Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I'm curious, if all the hardware gets power from the PSU's DC power lines, which provide a max. voltage of 12V, can you get electrocuted by touching some of the hardware?

share|improve this question
7  
For safety reasons, you might still want to turn off everything. Power spikes, lightning, malfunctioning equipment and such things can get you electrocuted. Better to avoid it even if it sounds safe than to be sorry. I had a electrical fly killer electrocute my entire right arm, and I thought it was entirely safe to hold it by the net... –  Tom Wijsman Oct 9 '11 at 19:32
3  
9 V is enough to kill you. –  stardt Oct 10 '11 at 1:14
1  
Except in extraordinary circumstances, a voltage in excess of about 50 volts is needed to cause electrocution. Not to say that such circumstances couldn't occur inside a running PC, but it would be a stroke of extraordinarily bad luck. –  Daniel R Hicks Oct 10 '11 at 2:16
2  
The darwin awards have proven anything is dangerous and fatal in the wrong hands. If you don't know what you are doing leave it to the professionals or take a few electronics classes. –  Moab Oct 10 '11 at 3:49
2  
Nitpick: electrocution means being killed by electrical shock. You can be shocked multiple times. You can only be electrocuted once. –  Steve Fallows Oct 11 '11 at 13:45
show 2 more comments

6 Answers

up vote 31 down vote accepted

The short answer is no, 12 volts isn't high enough and the 24 rail to rail voltage available is from an unbalanced pair of connections, -12V is pretty limited on current output.

The long answer is yes, a faulty power supply with a bad ground connection (third prong) can kill you. Older machines ran mains power up to the on/off switch and a hot wire laying over a sharp edge could put full mains power on the chassis. But then the system doesn't have to be open, just touching the metal case would do it.

While you hear about the occasional electrocution from computers in the news (bad ground circuits), the likelyhood of working on an open system harming you is extremely low to nil. Now where you get into trouble is working with old CRT monitors. The flyback transformer system output starts at about 8.000 volts and the tube is a neat Leyden Jar, just waiting to bite when fingers get into the wrong areas and kill when the current travels from arm to arm. Having taken 80kV from a General Motors HEI ignition system, I can tell you from experience you don't want to do this. LCD backlight systems also are another area to avoid as you're dealing with 90-400V depending on the system.

Protecting the system from you often is more likely to be an issue. Having drawn a one inch static spark on a motherboard once and not having anything happen for me was a stroke of luck. Ground straps are probably a good idea.

Update on human physiology. Grab an ohm meter with autoranging and perform the following experiment. With dry hands, put it on a fairly high setting with a lead in either hand and the test probes in contact with the backs of your hands. Note the resistance. Now hold the probe tips between your fingers and slowly squeeze. Get some salt water, soak your hands and hold the probe tips in your fingers. Note now the resistance changes.

Most people don't understand Ohm's law: voltage divided by resistance = current. So a 5 volt power supply capable of 50 amps is only able to push 0.0025 amp through a 2000 ohm resistance. 110 volts can push 0.055 amp through 2000 ohms and 1000 volts can push 0.5 amp through 2000 ohms. The general test for body impedance is from arm to arm (across your heart). People have died from as low as 32 Volts.

The usual accepted lethal current is 0.100 amp (100 milliamps) and is determined by voltage applied, skin resistance and the path it takes through your body in an attempt to find its opposite polarity or ground. Basically the rule is 100mA for fib and 200mA for total heart contraction.

And the reason static from the carpet doesn't kill you is because while you have a tremendous voltage buildup, there isn't enough charge there (quantity of electrons) to carry a sustained current, power supply systems on the other hand are designed to do this.

share|improve this answer
1  
+1 for mentioning CRTs, they can really hurt you. –  Handyman5 Oct 9 '11 at 23:11
4  
Human physiology. Grab an ohm meter with autoranging and perform the following experiment. With dry hands, put it on a fairly high setting with a lead in either hand and the test probes in contact with the backs of your hands. Note the resistance. Now hold the probe tips between your fingers and slowly squeeze. Get some salt water, soak your hands and hold the probe tips in your fingers. Note now the resistance changes. –  Fiasco Labs Oct 10 '11 at 18:59
3  
The problem with most people is that they don't understand Ohm's law. Voltage divided by resistance = current. So a 5volt power supply capable of 50 amps is only able to push 0.0025 amp through a 2000 ohm resistance. 110 volts can push 0.055 amp through 2000 ohms and 1000 volts can push 0.5 amp through 2000 ohms. The general test for body impedance is from arm to arm (across your heart). People have died from as low as 32 Volts –  Fiasco Labs Oct 10 '11 at 19:06
3  
The usual accepted lethal current is 0.100 amp (100 milliamps) and is determined by voltage applied, skin resistance and the path it takes through your body in an attempt to find its opposite polarity or ground. Basically the rule is 100mA for fib and 200mA for total heart contraction. –  Fiasco Labs Oct 10 '11 at 19:14
3  
And the reason static from the carpet doesn't kill you is because while you have a tremendous voltage buildup, there isn't enough charge there (quantity of electrons) to carry a sustained current, power supply systems otoh are designed to do this. –  Fiasco Labs Oct 10 '11 at 19:19
show 10 more comments

One tip if you happen to be working in an older (pre-ATX) computer - The wires going to and from the switch on the front of the case are carrying the same voltage as the electrical outlet. I shocked myself several times brushing up against the connection between the switch and the wire when the plastic guard had gotten dislodged.

I believe this goes back to the original PC-AT power supply design with the switch integrated into the supply. When they decided to move the switch to the front of the case, the simplest thing to do was to run the 120/240V power to the switch and then back to the power supply.

share|improve this answer
3  
You probably shocked yourself. Electrocution (in the proper use of the word) is usually fatal. –  Paperjam Oct 9 '11 at 19:32
2  
Excellent point @Paperjam, I have corrected my poor usage. –  dsolimano Oct 9 '11 at 19:39
2  
@Tom, the power supply was not broken, that's just the way they were wired back in the day - references added. –  dsolimano Oct 9 '11 at 19:40
1  
+1 Thank you for the references. –  Tom Wijsman Oct 9 '11 at 19:43
1  
The AT-standard only allowed for "hard" power switches, e.g. the switch physically cut the mains supply to the PSU like a light switch. The ATX standard defined a "soft" power switch that allows the motherboard to control the PSU power by supplying the proper voltage on the control line. The downside is that the PSU is required to provide a small line that is constantly on - to allow the computer to power up. The plus side is that you can shutdown/restart/powerup the computer in software. –  crasic Oct 10 '11 at 8:30
show 2 more comments

Yes, there is a chance to be injured

1. You can get burned badly by a CPU. (especially laptops)

2. Direct contact (for instance if a sharp a solder joint penetrates the skin allowing good contact & low resistance) with as little as ~20mA can stop or cause fibrillation of the heart. (I once touched a 12v solenoid and it caused me to accidentally hit my face with my wrist hard :D )

Assuming that the power supply is 100 Watts for each rail:

100w / 12v  = 8.3a
100w / 5v   = 20a
100w / 3.3v = 30.3a

Assuming that the power supply is 600 Watts for each rail:

600w / 12v  = 50a
600w / 5v   = 120a
600w / 3.3v = 182a

...

Table from: http://www.allaboutcircuits.com/vol_1/chpt_3/4.html

BODILY EFFECT     DIRECT CURRENT (DC)    60 Hz AC     10 kHz AC
--------------------------------------------------------------- 
Slight sensation     Men = 1.0 mA         0.4 mA        7 mA 
felt at hand(s)    Women = 0.6 mA         0.3 mA        5 mA 
--------------------------------------------------------------- 
Threshold of         Men = 5.2 mA         1.1 mA       12 mA 
perception         Women = 3.5 mA         0.7 mA        8 mA 
--------------------------------------------------------------- 
Painful, but          Men = 62 mA           9 mA       55 mA 
voluntary muscle    Women = 41 mA           6 mA       37 mA 
control maintained                                           
--------------------------------------------------------------- 
Painful, unable       Men = 76 mA          16 mA       75 mA 
to let go of wires  Women = 51 mA        10.5 mA       50 mA 
--------------------------------------------------------------- 
Severe pain,          Men = 90 mA          23 mA       94 mA 
difficulty          Women = 60 mA          15 mA       63 mA 
breathing                                                    
--------------------------------------------------------------- 
Possible heart        Men = 500 mA        100 mA             
fibrillation        Women = 500 mA        100 mA             
after 3 seconds                                              
--------------------------------------------------------------- 

...

Ever licked a 9v battery to test it? Low voltage can be quite shocking ;-)

...

Quoting Fiasco Labs

Human physiology. Grab an ohm meter with autoranging and perform the following experiment. With dry hands, put it on a fairly high setting with a lead in either hand and the test probes in contact with the backs of your hands. Note the resistance. Now hold the probe tips between your fingers and slowly squeeze. Get some salt water, soak your hands and hold the probe tips in your fingers. Note now the resistance changes. – Fiasco Labs

3. Accidental contact with fan (I did this before it HURTS!)

share|improve this answer
1  
Thanks for posting that chart, I had one in one of my old Electronics textbooks and was wishing to add it. As to the relay, another effect most people don't know about is that the pull-in coil has a self-induced voltage spike when the current is cut off to it and its field collapses. If you're driving this relay with a solid state circuit, you need to place a diode across the coil (reverse polarity) to effectively short the coil to itself so the voltage spike doesn't get back into your driver transistors and smoke them. –  Fiasco Labs Oct 10 '11 at 21:11
    
you're welcome, i remember hearing using a diode on inductive loads too (I wonder if speakers do it too) –  linuxrules94 Oct 13 '11 at 0:57
add comment

A spinning fan or sharp edge inside the case are more common "dangerous" concerns within a computer :)

share|improve this answer
2  
i know that) i have once broke a blade of a cpu's fan. instead, my finger was perfectly fine =) –  develroot Oct 10 '11 at 12:28
    
Had one of those first high speed ball bearing heat sink fans that run about 5500 rpm (whiny thing) on an Athlon processor. It drew blood as I saw the blade depart the case and head across the room. Thankfully it had an even blade count, so I snapped one off on the other side and filed the stubs down till both were in balance. It ran for years that way, and I swear just enough quieter to make it tolerable. Sharp case edges are another matter. I had a scar trail down my arm from a washing machine. All these things are built by a coining process that leaves razor sharp micro sawblade edges. –  Fiasco Labs Oct 13 '11 at 1:30
add comment
  1. You can get burned on a hot heat-sink.
  2. You can get a finger nicked on a fan blade.
  3. If you're wearing any jewelry you can get a ring or watchband or what-have-you between the main 5V supply and ground. There is enough current there to cause serious burns, perhaps enough to cause loss of a finger if, eg, your ring gets heated to red hot. (In general, modern desktop power supplies are sufficiently current-limited that the chance of serious injury is minimal, but in a high-powered box the 5V supply is still delivering enough current to produce this sort of hazard. It is REALLY a hazard in "big iron" boxes where hundreds of amps are available, and where safety rules demand that all jewelry be removed before working inside.)
  4. If you're wearing any jewelry you can get a ring or watchband or what-have-you between two points in the circuit and burn out a $500 component, after which you'll kill yourself.
share|improve this answer
add comment

You have asked the question which is a good reason to leave well alone. There are parts that could be trashed if you touch them (static), or if you drop something on them. Power supplies usually come with an extra level of metal shielding, but really, call in the repair man.

share|improve this answer
5  
I'd -1 if I had enough reputation. If everyone followed this advice, there would be no repair men (or women). Learning to service the insides of a machine is best done by experience. The inside of a PC is basically not dangerous to a healthy adult human, even with power running through it, but of course it's safest to turn it off and unplug it. –  Zach Oct 9 '11 at 22:04
2  
Hi Zach. Enough gung-ho individuals thought as you do to demote my answer anyway. I am concerned that the original poster seemed like a rookie and was not asking for help from a person but from the web. I don't want that kind of person in my computer room taking the back off any of my servers without a more experienced person guiding him. No drivers on the road without tuition and L'plates thank you! –  Paddy3118 Oct 10 '11 at 6:55
    
Fair enough... if the OP was someone who would be working on my machines, I would not want him or her asking this question. But if he or she is a hobbyist at home just curious about taking apart and working on computers, then I'm all for it. –  Zach Oct 10 '11 at 21:48
    
+1 for overly zealous -1'ers with no explanations. –  HaydnWVN Nov 28 '11 at 10:12
    
@Zach - Yeah, if someone was asking this over in Server Fault, I'd be worried. –  Fiasco Labs Mar 9 '13 at 2:47
add comment

protected by Journeyman Geek Mar 3 '13 at 5:51

Thank you for your interest in this question. Because it has attracted low-quality answers, posting an answer now requires 10 reputation on this site.

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.