ESD confirmed kills?

[ice595]

Its Winter time here in Canada, which means "nasty static shock season".
Every time when I take off or put on my sweater I have to have 1 hand touching a grounded surface to avoid any shocks later,
or kill any ESD sensitive devices.

I've never actually had any components or devices killed by ESD before. (not that I know of at least)
The things I've handled includes BJTs, FETs, MCUs, CMOS logic gates, PCBs and even high-end computer motherboards + CPUs 
My work bench's surface is just a particle board, no ESD mat on top, no anti-static wrist-strap.

Have you guys had any confirmed kills by ESD?
Please share your experiences with ESD.

[AndyC_772]

Two things which most people don't realise about ESD:

- it's possible to have an ESD event occur which is not felt, but which nevertheless contains enough energy to damage a sensitive device, and
- ESD damage often doesn't mean a device fails immediately, but its lifetime and/or operating performance can be severely degraded.

This is a big deal, because it means people can handle boards badly, and remain blissfully ignorant of the damage they're doing because they don't feel or hear any sparks, and the board still works afterwards. "Therefore", the deeply flawed logic goes, "no damage has been done, and ESD isn't that big a deal".

Many years ago, when I was the hardware guy in a room full of software engineers, people would occasionally hand me boards and complain that they "just stopped working". For some reason they expected me to care, or to do something about it.

I did. I ordered up a big roll of ESD matting, and cut everybody a sheet to cover their desks. I fitted earthing studs, connected them up, and provided everyone with a wrist strap. I also sent out an email to all staff, explaining a bit about ESD and why it was important to use the equipment which I'd provided.

It didn't go down well, but that didn't matter. What did matter was that nobody ever handed me a board again which "just stopped working".

[FrankBuss]

It didn't go down well, but that didn't matter. What did matter was that nobody ever handed me a board again which "just stopped working".

Now the question is, did it work or were you never asked again for a broken board because they feared what you would do next?

[Rick Law]

Its Winter time here in Canada, which means "nasty static shock season".
Every time when I take off or put on my sweater I have to have 1 hand touching a grounded surface to avoid any shocks later,
or kill any ESD sensitive devices.

I've never actually had any components or devices killed by ESD before. (not that I know of at least)
The things I've handled includes BJTs, FETs, MCUs, CMOS logic gates, PCBs and even high-end computer motherboards + CPUs 
My work bench's surface is just a particle board, no ESD mat on top, no anti-static wrist-strap.

Have you guys had any confirmed kills by ESD?
Please share your experiences with ESD.

I have my first immediate-death confirmed happened to me 3 or 4 days ago.  I walked back to my PC, reach for the mouse, zap!   I have it put aside waiting for available time to open it up and check it out.  Mean while, I plug the mouse back in on occasion (once a day) to see if (and hoping it is) merely an intermittent failure.  So far, dead as a door nail.  Some point in time, I will open it up and see if I can find the dead component.  I like that mouse, so I would actually put a bit of effort into seeing if it can be revived.

As to the "reduced life" kind, I probably had a lot and don't know.  This is not the first time I zap a mouse or laptop, but it is the first immediate failure.

[AndyC_772]

Now the question is, did it work or were you never asked again for a broken board because they feared what you would do next?

Either way, the *real* problem was solved 

[zapta]

Here is a confirmed kill in simulated conditions, around 13m30s.

http://www.eevblog.com/2012/02/25/eevblog-247-anti-static-bag-myth-revisted/

[AndyC_772]

Mosfets are particularly sensitive. Because the gate is fully insulated from the channel, there's no path for any charge which may build up on it to leak away. It acts as a very small capacitor, and by Q=CV, any charge on a small capacitor can result in a high voltage.

This voltage exceeds the breakdown strength of the gate oxide, and the result is a scrap component.

[edpalmer42]

Here's what HP has to say about ESD.  Thirty years old, but still applicable today.

http://www.hparchive.com/Bench_Briefs/HP-Bench-Briefs-1983-03-05.pdf

Ed

[SeanB]

I once tried to kill an old Celeron processor by giving it a massive ESD event using an electric fence energiser. After 5 minutes of flashes between pins I put it back in the socket , put the heatsink on ( otherwise they do pop rather spectacularly) and powered it up. The bloody thing still worked! Took a little more work involving a car battery and a spanner to melt the pins off before it failed, mostly because it no longer fitted the socket.

[leppie]

I once tried to kill an old Celeron processor by giving it a massive ESD event using an electric fence energiser. After 5 minutes of flashes between pins I put it back in the socket , put the heatsink on ( otherwise they do pop rather spectacularly) and powered it up. The bloody thing still worked! Took a little more work involving a car battery and a spanner to melt the pins off before it failed, mostly because it no longer fitted the socket.

SAA managed to kill my AMD CPU in one flight :p

It worked fine for most part afterwards, but some of the 'SSE' parts were physically damaged. The corner got crushed (the good ol days when CPU had no IHS).

OT to SeanB: Picked up a Philips PM3209 scope (with original probes and operator manual, excellent condition) and separate 2Mhz function generator for R500 on Gumtree this week Now I have 2 analog scopes.

[tszaboo]

I've seen an MCU resetting because of ESD. It did not die though. Yet?

[IvoS]

I managed to accidentally kill 2 house thermostats in 1 year,  but they both got a real spark zap by touching them.

[EEVblog]

I read that title as human kills!

[sacherjj]

$25 Cotton/Poly blend ESD smock, with tight wrist cuffs.  Snap onto the pocket and onto your board with a 10 feet cable.  Gets the job done. 

Really sucks when the last working of 4 $700 prototype boards gets toasted, because someone isn't doing simple ESD precautions.

The biggest problem is when you are making a product to last.  Small shocks can damage, but not enough to fail functional tests.  It will just fail in the field.

[Neilm]

I had an ESD event that killed the MP3 player in my pocket. It was probably about an 8kV spark.

(I now remove all electronics even my car keys from my pockets before doing ESD testing).

[wilheldp]

I've wondered about this too since Dave's video about the firmware issue in an Agilent DMM.  He said he may have messed it up during the teardown, but part of his "robustness" test when tearing down a DMM is to not wear a wrist strap and see if ESD kills anything.  I would assume there would be a lot of sensitive parts inside a DMM, so if ESD doesn't kill them, why does it kill other things?

[denelec]

It gets very dry at my home in winter.
A few years ago I killed my old electronic alarm clock when I touched it. 
I decided to repair it by replacing the electronic circuit with another I made using a Picaxe 40X2 and a veroboard. 
I kept the same case, buttons and led display.  Now, the alarm clock looks and sounds exactly as before. 
It was the electronic equivalent of a brain transplant.
It made no economic sense to do so, but it was fun.

[Nerull]

We had a big problem with mosfets on a PCB self immolating after handling at work before making some modifications to the work areas - a spray booth used for conformal coating, to be exact. Then the problem vanished.

It's a neat demonstration when you can pick a mosfet up off a board with tweezers because it's so hot its reflowing the solder on the board, but it gets expensive after a while...

[EEVblog]

Really sucks when the last working of 4 $700 prototype boards gets toasted, because someone isn't doing simple ESD precautions.

At a former company when we had a $10M military contract on the line, and only had ONE meticulously hand build prototype to rely upon, the company instigated a new rule that it was a instant sackable offence if you entered the lab without your ESD coat and tested wrist strap.

[scientist]

instigated

instituted?

[ice595]

I work at a computer retail store and every week or so we get returns on motherboards or memory that are "defective"
(and yes, we actually test it before accepting any defective product returns)
Even for brand name motherboards like ASUS and MSI we're still getting "defective" units back
I'm pretty sure those motherboard manufactures have tight quality control over their products (ie. test each board before shipping)

So I guess they've been zapped by ESD when handling and thats why its "defective"? 
Anyways, we can't really tell if was indeed defective or they've been zapped with ESD. 

[SeanB]

We had a big problem with mosfets on a PCB self immolating after handling at work before making some modifications to the work areas - a spray booth used for conformal coating, to be exact. Then the problem vanished.

It's a neat demonstration when you can pick a mosfet up off a board with tweezers because it's so hot its reflowing the solder on the board, but it gets expensive after a while...

I have had stuff that did that with a TO3 packaged power transistor. It got hot enough that the wires to the leads would unsolder themselves. Sometimes you just resoldered them, but normally they had run hot enough that they went short circuit.

[AndyC_772]

So I guess they've been zapped by ESD when handling and thats why its "defective"? 

More than likely. The PC 'enthusiast' crowd are the absolute worst offenders I've come across when it comes to ESD precautions, or lack thereof. Because they think they're experts, and that they "know about computers", their own personal experience trumps well established industry practice on handling ESD sensitive components.

[peter.mitchell]

I read that title as human kills!

I do believe it has happened,  indirectly though,  I think it was someone vaccuuming up wood dust, then ESD event,  ignited the dust,  explosion,  dead.

[FrankBuss]

I do believe it has happened,  indirectly though,  I think it was someone vaccuuming up wood dust, then ESD event,  ignited the dust,  explosion,  dead.

And flour mills can be more dangerous than you would expect:
http://www.wsoctv.com/news/news/local/dust-potential-factor-statesville-flour-mill-explo/nXHTK/

[richard.cs]

I'm not always as careful with ESD as I should be, but the only confirmed kill so far is an RF JFET. Unless you count lightning strikes as an ESD event.

[Gath]

Well, the only chips I have toasted were the ones I used in my very first project (back in early college days). I zapped a GAL22v10 and a handful of PIC 16F series This was my first encounter with Murphy! Ever since, I am trying to be a bit serious when dealing with sensitive parts (mostly FETs based and alike).

Really sucks when the last working of 4 $700 prototype boards gets toasted, because someone isn't doing simple ESD precautions.

I've a friend who zapped custom chips straight out of the foundry. Talk about a mood killer after finally receiving a tapped out chip you have worked so much on. Thank god you always order many. But still ...

[AlienRelics]

I know this is an old thread, but it bears reading again.

Sainsmart sent me two 3.2 inch TFT displays with the pins stuck in white foam, packed in clear ziploc bags. Both were damaged, the displays had lines and would not work correctly. They also shipped four 2.4GHz transceivers with the pins stuck in white packing peanuts, held on with plain clear packing tape and stuck in clear ziploc bags.

None of them worked properly.

[Coceth]

Blew up an LT1920 (instrumentation amp) when making a college project. It was the most expensive part on the board, about $7 at the time  . Luckily, I had a back up just in case.

[xani]

A lot of chinese sellers sadly do not pack things correctly, altho so far I have been lucky.

I've seen ESD power-cycling keyboard when I touched it after getting up the chair (yes, via plastic buttons), occasionally hanging the controller.

Got annoyed enough, bought a new, different one, same thing happens just without hanging it