*Rant* Question I asked at a Job interview.

[Freelander]

You haven't had sudden or infant deaths? ESD is insidious in the sense that damage might only show up later. Or your part is unstable without a real explanation.

I've had a couple of incidences where things quit working after handling them otherwise carefully. In one case, I actually registered the moment I zapped the part, which showed erratic behaviour afterwards. I've been more careful ever since.

Anecdotal and oft repeated semi-woowoo. 

[Mr. Scram]

Anecdotal and oft repeated semi-woowoo. 

I'm sure "woowoo" is why serious players spend millions on ESD prevention. I guess they don't know what they're doing, as opposed to a bunch of amateurs and one man bands on a forum somewhere.

I don't suppose you can do better than anecdotal evidence and "I've never killed anything so it's not a problem"?

[coppice]

You haven't had sudden or infant deaths? ESD is insidious in the sense that damage might only show up later. Or your part is unstable without a real explanation.

I've had a couple of incidences where things quit working after handling them otherwise carefully. In one case, I actually registered the moment I zapped the part, which showed erratic behaviour afterwards. I've been more careful ever since.

Anecdotal and oft repeated semi-woowoo. 

Mr Scram's position agrees with extensive industry research. Your's doesn't.

[Freelander]

Mr Scram's position agrees with extensive industry research. Your's doesn't.

Afraid not. The woowoo is the 'later failure' or the oft recited 'insidious' type of damage.
Protection is usually employed for the attempted elimination of 'instant' failure. NOT woowoo failure.
 

[Mr. Scram]

Afraid not. The woowoo is the 'later failure' or the oft recited 'insidious' type of damage.
Protection is usually employed for the attempted elimination of 'instant' failure. NOT woowoo failure.
 

You mean RS components is wrong when it says both immediate failures and latent defects are caused by ESD damage? You might want to shoot them an email, because that's probably something they want to know.

https://www.rs-online.com/designspark/what-is-esd-and-what-damages-can-it-cause

A latent defect can occur when an ESD sensitive item is exposed to an ESD event and is partially degraded. It may continue to perform its intended function, so may not be detected by normal inspection. However, intermittent or permanent failures may occur at a later time.

Or maybe Electronic Design?

http://www.electronicdesign.com/power/understanding-esd-and-eos-failures-semiconductor-devices

If a device is exposed to a weak ESD pulse and is partially damaged, it may continue to function well enough and pass production automated test equipment (ATE) tests meeting datasheet specifications. The defect may then extend over time with the device failing after several hours. These kinds of defects are known as latent defects, and failures are known as latent ESD failures. Latent defects are hard to detect, especially after the device has been assembled into a finished product.

Or maybe the Scottish Qualification Authority?

https://www.sqa.org.uk/e-learning/HardOSSupp05CD/page_27.htm

The damage caused to components may be severe enough to cause the component to fail completely. There is also a very high risk of latent failure where the damage is not immediately noticeable but still the device could fail later, either completely or intermittently. ESD damage will at best reduce the working lifetime of the component and at worst cause permanent, immediate failure.

[coppice]

Mr Scram's position agrees with extensive industry research. Your's doesn't.

Afraid not. The woowoo is the 'later failure' or the oft recited 'insidious' type of damage.
Protection is usually employed for the attempted elimination of 'instant' failure. NOT woowoo failure.
 

Try looking through the research. It is common for protection diodes to be partially or fully damaged, without the chip failing. Later, even a modest EOS situation can result in failure of the device, because the protection is either weakened or missing. However, research in the 80s by British Telecom, AT&T and others showed that partially damaged devices would often fail later in normal use, even without further EOS events occurring. Once the results of this work lead to better static control being introduced at their suppliers, long term in use failures of ICs in the telephone network reduced considerably.

[Freelander]

 

Some packets of nuts have 'may contain nuts' written on them......................
There has been a HUGE amount of work on IC protection - INTERNALLY. There are still warnings for the CYA factor of course.
In the 'real world' the ridiculous practices of supposed esd protection simply do not happen.
Some workshops though, do contain nuts and people who throw tins away after the sell by date ...............
 

I believe 'dowsing' may help detect insidious damage though ? (according to the SQA)

[NivagSwerdna]

There is clearly a continuum across the best of breed industrial environment. down to the repair workshop above a retail store and down through to a bunch of amateurs and one man bands... down to the polyester under-pant wearing tech with their Styrofoam coffee cup learnt up against the device they are tinkering with.

Sounds like the OP wants to go up-market.  Nothing wrong with that aspiration.

(I seem to remember a real world example of someone here describing something very expensive damaged by ESD.... so it's definitely a real scenario and if you care about MTBF then you should care about working practice).

[Mr. Scram]

Some packets of nuts have 'may contain nuts' written on them......................
There has been a HUGE amount of work on IC protection - INTERNALLY. There are still warnings for the CYA factor of course.
In the 'real world' the ridiculous practices of supposed esd protection simply do not happen.
Some workshops though, do contain nuts and people who throw tins away after the sell by date ...............
 

I believe 'dowsing' may help detect insidious damage though ? (according to the SQA)

Simply not admitting you're wrong won't change you being wrong about this. It's been looked at very well by various parties who have a huge incentive for it not to be a problem. Yet it turns out it is, and people are spending billions to mitigate the issues. Calling it "woowoo" does nothing but show your own lack of understanding the subject.

I'll leave it up to you to figure out how warning labels on products intended for the general public are different from industry standards followed by professionals, and why board houses and assembly plants still heavily invest in ESD mitigation, despite the fact that they're only handling fully built ICs with the built in protection in place.

[Freelander]

Simply not admitting you're wrong won't change you being wrong about this. It's been looked at very well by various parties who have a huge incentive for it not to be a problem. Yet it turns out it is, and people are spending billions to mitigate the issues. Calling it "woowoo" does nothing but show your own lack of understanding the subject.

I'll leave it up to you to figure out how warning labels on products intended for the general public are different from industry standards followed by professionals, and why board houses and assembly plants still heavily invest in ESD mitigation, despite the fact that they're only handling fully built ICs with the built in protection in place.

[dmills]

We found one of our products had a high field failure rate (It would pass soak then fail after a few weeks to a few months in in the field), all the failures were the SSD disk modules showing corruption.

The cure? ESD mats and wrist bands in the area where the disk images were being loaded in production, problem stopped instantly and we have never had another early life failure on the disks in those products.

No woowoo involved, we had an issue, we took ESD measures the issue went away, end of.

Regards, Dan.

[Freelander]

We found one of our products had a high field failure rate (It would pass soak then fail after a few weeks to a few months in in the field), all the failures were the SSD disk modules showing corruption.

The cure? ESD mats and wrist bands in the area where the disk images were being loaded in production, problem stopped instantly and we have never had another early life failure on the disks in those products.

No woowoo involved, we had an issue, we took ESD measures the issue went away, end of.

Regards, Dan.

So, are you saying the SSDs were irreparably damaged............... ?

[Mr. Scram]

So, are you saying the SSDs were irreparably damaged............... ?

He's saying they showed latent failures, exactly like the ones you said earlier are "woowoo".

You know, it's pretty awesome to be wrong. Being wrong means learning something you didn't know before. Being wrong means being a more knowledgeable person than you were yesterday. The only premise it that you have to concede to being wrong. Otherwise you're just condemned to being wrong every time.

[Freelander]

So, are you saying the SSDs were irreparably damaged............... ?

He's saying they showed latent failures, exactly like the ones you said earlier are "woowoo".

You know, it's pretty awesome to be wrong. Being wrong means learning something you didn't know before. Being wrong means being a more knowledgeable person than you were yesterday. The only premise it that you have to concede to being wrong. Otherwise you're just condemned to being wrong every time.

Latent failures  - (in the context described) - old boy, are not necessarily damage to the SSD. No damage to the SSD is not a hardware failure.
Stop making yourself so 'awesome'

[dmills]

We found that we could reimage them, but that they would fail again sooner of later.

These were MSATA modules in an industrial control system.

Regards, Dan.

[Freelander]

We found that we could reimage them, but that they would fail again sooner of later.

These were MSATA modules in an industrial control system.

Regards, Dan.

Probably a batch issue. Occam's Razor.
Highly unlikely to have damaged the mSata modules. They are extremely robust.
If you are happy with your solution though that is great.

 

[tggzzz]

I've been working with electronics my entire adult life, and I yet to kill anything due to ESD.

My daughter has been walking into roads without looking all her life, and has yet to be killed.

With normal precautions (lay off wool clothing in winter time), ESD is not all that dangerous.

ESD can cause subtle parametric shifts that may not cause problems until later.  Low-end repair shops are unlikely to be concerned by that!

[dmills]

Not a batch issue, we checked our records for that, having the ERP system record batch numbers is a good thing in general, it makes identifying when you **do** have a batch issue so much easier (And we have had that source of 'fun' as well)....
One other tell was looking at failure against production month, we were seeing far more returns on stuff built Oct/Nov/Dec/Jan/Feb then we were in the summer months! Coincidence? I think not!

Incidentally, be careful of consumer Flash cards and SSDs, they have a nasty habit of changing the technology while keeping the same model number, very annoying. The industrial vendors tend not to play this game for all that you are paying quite a lot more money (SD card vendors, looking at YOU!).

Having records of who, when and which parts for everything going thru production and then being able to do stats on the returns is way cool BTW.

Regards, Dan.

[Mr. Scram]

Not a batch issue, we checked our records for that, having the ERP system record batch numbers is a good thing in general, it makes identifying when you **do** have a batch issue so much easier (And we have had that source of 'fun' as well)....
One other tell was looking at failure against production month, we were seeing far more returns on stuff built Oct/Nov/Dec/Jan/Feb then we were in the summer months! Coincidence? I think not!

Incidentally, be careful of consumer Flash cards and SSDs, they have a nasty habit of changing the technology while keeping the same model number, very annoying. The industrial vendors tend not to play this game for all that you are paying quite a lot more money (SD card vendors, looking at YOU!).

Having records of who, when and which parts for everything going thru production and then being able to do stats on the returns is way cool BTW.

Regards, Dan.

Figuring out the technology in an SD card can be nearly impossible, even though this can be an important indicator of performance and reliability. It often seems to be a closely guarded secret. SSD manufacturers seem to be much more open about his, though some occasionally pull a bait and switch.

[SeanB]

I know that. 2 nominally identical Sandisk thumb drives, yet one has a quiescent current draw of around 50mA, and the other one is only 20mA. 50mA one rises to around 100mA on write, while the other will not exceed 40mA on a similar size similar speed block write.

[nctnico]

Raw components are more vulnerable to ESD, but assembled boards are not.

Oh sure they are! About 2 decades ago I sold self assembled PCs. I assembled one in a non-ESD safe environment due to lack of space and almost every component in that PC broke down (I used the same components in every PC; it has to be ESD).

[coppice]

If you've been involved in customer support for a silicon vendor, you may well have found that the most frustrating activity can be dealing with customers who have failed to take adequate control of their static. You'll get a a complaint that someone is seeing a high failure rate of one of your devices at production test, and they want to know why your device is so crappy. You get a few failed devices back from the customer, often horribly packaged, and send them to the analysis lab. In a high percentage of cases the analysis comes back "EOS" (electrical over stress). Some of the time the reason is a poor circuit design, that does overstress the part under some circumstances the customer hasn't allowed for. Much of the time its because the customer hasn't controlled static properly in their production facility.

If you are working with a reasonable customer things mightl go pretty smoothly from there. You do a site inspection, make some recommendations, and things are brought under control. However, in a percentage of cases you end up dealing with people in total denial. They've been building a billion boards a day for the last thousand years, and have never encountered such a problem before. It might be they've just moved from a more humid place, where they could get away with bad practices, and genuinely are seeing issues for the first time. Whatever the background might be, you need to get them past denial, because if you don't they will stop manufacture, and your revenue will fall. So, you have this difficult situation where the customer is being an idiot, but if you call them an idiot upper management will rush down to fire you.

[CopperCone]

how the real world works lol

it works because people don't know what the fuck their doing nor do they understand the ramifications of their actions

no one cares if something gets zapped and gets sent to the customer in a degraded state. $$$$

sad shit when you see it done to a 50,000 dollar product that weighs 700 lbs..........

esd does not kill instantly usually, it just degrades circuits, so they fail later. soft errors, etc can occur because of this, the impedance of something changed. if you say yea it works after i put it together, you don't really know whats going on, you would need a electron microscope and decapsulated circuits to see

[vk6zgo]

how the real world works lol

it works because people don't know what the fuck their doing nor do they understand the ramifications of their actions

no one cares if something gets zapped and gets sent to the customer in a degraded state. $$$$

sad shit when you see it done to a 50,000 dollar product that weighs 700 lbs..........

esd does not kill instantly usually, it just degrades circuits, so they fail later. soft errors, etc can occur because of this, the impedance of something changed. if you say yea it works after i put it together, you don't really know whats going on, you would need a electron microscope and decapsulated circuits to see

I've worked on a lot of stuff which fits your criteria, both on price & weight
Much of the original ESD stuff was driven from experience in really good labs where every effort was made to have a clean environment, including minimising humidity!

Much to their horror, after doing all the things which seemed correct, they discovered they were having problems from "static electricity".

The first attempts in ESD protective packaging was the "Poly pink"material, which deliberately reintroduced
humidity, in the form of a film of moisture on the inner surface of the material.
(At least, that is what the film we watched  back in the 1970s said, & it was produced by the manufacturer of the material).

Out in the world, we get away with a lot, because the environment is almost always far more humid than a
proper lab.

At the end of the day, you do the best that you can, & as wristbands & static mats aren't expensive, people use them most of the time, but sometimes, if you are up to your armpits in a very large non-modular piece of gear, there is no room for such things.

I almost always had the advantage of being able to track the stuff I fixed, & if it lasted for 5 to 10 years after my repair, I could probably say there were no subtle ESD faults introduced.
(If they were so subtle that they didn't cause any degradation in performance, in that time frame, they could
be ignored).

Not a conspiracy theory, but component manufacturers have a vested interest in blaming unexplained failures on the customer, as it costs them money to review their own procedures.

[philpem]

Welcome to university, where students leave $500 Metcal irons hot and untinned for an entire day, and occasionally some idiots try to probe signals with elevated ground. Price of equipment doesn't guarantee the operator has any sense of what they are doing.

You can lead a lazy engineer to the workbench, but you can't teach them to give two hoots about taking care of shared tools.

In my experience though, getting all new tools and sticking them up in plain sight on a shadowboard works damn well. Nobody wants to be "the first guy to knock a chip off the blades on the new Jokari stripper" or "that guy who scuffed the new JBC".