Slashdot Mirror
How Serious is Static Electricity?
seanadams.com asks: "My company is considering the purchase of a small surface-mount assembly line so we can do our manufacturing in-house, and the issue of static control has come up. We've all been told to take ESD precautions when handling electronics, but how much precaution is enough? Obviously we plan to do the easy stuff like making sure that equipment and work surfaces are properly grounded. However, many shops go even further - conductive shoe straps, wrist staps, special flooring, humidity control, etc. The SMT equipment vendor says that it's unnecessary, and I would tend to agree. I've handled tons of electronics over the years and have never been able to attribute a single failure to ESD damage. Granted, Silicon Valley is a fairly humid area so that may be a contributing factor. Has the ESD threat been blown way of proportion by the guys who sell those little grey bags?"
Basically, you don't have problems with static when your only dealing with a small sample, but increase the sample size, you are increasing your chances of causing damages to something within that sample. So, for a manufacturing operation, I'd say it is worth it.
Perform a benefits/loss analysis. If you handle extremely expensive equipment that can't be damaged, then maybe it would be wise to invest a little more into ESD protection. On the otherhand, if you're manufacturing small boards with little value, then don't invest in ESD protection. You said it yourself, that you have NO cases of ESD damaging electronics as long as you've been there. If one part per million gets damaged and it doesn't cost much to replace, then forget about ESD protection.
ESD is a very real problem. Don't trust a conveyor belt salesman who says it is not. He will say anything to get the product sold. If that conveyor belt is made of something non-conductive, plastic bins placed on it will generate static. Soles of shoes, if not grounded with a heelstrap geneate static. Once at work, My heelstrap had slipped off. I found this out when I touched my workbench, and got a nasty shock. If I had touched any electronics, they would have been toast! Fortunately, I did not have a system on the bench at that time.
How ya like dat?
Just a note here... unless you can actually explain, with 100% certainty (or something close to it) the causes of all the other hardware failures, then you can't say for certain that any of the failures that you have observed are due to damage done from ESD. Just because a failing component doesn't sit up and shout "you shocked me with static three months ago and now I'm on my death bed" doesn't mean that a damaging static discharge didn't occur.
Additionally, note that what might be explained at a higher level as the failing of a certain component doesn't mean that the root cause wasn't a static discharge. Moreover, what if there was a ESD that you didn't notice. Just because you don't see a spark or feel one, it doesn't mean that there wasn't an ESD.
just some thoughts... -tcp
Are you serious? ESD is an extremely serious threat. You can't possibly think of setting up a commercial assembly line for electronics and not worry about things like anti-static wriststraps. Serious places install expensive conductive flooring just to mitigate the buildup from walking around.
Why are you asking Slashdot about this? Ask people who know about this.
--
Twoflower
I worked for many years as a contractor at a large US-based defense company. They make lots of neat useful things, and some nasty things.
One thing was always stressed, in the hardware departments, in the software departments, in the finance departments, wherever. If you go into a lab, you must have ESD training. At least 3 levels of training existed. Level 1 was little more than awareness training. If something had an ESD warning label, stay clear of it. Don't touch. Etc. Why? The training also emphasized the costs associated with ESD damage to components. A great deal of effort was spent making sure that we all understood that ESD damage to components might not be visible or even detectable at test / QA time, HOWEVER, in the field, the defect rate over time was dramatically lower when ESD controls were in place on the assembly and test / QA lines. This was serious stuff, the examples ranged from deployed PCs going inop after years of reliable service up to air-to-air missiles not functioning due to static damage. In the end, a very large sum of money was spent investigating the effects of ESD on the reliability of components in the field and it was determined that the benefits far out paced the costs of training everyone and taking precautions in the labs.
I now work somewhere much smaller and have a really hard time getting people to believe that ESD is real. I even had to fight a bit to ESD mats at the workstations where we do assembly.
There are a lot of myths and misperceptions surrounding ESD incidents, and I think that people would be well served by understanding that damange to electronic devices is not either fatal or non-fatal. A FET device might have it's gate region severely weakened by an ESD incident, but it would appear to function normally for an extended period of time. Perhaps the thermal efficiency has been compromised because the gate has partially broken down. The added thermal stress on the part over time will lead to early failure. The reason, naively, would look like a bum part or a thermal problem. The ESD problems don't always reach out and slap you across the face with a sign that says: "Zapped by poor assembly / handling techniques".
Just because you can't immediately establish a cause and effect with ESD, doesn't mean it's not causing damage. ESD can happen without you even noticing and it's especially hard to detect if it doesn't cause a failure outright. I.e. if a discharge doesn't outright kill a product, it can still do damage that increases the likelyhood of failure further on in the life of the product, or even worse, variability. Far worse than a product that you know doesn't work, is one that appears to work but produces wrong or variable output. Combined with the sheer number of products that have the potential to be damaged, increasing your ESD defence will likely give a decent increase your product's overall reliability.
If your product is so cheap that you don't care about failure and just want to produce as many of them as cheaply as possible, then go ahead and skimp on ESD. Just make sure you have a good system to deal with the failing products that will cost you less than the increased ESD defence.
This means, in practice, that there must be a conductive, resistive path between them prior to this happening.
This is easiest to achieve if you just connect everything together, the wrist straps, the conductive shoes, the whole 9 yards.
For example of what can happen. I have a component in my hand- but that's fine, I have my strap on, and the component is at the same voltage as me. I take off my strap with the component in my hand, and walk across the floor and touch someone on the shoulder with my other hand.
Zap! That component is now toast! The problem is that myself and the component built up a few thousand volts when I walked across the floor, and the person I touched instantly discharged me, and a large current came out of the component too... 90% of such components work for a while before failing.
You don't have to have all of these gadgets, but if you don't; you have to move very deliberately, and even then you can screw up sometimes.
-WolfWithoutAClause
"Gravity is only a theory, not a fact!"I recently toured at a major manufacturer of airbag controls...definitely a very critical and defect sensitive operation. Approximately 3 million parts are placed every day at this facility.
The floors have conductive particles embedded in them, all employees are required to wear either two ESD sole straps or take advantage of the conductive-sole shoe discount program. In addition, all employees venturing onto the production floor are required to wear an ESD smock. All handling equipment is grounded, conductive rubber used for conveyors, and many transport mechanisms are wire brush rollers. The environment is rigidly temperature and humidity controlled.
It's easy to develop huge static potentials when you have equipment running for a long time. Look at devices intended to generate high voltage static electricity: Wimshurst generators, Van de Graaf generators, etc work with simple components such as a rubber belt or rotating disk. Even dropping components can generate electricity; one high voltage generator works by dripping water from two tin cans.
Since the effects of static electricity may not show up until later, when the operating limits of a gate or two are much lower than necessary, it makes no sense to manufacture any electronic devices without rudimentary ESD protection.
...
I fried a dozen uVAX 3100e motherboards (back when they were cutting edge and majorly expensive) with static electricity. We had a maintenance contract with DEC, and they kept coming out to replace 'em.
We tried conductive floor materials, jumpering all the machines in the room together, stringing tinsel across all the paper trays, etc. etc. etc. ad nauseum but we kept frying motherboards every time a jumper wire slipped off or a floor mat got insulated by shoe crud.
Eventually I put a bunch of spider plants in the room and watered them every day. Humidity in the room went up to a reasonable level immediately from aspiration of water through the plant leaves.
Never burnt a single board after that. The plants were still on the job five years later when I visited the site again.