30, 2007 1:56 am

there doesn't seems to be a definitive answer to this question...

which question? how to work safely on a PSU? SilentKev gave good advice,always worked for me:

Unplug the PS from the wall outlet and leave its own switch in the ON position. Next, press the computer's power switch in and hold it in for 10 seconds. This will drain most of the power from the PS. Your fans or lights may come on very briefly.

However I would press the PC ON switch repeatedly rather than hold it down.

Basically what I'm doing is leaving a PS open, only thing left will be the grill with the power cord slot and the circuit board

I did that w/ an el cheapo PSU once, had live heatsinks, shocked myself a couple of times!

how are you cooling it BTW?

EndoSteel · Post by **EndoSteel** » Mon Apr 30, 2007 2:36 am

Well, I'm not that good at electronics.. But AFAIK, all pulse PSU circuits have these (marked red)...

...and they are there for the sole purpose of draining the caps when the unit is turned off. Correct me if I'm wrong.

Mr Evil · Post by **Mr Evil** » Mon Apr 30, 2007 3:23 am

EndoSteel wrote:Well, I'm not that good at electronics.. But AFAIK, all pulse PSU circuits have these (marked red)...

...and they are there for the sole purpose of draining the caps when the unit is turned off. Correct me if I'm wrong.

There is no guarantee that those resistors will be present in any particular power supply. The one that I received a shock from had been switched off for several days.

SilentKev · Post by **SilentKev** » Mon Apr 30, 2007 5:56 am

Here is what I use to discharge a PS:

http://www.antec.com/us/productDetails.php?ProdID=77003

It is a simple device that jumpers the ATX connector pins that turn the PS on, just like the computer power switch. It then has a "Built-in 25W 5.4 ohm load to generate steady outputs." This load will drain the PS of any stored charge. I have seen similar devices at CompUSA and Best Buy.

Chocolinx · Post by **Chocolinx** » Mon Apr 30, 2007 8:49 am

jaganath wrote:I did that w/ an el cheapo PSU once, had live heatsinks, shocked myself a couple of times! how are you cooling it BTW?

It's basically in an upper chamber of the case closed off with a 120 Yate Loon at probably 4.x volts. Or at least will be when I get the PSU. I'll definitely take pictures when its all done with. Oh another thing, if the PSU is open and surround by "Bristle Board (Black)" what are the chances of the bristle board catching on fire? (When I get home I'll take pictures of the case I'm using)

Mr Evil · Post by **Mr Evil** » Tue May 01, 2007 8:34 am

Chocolinx wrote:...Oh another thing, if the PSU is open and surround by "Bristle Board (Black)" what are the chances of the bristle board catching on fire? (When I get home I'll take pictures of the case I'm using)

I had to look up "bristle board" - it seems that its' the stuff used to make dartboards and may be made out of animal hair or other natural fibres, which might be quite flammable. While it's probably fine under normal conditions, it's fault conditions that you have to think about: If something overheats then it might set it on fire. If it's kept far enough away from anything that may become hot then it is not likely to be a problem. In general, be wary of using anything flammable in electronic equipment.

psiu · Post by **psiu** » Tue May 01, 2007 9:36 am

Follow SilentKev's advice, and don't just go touching stuff in it if don't have too...like heatsinks. Some of them are most definitely live and will leave your arm tingly for awhile.

I believe my opinion on grabbing a live heatsink is: YEEEOOOOOOOWW!!!

EsaT · Post by **EsaT** » Tue May 01, 2007 10:24 am

Mr Evil wrote:Not a safe assumption. Depending on the design of the circuitry, capacitors may remain charged for months after power has been removed.

Could you show me that capacitor?

No capacitor is perfect capacitor and IIRC normal electrolyte capacitors have highest leakage current of all capacitors so I don't think those can hold much charge for more than few days days.
(beside low energy density leakage current is why capacitors aren't used for long term electricity storage)

Chocolinx wrote:This is what I'm most worried about. I heard that getting shocked by a toaster can kill you becuase of only 3A going straight to the heart.

50mA of AC is enough for "re-synchronizing" heart. 3A current going through body is capable to causing severe internal burns and killing by those even if it doesn't go through heart.

Mr Evil wrote:Don't worry about it too much though: a shock from a charged PSU capacitor can be painful (did it to myself once and had a nice pair of burn marks on one finger to show for it),

Getting shock to hand and small burns are worst thing what can happen if PSU is unplugged and on non-conducting surface and you're using only one hand.

If other hand holds metal casing of PSU then it's possible for current to go from arm to arm through hearth but that kind short single pulse shouldn't cause much problems for healthy people...
In fact electric fences of cattle give 5kV+ ~5-20 Joule pulses which make you "jump" rather nicely.

Mr Evil · Post by **Mr Evil** » Wed May 02, 2007 8:42 am

EsaT wrote:
Mr Evil wrote:Not a safe assumption. Depending on the design of the circuitry, capacitors may remain charged for months after power has been removed.
Could you show me that capacitor?

No capacitor is perfect capacitor and IIRC normal electrolyte capacitors have highest leakage current of all capacitors so I don't think those can hold much charge for more than few days days.
(beside low energy density leakage current is why capacitors aren't used for long term electricity storage)

Electrolytic ones leak a lot, but they're not the only type that may be charged to a high voltage (and high-voltage electrolytics leak less due to the bigger gap between electrodes). Other types of capacitor may tend to be of relatively low capacitance, but can still be big enough to hurt, or vapourize the tip of your screwdriver.

But no matter what, it's just safer to never assume that a capacitor is discharged until you've taken steps to do it yourself.

wwenze · Post by **wwenze** » Thu May 03, 2007 6:52 am

Just wear gloves... a few mm of rubber/plastic is enough to insulate 230V, that's what I always say...

Mr Evil · Post by **Mr Evil** » Sat May 05, 2007 1:00 pm

Just out of interest, I charged up a couple of high-voltage capacitors to 25V (I don't have a high-voltage DC supply handy or it would have been a couple of hundred volts) and left them for 48 hours, to see how much leakage there is.

1) Aluminium electrolytic, 33uF: 19V
2) Polypropylene film, 6uF: 25V

As you can see, the electrolytic lost a lot of charge, but the film capacitor remained fully charged.

A couple of things to note about the test: a) The electrolytic capacitor is one that I pulled out of some old equipment which hasn't been used for many months. Both age and lack of use causes leakage to increase - a new capacitor would leak quite a bit less. b) Leakage would have been faster if the voltage was higher, i.e. it takes less time to go from 250V to 200V than it takes to go from 25V to 20V