The Filthy, Rotten, Nasty, Helpdesk-Nightmare picture clubhouse

[80251]

No one talks about the fact dust is a natural sound deadener -- the more dust the better!

 

[Jetster]

No one talks about the fact dust is a natural sound deadener -- the more dust the better!

It's also a fire hazard

 

[Athlonite]

It's also a fire hazard

but on the other hand there's nothing more silent than a burnt out PC though

 

[micropage7]

That's a Soyo case. Remember those well

soyo and soltek is one of my fav beside DFI
yeah that's old school memories

 

[Valantar]

Thought I'd post a different type of tech gore here than what's usual:

http://imgur.com/QpyNxSG

Sound on.
For reference: LGA 1155 pinout.
What you're seeing there is a CPU with VCC (aka. VCore) and VCCAXG shorted to ground.

 

[80251]

What kind of voltage does that multimeter put out in ohmmeter mode? If it's anywhere around 5V is a CPU able to take that kind of voltage?

 

[Valantar]

What kind of voltage does that multimeter put out in ohmmeter mode? If it's anywhere around 5V is a CPU able to take that kind of voltage?

The CPU was dead long before I started probing it, so no worries there. The PC was refusing to boot (not even the faintest sign of life on the motherboard), and I initially thought it had a shorted VRM as they all measured as continuous with ground. Imagine my surprise when the shorts disappeared when I removed the CPU! That led me to probing its pads, confirming it was dead.

From a bit of searching, the continuity test voltage of multimeters seems to vary a lot - a few Fluke models run in the 7-9V range, while many others are between 0.4-0.6V. And some are in between. The current limits on the output from the meter should be sufficient to avoid damage outside of the most extreme cases, but I guess it's a good idea to not run continuity testing on the pads of a known good CPU unless you know your meter outputs low voltage.

Edit: my meter's manual says it's open circuit voltage for continuity testing is 2.3-2.5V. High, but definitely not high enough to burn out anything in a CPU at near-zero current, or cause a short, especially considering that the componentry inside a working cpu would cut that voltage noticeably if it wasn't already shorted.

 

[Frick]

Someone's been to gubbdagis.

 

[Valantar]

Someone's been to gubbdagis.

Home away from home!

 

[micropage7]

the first step every buy 2nd hand stuff is open it, check it and clean it

 

[AsRock]

the first step every buy 2nd hand stuff is open it, check it and clean it
View attachment 255652

Battery's last for ages in them, lil small but dam they used to sell about $17.

 

[witkazy]

Poor little bugger...

 

[Edwired]

Thought I'd post a different type of tech gore here than what's usual:

http://imgur.com/QpyNxSG

Sound on.
For reference: LGA 1155 pinout.
What you're seeing there is a CPU with VCC (aka. VCore) and VCCAXG shorted to ground.

That ohm reading isn't exactly voltage reading

 

[Valantar]

View attachment 257065
Poor little bugger...

Wow, took me quite a while to see what that was. Poor thing :/

That ohm reading isn't exactly voltage reading

Did you have the sound off? It's a continuity test, with the black probe on ground. And it's beeping consistently when touching Vcore. I.e. there's a short between Vcore and ground on the CPU, which also read as a short between Vcore and ground on the motherboard when installed. The numbers shown by a continuity test aren't ohms, but measured voltage drop AFAIK.

 

[Edwired]

It not reading voltage drop you can see it in the multimeter screen it in ohm reading. If the V symbol was shown it would show voltage drop

As well the probes you mentioned is in the wrong way as for the diode mode you suppose to place the red probe on ground and black on the on the source in question as you have it in a way shows constant short in a wrong direction

 

[Valantar]

It not reading voltage drop you can see it in the multimeter screen it in ohm reading. If the V symbol was shown it would show voltage drop

As well the probes you mentioned is in the wrong way as for the diode mode you suppose to place the red probe on ground and black on the on the source in question as you have it in a way shows constant short in a wrong direction

The only on-screen symbols are the diode symbol (for diode/continuity testing) and the speaker symbol for the audible beeps it makes when doing said tests - there's no ohm sybol anywhere near the screen. The red probe is connected to the header that does diode/continuity, resistance and current testing, but that's simply because voltage testing (which is not the same as voltage drop testing) requires different circuitry, it doesn't preclude the continuity test from checking for voltage drop. And that is what it does, as you can put the meter in that mode across any diode (including an LED) and it will tell you the voltage drop across that diode, not its internal resistance.

 

[Edwired]

The only on-screen symbols are the diode symbol (for diode/continuity testing) and the speaker symbol for the audible beeps it makes when doing said tests - there's no ohm sybol anywhere near the screen. The red probe is connected to the header that does diode/continuity, resistance and current testing, but that's simply because voltage testing (which is not the same as voltage drop testing) requires different circuitry, it doesn't preclude the continuity test from checking for voltage drop. And that is what it does, as you can put the meter in that mode across any diode (including an LED) and it will tell you the voltage drop across that diode, not its internal resistance.

Are you testing a motherboard socket while it powered on? Or testing the cpu for short as it isnt going to give a reading due to low resistance that the multimeter can't read below a certain value range?

 

[Valantar]

Are you testing a motherboard socket while it powered on? Or testing the cpu for short as it isnt going to give a reading due to low resistance that the multimeter can't read below a certain value range?

CPUs are quite low resistance, but not low enough to read as a dead short in continuity mode (a GPU would though). And no, I'm obviously not doing continuity testing on a powered-on motherboard - how would that even work? Testing was done as the system was dead, with zero signs of life.

 

[lexluthermiester]

View attachment 257065
Poor little bugger...

That is crazy!

 

[Edwired]

Unless he swiping all the grounds at the same time that is not the way to test it

 

[Valantar]

Unless he swiping all the grounds at the same time that is not the way to test it

What? I have the black probe on a ground pad. All ground pads are commoned together across the package, so it makes no difference which one of them is being measured against - they all connect to the same ground plane. You do absolutely not move around your ground probe when testing for continuity - you find a ground point and use that as your reference. Finding another and testing against that afterwards is obviously a good idea - which i also did. Then again, all the VSS (ground) pins also being continuous to ground kind of removes the usefulness of that.

I linked the pinout of the socket in that post (mirror it to get the layout for the bottom of the cpu), which will show you which rails are being measured. What the video shows is that the meter beeps on every single pad in each bottom column, despite two out of three columns in that area being VCC (Vcore). Every third column is VSS (ground). Yet they are all continuous with ground.

 

[Edwired]

What? I have the black probe on a ground pad. All ground pads are commoned together across the package, so it makes no difference which one of them is being measured against - they all connect to the same ground plane. You do absolutely not move around your ground probe when testing for continuity - you find a ground point and use that as your reference. Finding another and testing against that afterwards is obviously a good idea - which i also did. Then again, all the VSS (ground) pins also being continuous to ground kind of removes the usefulness of that.

I linked the pinout of the socket in that post (mirror it to get the layout for the bottom of the cpu), which will show you which rails are being measured. What the video shows is that the meter beeps on every single pad in each bottom column, despite two out of three columns in that area being VCC (Vcore). Every third column is VSS (ground). Yet they are all continuous with ground.

What are you trying to prove in a concept of what you just said?

 

[Valantar]

What are you trying to prove in a concept of what you just said?

What? I honestly do not have a clue what you're asking. I'm simply explaining the very basic methodology behind checking for shorts to ground with a multimeter.

 

[Edwired]

To be honest it isn't exactly clear the way you said it and shown the video. You should have two exact cpu side by side comparing the possible dead CPU to the working cpu. But you didn't do that so pointless

 

[Ferrum Master]

Well he has a point. It was a pretty pointless stuff...

It just amuses me that you have time for that. Not the technical side of it, that doesn't matter here at all...

But the frog... @R-T-B was it your relative?