What would happened if you pulled the processor out while pc was on?

panyan · 15 Jul 2011 at 23:52

Phenomenologica said:
DC electricity requires about 10,000 volts to arc a centimetre, so no significant arcing is going to occur at 1.4 odd volts. Also, good advice when dealing with electricity if you don't have gloves is only use one hand. That way the power will either pass through a couple of fingers or down the side of your body to ground. If you use two hands and bridge something the current can travel straight across your chest, and your heart happens to sit there. But nothing deadly is going to come out of a CPU socket.

+1 the risk of shock is almost nil

madindehead · 16 Jul 2011 at 01:51

I think you would be surprised at how low a current needs to be for it to actually be fatal

http://www.physics.ohio-state.edu/~p616/safety/fatal_current.html

jonny512379 · 16 Jul 2011 at 02:59

madindehead said:
I think you would be surprised at how low a current needs to be for it to actually be fatal

http://www.physics.ohio-state.edu/~p616/safety/fatal_current.html

but that is with AC electric
The little i know says DC = OK (unless stupid voltage and current) AC = death possible!

jonny512379 · 16 Jul 2011 at 03:02

Greboth said:
For those asking why? The correct question to ask is - Why not? And for those saying about getting a shock, just insulate yourself from earth and problem solved. If I had a HSF lying around I would be tempted to try it on my old athlon just to see what actually does happen, but I don't so I can't.

Hehehe wont the OPs youtube video look silly when he is standing in wellington boots on a pile of books to insulate himself from a CPU

jonny512379 · 16 Jul 2011 at 03:03

OLDPHART said:
How do peeps think of these ideas.

bet they said this about Einstein

OLDPHART · 16 Jul 2011 at 11:30

Einstien is a little way above this chap.

Theres a big difference hot swapping a HD > GPU to a CPU. The OP does not even know how much juice goes through a CPU, hence he thinks he will get shock.

marc1024 · 16 Jul 2011 at 13:54

OMG i didnt say i would do it! please qoute me saying i would do it (excluding the time where i said i would do it if i got a free pc with the same spec)!

WingZero30 · 16 Jul 2011 at 14:03

Dervious said:
I still find it slightly incredible that a modern CPU can draw 100amps. That's a lot of current, even if it is shared between several million transistors.

drunkenmaster said:
so 100amps at 12v is a lot more dangerous than 0.5amps at 240v.

Hang on!!

If a PSU 12V rail is rated to supply 25Amps or two 12V rails with 25Amps each as in the case of my PSU, then how can it supply 100amps on 12V rail(s)? :confused:

namnoc · 16 Jul 2011 at 14:58

In 1987 I did this with an ARM 32016 chip to move it from the host system to a live prototype subsystem and it (eventually) worked fine. Of course the socket was engineered for the purpose, power lines and clock signals were synced between the systems and we had spend 6 weeks planning it. Still quicker than the alternative of designing the support chips for the test system though.

Game · 16 Jul 2011 at 15:56

more to the point why would anyone even ask this question

MagicBoy · 16 Jul 2011 at 16:00

System would stop responding. Simples!

marc1024 · 16 Jul 2011 at 23:06

Game said:
more to the point why would anyone even ask this question

curiosity. I have mind that can ask very stupid questions to very interesting and crazy questions. People are actually struggling to agree on what would happen so, i think somebody has to actually do it with a new age processor (on a socket 1155 or something) before we can know for sure what will happen.
All we know it will cause the computer to stop responding.
Will there be a black screen, would the screen freeze on the frame its on,
would it kill the person taking it out?
would there be electrical arcs between the motherboard and processor?
would the pc blow up?
would the CPU work after being removed from a hot system? (hot as in being powered on)

Jony27 · 16 Jul 2011 at 23:18

curiosity killed the cat

jak731 · 16 Jul 2011 at 23:30

I think some multi socket server boards support hot swapping of cpus, upgrade with no downtime

andy_mk3 · 16 Jul 2011 at 23:36

I have an old AMD althlon mobo and cpu somewhere at home, I'll get it up and running on XP and try it

MagicBoy · 16 Jul 2011 at 23:41

I've done the equivalent on a testbed P2 system at work years ago when I knocked the slot-type CPU by accident. Screen went black then the display powered down after a few seconds. Nothing blew up. System worked perfectly when the CPU was re-inserted.

That PC got all sorts of abuse, even shoving blanking plates in the PCI slots didn't phase it. Some plank even shoved an Athlon CPU in it by accident - it sort of fitted when inserted back to front and given a shove. The Athlon CPU blew, the rest of the system was right as rain.

The voltages are tiny and the amps are not enough to hurt anyone. The biggest risk is burning your finger on a hot component. It's very unlikely the pulling a CPU or any other component would cause the PC to "blow up" or damage any other components. What's the difference between pulling the CPU out and turning the system off at the wall while it's running? Very little.

ScratchFive · 17 Jul 2011 at 00:05

WingZero30 said:
Hang on!!

If a PSU 12V rail is rated to supply 25Amps or two 12V rails with 25Amps each as in the case of my PSU, then how can it supply 100amps on 12V rail(s)?

because its converted to 1.4 or whatever your running your cpu at, in the same way as you have 100A through your cpu and 0.58A from the wall, the 12v rail is a mid stage where you would have 140/12 = 11.7 amps (P = IV) going through your rail.

also, people talking about shock risks; you are indeed right that its current that kills, but its current flowing through your body (a fixed resistance) and as V = IR its really the voltage you have to worry about! i.e you cant have a large current through a large resistance without a large voltage...

now... finger to finger (easiest to measure...) I am about 1 MOhm, this means at 240v there would be a current going through me of 2.4*10^-4A, as I understand, not enough to kill me, but enough to hurt like hell (please understand that i measured this resistance with mulitimeter probes so there was minimal skin contact, I am in no way saying if you stick your finger in a socket you will survive!)

however at 12v (no matter whether 25A can be pulled from this socket) i will only have 1.2*10^-5A running through me; significantly less, and it will be 1.4*10^-6A at 1.4v even though you could be pulling 100A through your CPU at the same time from the same source

hope this makes some kind of sense

strumpusplunket · 17 Jul 2011 at 00:15

Oh my God! Don't try this, whatever you do! You will more than likely be pulled into the digital world through the electron transfer across the CPU/board divide. As a User, you might prove capable of manipulating the laws of the digital world, but you'd be in serious danger from the Master Control Program.

jonny512379 · 17 Jul 2011 at 01:24

strumpusplunket said:
Oh my God! Don't try this, whatever you do! You will more than likely be pulled into the digital world through the electron transfer across the CPU/board divide. As a User, you might prove capable of manipulating the laws of the digital world, but you'd be in serious danger from the Master Control Program.

WOW! get my bike ready, i'm going in!!

panyan · 17 Jul 2011 at 01:24

Schrödinger's Computer:

From what i understand, in the CPU is a vial of poison and the silicone is naturally radioactive, when the cpu is in the board and the machine turned on, we cannot know what state the computer is in - and by default - must also be in both states at once. By removing the cpu, we have essentially disrupted The Copenhagen Interpretation and by doing so - have limited the state in qhich the computer is in. It has gone from being in simultaneous states, to being in a singular state.