Sign in with
Sign up | Sign in
Your question
Solved

Static damage to the CPU

Tags:
Last response: in CPUs
Share
September 11, 2012 1:31:00 AM

I cleaned the dust out of my PC two days ago (not much dust thanks to my demciflex filters). Now I'm worried that I may have done some damage to it because I did not take too many anti-static precautions (no wrist strap for example). My fingers never touched the motherboard or anything, just the bare metal case on the inside and the plastic parts of the fans. Also used a damp paper towel to clean some of the bare metal. But I think I may have went too far when I used a damp Qtip (100% cotton with a plastic tubular body) to clean off the slightly dusty heatsink. I'm afraid that may have been enough to transfer some harmful volts to the CPU, shortening its life. It's a Core 13 2120.

I know it's hard to tell if you will have latent ESD failure in the future, but I've tried testing my CPU by running it at full load. I figured that if I damaged an IC, maxing out the CPU would reveal it because that circuit would have to be used. The result, the CPU performed at full load for quite a while with no issues. Do you think that this is a scientifically valid way to test for ESD damage?

Further, does anyone know if this CPU has built-in ESD protection and if it does, the ESD protection rating? And do you think my actions were potentially harmful to the CPU or am I overreacting?

More about : static damage cpu

a b à CPUs
September 11, 2012 1:49:11 AM

If it works, there's no problem....

You'r overreacting a bit. A Q-Tip is made out of plastic, which can't conduct electricity,.
m
0
l
a b à CPUs
September 11, 2012 1:53:05 AM

This seems like severe paranoia, the CPU is perfectly fine. If it boots up then you haven't damaged it via ESD.
m
0
l
Related resources
September 11, 2012 1:59:49 AM

I don't think you could even damage the CPU with ESD, just by touching the heatsink... You would need to touch the pins.
m
0
l
September 11, 2012 2:00:16 AM

There is a thing called ESD latency where a CPU will still function with static damage but its life will be shortened. It could die in two weeks.

It's explained here, "http://www.ce-mag.com/archive/02/Spring/dangelmayer.htm..."

However, the sources and studies that are cited are outdated. I am hoping that built-in ESD protection has improved since then, kind of like how Moore's law improves CPU power. I've been searching for the ESD protection ratings of Intel CPUs but have had no luck.
m
0
l
a c 116 à CPUs
September 11, 2012 2:05:32 AM

ammonitida said:
I know it's hard to tell if you will have latent ESD failure in the future, but I've tried testing my CPU by running it at full load. (...) Do you think that this is a scientifically valid way to test for ESD damage?

No. If the ESD has caused any damage, it would be by breaking down insulating material within the IC. This damage could manifest itself as a catastrophic failure, as a non-fatal increased current draw in the affected transistors/traces, as an eventually fatal failure later or not at all.

ammonitida said:
Further, does anyone know if this CPU has built-in ESD protection and if it does, the ESD protection rating?

CMOS transistors have an intrinsic body diode which would clamp ESD to Vcc/ground so the likelihood of damage while the CPU is in-socket would be very low.

BTW, a wet surface (damp Qtip/towel) would be unable to build up a static charge in the first place and since you touched the "bare metal" which I presume means IHS then any ESD you might have built up would have been "grounded" to the IHS.
m
0
l
a b à CPUs
September 11, 2012 2:10:31 AM

You're over thinking this. Basically, if it still works you have nothing to worry about. Worry, when something actually stops working. Besides, if you did shorten it's life, what can you do about that now? Basically nothing.
m
0
l
September 11, 2012 2:21:55 AM

InvalidError said:
No. If the ESD has caused any damage, it would be by breaking down insulating material within the IC. This damage could manifest itself as a catastrophic failure, as a non-fatal increased current draw in the affected transistors/traces, as an eventually fatal failure later or not at all.


CMOS transistors have an intrinsic body diode which would clamp ESD to Vcc/ground so the likelihood of damage while the CPU is in-socket would be very low.

BTW, a wet surface (damp Qtip/towel) would be unable to build up a static charge in the first place and since you touched the "bare metal" which I presume means IHS then any ESD you might have built up would have been "grounded" to the IHS.


Well, that's good news about the wet surface issue. That damp Qtip cleaning was especially bothering me. One more concern I have is that my finger got close to the exposed circuit board of the HDD (and the wifi card). I recall that even close but non physical contact can transfer volts to a component, potentially causing an ESD failure. Something to do with a charged object (your body for example) generating an electrostatic field. I got this information from an old Apple training video called "The Shocking Truth". Still true with modern hardware?
m
0
l
a c 116 à CPUs
September 11, 2012 3:28:03 AM

ammonitida said:
Still true with modern hardware?

Most likely still is for the most part, except that today's devices are far more rugged than late-80's CMOS.

Another contributing factor is that almost nobody (other than electronics hobbyists who assemble their own PCBs) handles "raw" chips anymore. Today, CPUs are on substrates with decoupling capacitors, RAM chips are on DIMMs with decoupling capacitors, everything else is mounted on a board of some kind with decoupling caps before making it into customers' hands. All of these provide ample trace and decoupling capacitance to sink trivial amounts of ESD which makes them far less likely to get damaged than loose chips.
m
0
l
a b à CPUs
September 11, 2012 5:16:55 AM

I have to agree with the other posters. You are way overreacting.
I could sort-of understand if it was not working correctly but even then I would first guess you knocked something loose.

Also you said you did alot of touch of the metal case which would ground any static discharges.

Also you claim not to have touched the board making it even less likely
m
0
l
September 11, 2012 4:59:30 PM

smeezekitty said:
I have to agree with the other posters. You are way overreacting.
I could sort-of understand if it was not working correctly but even then I would first guess you knocked something loose.

Also you said you did alot of touch of the metal case which would ground any static discharges.

Also you claim not to have touched the board making it even less likely


Except that the case was only plugged into the surge protector and that surge protector was not plugged into the wall.
m
0
l
a b à CPUs
September 11, 2012 9:05:09 PM

You're overreacting. Even at full load, nothing happened. 99% percent chance it doesn't fail, and if it does, RMA it.
m
0
l

Best solution

a c 116 à CPUs
September 12, 2012 3:45:33 AM

ammonitida said:
Except that the case was only plugged into the surge protector and that surge protector was not plugged into the wall.

That does not really matter. By touching any metal part bound to chassis ground, you equalized the voltage between you and the case and from that point on, you are effectively grounded (common floating ground) until you break physical contact and either you or the case increase or dumps its static charge.

The electrical/earth ground simply provides a universal reference voltage for everything but is not absolutely necessary. Sharing an isolated/local ground with the workpiece(s) is sufficient ESD-wise. Planes aren't grounded to electrical ground but the whole airframe acts as a local ground for everything that needs grounding and this is enough to successfully protect most planes from the mother of all ESDs: lightning strikes.
Share
September 13, 2012 3:40:46 AM

Best answer selected by Ammonitida.
m
0
l
!