Shipping computer components in foil

[aaron]

Okay so I asked this a while ago on a more computer related forum and most people thought it was a stupid idea but I'm skeptical. I still lean towards the "It's a good idea, as long as ..." side. I'm definitely not saying I'm right though, I just wanted a few more opinions.


Just say I'm shipping RAM, or a CPU. I don't have any anti-static bags and I can't source any. Is there anything wrong with carefully wrapping hardware in foil to protect it from ESD?

Please note, of course if it's a motherboard or any type of board/device that has an internal power supply, it would have to be removed for obvious reasons.

A second concern is (and this is once again more relevant to a motherboard for example) if the foil is pressed onto the back of a device which has many sharp solder through-hole points on the back, in transit foil my catch, pull off and stick to the joints. This could cause a short/damage once the device is powered up if not caught.


However, for a CPU or some sticks of RAM, wrapping them in foil like a nice sandwich would be okay, right? If not, why? 

[aaron]

So it would be a good idea to wrap the ram in foil? As long as it's neat, maybe a few layers thick for rigidity?

[XOIIO]

I'm thinking since they pack IC's with all the pins in electrostatic foam, maybe you should try and make sure all the pins are in contact with the aluminum, to make sure there isn't any potential for build up?

[coppice]

It's okay, but you do not have to do that. CPU is quite ESD insensitive, unless you have a direct 10+kv HBM discharge on it. CPU rubbing its packaging can not generate enough energy to breakdown it.

Mobo is usually quite tough as well. RAM is NOT. They must be packages carefully.

Since the pin protection specs for I/O pins on most complex ICs are similar, why would you expect a RAM to be far more susceptible than a CPU? Also where do you find these CPUs that would rub against arbitrary packaging materials in an arbitrarily dry climate and not suffer damage?

[coppice]

Since the pin protection specs for I/O pins on most complex ICs are similar, why would you expect a RAM to be far more susceptible than a CPU?

I do not know, but I handle computer parts equally for the past decade of building custom rigs, and I found the failure rate of RAM, both mfg defects, and defects caused by use/reinstall/transfer is the highest.

Also where do you find these CPUs that would rub against arbitrary packaging materials in an arbitrarily dry climate and not suffer damage?

I never seen any CPU fails besides being cooked or voltage got cranked too high. In my hometown, PC dealers pack CPUs in 5* or 10*pack, bundled with rubber bands.

RAMs just fail more because people abuse them more. The need to drop a big square object accurately into a square hole, and the messiness of putting thermal goo on it, limits the amount of handling a CPU typically gets. People really love polishing the edge connectors of RAM sticks on their sleeve, for some reason.

If a production line is finding a significant failure rate in some semiconductor part, EOS (electrical over stress) is the reason 9 times out of 10. Sometimes the EOS issue is a design problem (e.g. a startup surge), but much of the time it comes down to poor handling. Its often seasonal - the dry season comes along, and production failures skyrocket. If you live in a humid place you can get away with a lot of things, but don't expect that to be true for everyone. Just because a part works, that does not mean it has not been damaged. If you look at micrographs of poorly handled, yet still functional, devices you can often easily see discharge marks. Those parts can have much higher failure rates in the field.

EOS in ICs is still a huge problem, and protection is still the subject of continuous research.