Computer upgrade - Static (OT maybe)

[Bob Eager]

Yup, I still have a 512B memory card!

[Andy Champ]

.. and in between the two is to roll your sleeves up, and keep a bit of bare arm on the chassis.

I'm in favour of pulling the plug too. You just want to make sure that the first contact is you to the case, and you maintain that contact.

Andy

[Steve Walker]

Doesn't really matter, but won't hurt to.

[John]

What about unpacking the memory - it may have a charge that is different to mine and the case. How is this equalised?

[Clint Sharp]

In message , John writes

Sound exactly right to me. Make sure the mains switch is off and hold the chassis when you need to handle the components or touch the motherboard.

>

[Clot]

Via your hand holding it and another part of your body grounded to the chasis.

[Andy Champ]

The container should be ever-so-slightly conductive. So pick up the container, and take the RAM out of it, DON'T put the RAM down and it'll be fine. Then connect yourself to the chassis before touching the RAM to the motherboard, and maintain your connection (holding the chassis, or touching it with an elbow is fine, but requires discipline). If you feel an urge to put the RAM down anywhere, put it back in the packet - picking the packet up before touching the RAM to it - or put it down on a spare bit of the chassis *while touching the chassis*.

Andy

[John]

I'll wear my shorts and put it on my knee.

[John Rumm]

A common misunderstanding is that static damage results in fatal component damage. Research in the defence and aerospace industries shows that is only true in a little as 20% of the cases. The other 80% of the time there will be damage caused that does not result in immediate failure, but may manifest later as reduced component life, or other malfunction.

So the moral of the story is that 8 times out of ten you will be blissfully unaware that you have caused static damage at the time.

[Graham.]

They has DIMMs in 1967? ;-) They did have 40 series C-MOS chips 20 years ago, I seemed to destroy those by just looking at them.

[Andy Burns]

You what, with 1GB of RAM for about =A315? I remember worrying about=20

512byte chips ...

[TheOldFellow]

Now that is interesting, and to me, new information. Thanks, John. It makes a lot of sense. And goes to prove what a useful Usenet Group this is.

I'd just like to point out, saving another post, that I never said that DIMMs were around in 1967, but that I was. Old CMOS chips are another question, the OP was asking about DIMMS.

R.

[Paul Matthews]

Isn't it in electrostatic packaging?

Open chassis, get hold of chassis, place ram (in packaging) on chassis, open, install.

[Paul Matthews]

Ever wonder why hobbyists often have unreliable PCs? part of is it constantly fiddling with settings, and part poor staic precautions.

Static damage rarely manifests immediately.

P.

[The Natural Philosopher]

He did say switched off at the wall IIRC.

Read the original post.

Gosh. And there I was thinking that everybody already knew all that ..I giess spendong 30 years as an electronic engineer isn;t evrybodies expereince.

That works.

[The Natural Philosopher]

Andy has some gripe with me I think.

[The Natural Philosopher]

Slightly less than the petrol burnt to drive to the shop and buy one.

[Clint Sharp]

In message , The Natural Philosopher writes

If the OP has a wrist strap I'd suggest using it, it's only overkill if he doesn't have one, otherwise it's a sensible precaution. A bit like having seatbelts in your car but not using them because 99.99% of the time you'll be fine. If he doesn't then fine, just accept a slightly higher risk of damaging the equipment. Wrist straps are cheap and easy to get these days so if anyone's doing PC maintenance there's no excuse for not having one.

Better is to hold the case or rest a forearm on it whilst handling the components.

[Andrew Gabriel]

Likewise, having worked in the defence and aerospace industries, I'm all too well aware of this. There aren't many industries which can afford to investigate semicondutor failures. Depends on the extent of the damage of course, but a ballpark figure is that static damage (which hasn't instantly destroyed the semicondutor) increases its chance of failure during service life by something of the order 10,000 - 1,000,000 times. When a component does fail, high magnification of the junction can be quite interesting. It usually looks to have melted, and you can find nearby junctions which have also melted but are amazingly still working, which is the givaway that it was static damage. If these junctions which are still working are tested, you find they won't conform to their datasheet specs anymore. This means the component was out of spec before the complete failure, and is a typical cause of computers exposed to poor static handling procedures behaving badly, as Paul Matthews mentions.

[tinnews]

While I agree with the comments about static damage not manifesting itself immediately I'm not at all sure about "wonder why hobbyists often have unreliable PCs", do you have any evidence for this at all?

In my experience (only anecdotal but I have been working and playing with PCs since the 1980s, and in the computer industry before that) the *most* unreliable and full of glitches machines are those which have messy software not those which have been taken apart and put together lots of times.