static damage from scrubbing a CPU?

[OS]

I was scrubbing the thermal pad gunk off my XP2000+ cpu with a rag and alcohol. Will this sort of thing cause static damage? I've read about latent defects which are harder to detect.

Yeah rubbing my cpu brings up images of science fair experiments where you rub a balloon with a plastic rod and generate a butt load of static charge. :Q The CPU still boots fine and stuff.

 

[EKAtBzboyz]

as long as you didnt hear a *shock* your fine

and if it still works, then you are ok 🙂

you are not using plasic so i dont think it would even cause static electricity....plus theres alcohol involved

you just have to be careful about your feet if you are wearing socks, as sock+carpet can create lots of static electricity

 

[pm]

Since it works, it's probably ok. If you need to do it again in the future, cover all of the pins on the bottom with aluminum foil by taking a small square of foil and wrap it from the bottom around to the top leaving only the core exposed. This should prevent any static damage.

 

[OS]

Originally posted by: pm
Since it works, it's probably ok. If you need to do it again in the future, cover all of the pins on the bottom with aluminum foil by taking a small square of foil and wrap it from the bottom around to the top leaving only the core exposed. This should prevent any static damage.

Thanks Pat!

Yeah I tried not to touch the pins since that's where you'd get conduction. :Q

I've been told all it takes is 40 volts of static in the right spot to zap an IC, and that's not something you'd feel.

 

[pm]

It's substantially more than 40V, but it is true that you probably wouldn't be able to feel the zap.

The foil trick, incidently, is the way that we usually handle something like this at Intel. We have double-sided shiny foil that is thicker than kitchen foil that is used, but regular kitchen foil should work.

 

[OS]

Originally posted by: pm
It's substantially more than 40V, but it is true that you probably wouldn't be able to feel the zap.

The foil trick, incidently, is the way that we usually handle something like this at Intel. We have double-sided shiny foil that is thicker than kitchen foil that is used, but regular kitchen foil should work.

So how many volts does it take about? Or is that proprietary information? 😉

 

[pm]

I can't think that it would be proprietary - since it's a physical characteristic of the chip that we would probably want out customers to be well aware of.

But I have no idea what it would be off-hand. Let me go look it up and get back to you.

 

[OS]

Originally posted by: pm
I can't think that it would be proprietary - since it's a physical characteristic of the chip that we would probably want out customers to be well aware of.

But I have no idea what it would be off-hand. Let me go look it up and get back to you.

Thanks Pat! 🙂

 

[Viper GTS]

I always used goo gone or something similar on my CPUs, then washed them with running water (dead serious), dried them with a hair dryer, & plugged them in.

Never had a single problem.

Viper GTS

 

[pm]

Compared to a plasma etch, a blow drier is mild. 🙂

But still, ESD is a serious, important problem and users should take as much caution against ESD as they would from dropping it from a height, or from spilling (for example) Pepsi all over it. It takes surprisingly little static charge to disable a chip since static voltages are extremely high. And the problem with ESD is that it often renders the part merely unstable - not dead. In the labs at Intel, not wearing a grounding strap is likely to get an employee a sstern lecture and possibly attendence in an ESD precaution class.

In one case I personally destroyed a unique prototype processor by reaching out and take it from a coworked while we were walking down the hallway at Intel. It worked before the hand-off. When I touched it, I felt the spark, it didn't work 5 minutes later. Then some time later, I watched a fellow engineer disable a multi-million dollar defect detection system by ignoring all of the "ESD Protection Required!" signs and loading the part without using a grounding strap. You could hear the spark from several feet away and then I watched a whole array of diagnostic lights come on. It took the better part of a day and a half for a technician to fix. From personal experience, I have learned to respect static electricity's ability to destroy semiconductor components .

As far as what the limits are, I had a hard time finding them. I am not sure why they are not in the databook or the design guides... Anyway, the details of the ESD testing methods for some older processrs are here. The details of the newer ones are apparently not publicly available. Summarizing the link above, the standard method of test is the "Human Body Model" (HBM) which models the static electricity of a person by by a 100 pF capacitor discharged through a switching component and a 1.5kW series resistor into the component. Intel, according the older document reference above, used to (and presumably still does) test to the 2000V HBM with plenty of extra margin (ie. 4-8kV HBM).

Patrick Mahoney
Microprocessor Design Engineer
Intel Corp.

 

[everman]

Originally posted by: pm Then some time later, I watched a fellow engineer disable a multi-million dollar defect detection system by ignoring all of the "ESD Protection Required!" signs and loading the part without using a grounding strap. You could hear the spark from several feet away and then I watched a whole array of diagnostic lights come on. It took the better part of a day and a half for a technician to fix. From personal experience, I have learned to respect static electricity's ability to destroy semiconductor components .

:Q

 

[DaFinn]

...plus theres alcohol involved

Hehe,

personally I have learned that alcohol and computers dont mix well... 🙂


-DaFinn