ESD safety

[mike7877]

Winter is coming! And we all know what that means... More static shocks!

When looking in datasheets, I notice that op-amps are rated for voltages in the 1000V-5000V, usually on the lower end though (1-2kV)

We've all stood up from our desks with headphones on and heard the whining noise of hundreds to thousands (or hundreds of thousands!) of little shocks discharging one after another in rapid succession...

Now that I've got some exceptional gear, I'm kind of concerned... Why? It doesn't take much sometimes - just moving around in my chair I can hear a few discharges depending on how dry the air is.


My question is this: has anyone ever damaged the analog output from an op-amp or low-powered discrete amp with ESD this way? I believe damage would manifest as lower level output from increased resistance from damaged semiconductor

 

[solderdude]

The original L30 could be damaged this way and some did.
The ESD path was an unusual one via the potmeter shaft to audio ground via the wiper most likely.

Normally, when the designer has done a good job, ESD should flow around the PCB through the enclosure.
I can't say if very expensive custom made hifi gear is ever tested for ESD. Most likely not.

No need to worry about components (not only opamps) that are in a circuit. They all should survive 2kV.

Not all are tested for 4kV or even 8kV where testing could be up to 15kV (contact) to 30kV (air) for automotive.

Damage to semiconductors usually results in catastrophic failure (short).
Sometimes it leads to latent failures so a semiconductor material could be damaged a bit which 'grows' in the material until it breaks.

 

[fpitas]

It's a valid concern. If you think you need them, there is a world of inexpensive ESD suppressors available. Here's one:

https://www.mouser.com/datasheet/2/240/Littelfuse_Polymer_ESD_Suppressor_PESD0402_240_Pro-3312611.pdf

 

[solderdude]

A valid concern during production and handling of semiconductors and also needs to be taken into account during the design phase (schematic and PCB layout as well as the enclosure).
A product bought in a store should not need any additional parts. Sufficient counter measures should be present in completed products. At least up to 2kV that is.

 

[fpitas]

A valid concern during production and handling of semiconductors and needs to be taken into account during the design phase.
A product bought in a store should not need any additional parts. Sufficient counter measures should be present in completed products. At least up to 2kV that is.

I agree. But I wonder how true that is.

 

[MCH]

I agree. But I wonder how true that is.

For a diy project I was doing recently, I had a look at different HDMI extractors and got surprised how many had the ESD protection footprints in place... but left unpopulated. I even wonder if there was a reason behind...

 

[MaxwellsEq]

A valid concern during production and handling of semiconductors and also needs to be taken into account during the design phase (schematic and PCB layout as well as the enclosure).
A product bought in a store should not need any additional parts. Sufficient counter measures should be present in completed products. At least up to 2kV that is.

In my experience, pro production gear was largely insensitive to static shocks. Computer and network gear less so. Wearing an earth wrist strap when the lid was off, was usually good practice just in case.

One expert I spoke to said that statistically, with a large volume of devices, static discharge can affect the distribution of failures, with affected devices dying earlier than non-affected devices. This is probably true, but I never experienced it in practice.

 

[solderdude]

For a diy project I was doing recently, I had a look at different HDMI extractors and got surprised how many had the ESD protection footprints in place... but left unpopulated. I even wonder if there was a reason behind...

View attachment 324222

It could be that the used electronics connected to it did not need additional components and might comply to EMC/ESD on its own.
Could also be cost savings and relying on the screen making contact first.

 

[solderdude]

In my experience, pro production gear was largely insensitive to static shocks. Computer and network gear less so. Wearing an earth wrist strap when the lid was off, was usually good practice just in case.

One expert I spoke to said that statistically, with a large volume of devices, static discharge can affect the distribution of failures, with affected devices dying earlier than non-affected devices. This is probably true, but I never experienced it in practice.

During EMC and ESD testing for railway monitoring systems (much harsher environment than home gear) all designs passed ESD with flying colors (especially blue sparks when testing air discharges).
Only some (in the systems integrated) PC's gave some problems.
We used professional PC's that could be opened on the front side (with a key) and every time a discharge was made (-8kV) the PC 'hung'.
(I witnessed all the EMC tests when they were done)
The test condition allowed momentary non latching errors but the key lock (just a metal part making contact with a metal enclosure) seemed to do that.
The solution was ... fit a wrist wrap and a note to wear it when servicing.

 

[DVDdoug]

The ESD path was an unusual one via the potmeter shaft to audio ground via the wiper most likely.

Normally, when the designer has done a good job, ESD should flow around the PCB through the enclosure.

Right... It the "path" to the circuitry that's important. You're not normally touching the circuit board and the most vulnerable are the inputs & outputs, and once it's connected you aren't normally touching the connections, and the inputs are normally shielded so you can only touch ground.

Once an op-amp is installed on a board in a circuit, the impedances are lower and there are parallel current-paths. You still shouldn't touch a circuit board without being properly grounded but it's not quite as "dangerous" as touching the raw parts.

I work in electronics and it's one of my hobbies. I live in California and I occasionally get a static shock, but it's rare here. My ESD practices at home are rather lax. I usually just touch ground to discharge myself before touching the circuitry. Over many years I have may have had two "unexplained failures" that where were possibly caused by static discharge.

I've NEVER killed a piece of "complete" equipment that was in its case/cabinet with ESD. I'm ONLY concerned & careful when touching circuit boards.

Where I work now, we have strict procedures. We have ground foot-straps (I actually have ESD shoes) and treated-conductive floors and we have to test the straps/shoes every morning, and we have to wear an anti-static smock.

I've worked at other companies that were more lax and I've come to the conclusion that it's an over-blown problem, at least with the environment/weather we have here. I'm NOT dismissing it altogether but it's one of those things that's easy to enforce, and QA and management like things they can easily check & control, even if they are not the most important things.

 

[mike7877]

It's a valid concern. If you think you need them, there is a world of inexpensive ESD suppressors available. Here's one:

https://www.mouser.com/datasheet/2/240/Littelfuse_Polymer_ESD_Suppressor_PESD0402_240_Pro-3312611.pdf

I took this snip from the datasheet:

I don't know if you've personally heard what I'm about to describe, because it hasn't happened to me since maybe 15 years ago, and back then, I only experienced the problem temporarily.

My most concerning ESD issue is sitting in my recliner chair. To the right of the chair, I have my fast (4.8GHz) vintage 2500K system with 32GB DDR 3 2133 CL10 - I stuck a 6600XT in it to keep it modern, and a good Adaptec RAID card with 512MB cache (RAM), PCIe 2.0 x8 (for up to 4GB/s to the CPU) driving 6x 1TB WD SATA SSDs in RAID 5 lol). Anyway, there is a side table in front of the chair, as close as it can be without the foot recliner part hitting it when it goes up. On that table is my monitor, with its base on the front left corner so the screen is as close to directly in front of me as possible. I have to turn my head maybe 12-15 degrees to the right to be looking directly at the centre of the screen. When I move around in the chair, say I reach for something beside me on the left side of the chair and I have to lift my arse 2 or 3 inches, when the humidity is low (which it is now and will be until March or April in all likelihood), I'll get maybe 20-30 static shocks happen in about 200-300ms. I believe it is one of the earbuds shocking the inside of my ear. The way the earbud is designed, is the entire thing is rubber and plastic (mostly plastic covered in rubber) except for where there is a very fine grill which covers the driver. This grill is recessed about 0.15-0.2mm from the rubber, and doesn't touch my skin directly. So.. once the earbuds have been in my ear for 15 minutes or so, the humidity between them and my eardrum rises quite a bit (probably to at least 70%). Because of this, I don't think resistance is very high for the static, so the ESD voltage, I'm thinking (/hoping) is probably low. Something which also made me think the voltage isn't very high was: I don't feel the electric shocks as they happen. But I thought about this some more, and, although the outer ear is quite sensitive to pain (from static shocks or anything really), the ear canal itself, is not. It doesn't seem to be, anyway. The threshold for sensation seems to be pretty high, but once it's reached, it's as sensitive a spot as any. I think that's how it works anyway. Maybe it's just me and I'm weird lol. So the voltage could be semi-high. (after all that I just googled threshold to feel for electric shock, and 2-4kV was returned, so maybe it's good to assume this is less than 4kV)

Anyway... if I stand up from the chair to grab something from a little ways away, or turn on the light, or anything... THAT's when I don't get 20 or 30 shocks, but 1-3 thousand of them! These extra shocks are over more time, but not much... maybe 1500-2000ms. The reason I say there are 1000-3000 of them is the frequency from all the pulses is in the 800-1500Hz range...

tl;dr:
Do you think these ESD saving devices are good for thousands of shocks within seconds? How does a static shock dissipate? A random memory from the past when I was discussing static discharge with a prof at school, he described it as high frequency AC because it's a pulse which rises really fast. I wonder if this is still the case (considered AC) when you get thousands of them in a row into a capacitor... (does voltage rise and fill up the cap, or does it behave like AC)

I wonder, also, if there are any (cheap) devices which I could put in line with the headphones that would measure the voltage?

 

[mike7877]

The original L30 could be damaged this way and some did.
The ESD path was an unusual one via the potmeter shaft to audio ground via the wiper most likely.

Normally, when the designer has done a good job, ESD should flow around the PCB through the enclosure.
I can't say if very expensive custom made hifi gear is ever tested for ESD. Most likely not.

No need to worry about components (not only opamps) that are in a circuit. They all should survive 2kV.

Not all are tested for 4kV or even 8kV where testing could be up to 15kV (contact) to 30kV (air) for automotive.

Damage to semiconductors usually results in catastrophic failure (short).
Sometimes it leads to latent failures so a semiconductor material could be damaged a bit which 'grows' in the material until it breaks.

Dang, so the pot got wrecked? I'm thinking of getting the L30 II.. do you know if the problem was fixed for sure?

Random: I accidentally bought the E30 II instead of the L30 II lol. Before returning it I decided to try it out (why not, right?)
I would recommend it for the price... The DX1 sounds better IMO (at least in my system)
I know, not very elaborate, but I don't want to derail this thread lol

 

[mike7877]

One expert I spoke to said that statistically, with a large volume of devices, static discharge can affect the distribution of failures, with affected devices dying earlier than non-affected devices. This is probably true, but I never experienced it in practice.

Devices which experienced ESD dying earlier? Or devices affected by ESD with an effect from the ESD, dying earlier?
Mean all devices experienced ESD and they were all more likely to die earlier?

 

[mike7877]

Sometimes it leads to latent failures so a semiconductor material could be damaged a bit which 'grows' in the material until it breaks.

Like a crack?

 

[solderdude]

Like a crack?

Yes, sort-off... microscopic level.

Dang, so the pot got wrecked?

No the shaft of the pot and the enclosure were not grounded. If it were (this was fixed later on) then any ESD would be routed through the metal enclosure and not enter the PCB/parts.

E30 and L30-II do not have this issue. Only the early L30 did.

 

[MaxwellsEq]

Devices which experienced ESD dying earlier? Or devices affected by ESD with an effect from the ESD, dying earlier?
Mean all devices experienced ESD and they were all more likely to die earlier?

It's a statistical thing. If you make 10,000 switch ports (that are fitted by the end user into a chassis) and half of them are protected during installation and half are not, the poorly protected kit will have a statistically shorter life (see bathtub curve).

You can't work backwards from the general to the specific in this case, so a specific poorly protected device may last 10 years and a specific well protected device last 8 years, but in general, you would need to refresh the poorly protected kit earlier.

 

[mike7877]

It's a statistical thing. If you make 10,000 switch ports (that are fitted by the end user into a chassis) and half of them are protected during installation and half are not, the poorly protected kit will have a statistically shorter life (see bathtub curve).

You can't work backwards from the general to the specific in this case, so a specific poorly protected device may last 10 years and a specific well protected device last 8 years, but in general, you would need to refresh the poorly protected kit earlier.

I see. Very good!

 

[Berwhale]

Has anyone tried to ESD protect a standard office chair? I have a Sidiz T50 in my study and I often get loud crackling through my iLoud monitors when I get up and then a shock from the next metal thing I touch - this has been enough to condition me to be wary of metal door handles regardless of my location.

I know that ESD office chairs are a thing. Has anyone tried to ESD proof a standard chair? I happy with my T50 in all other respects and loathed to replace it with something else without attempting to fix the ESD issue.

I was thinking of purchasing some 'Faraday' tape and attaching is to some strategic locations on the chair and then grounding it using an ESD wrist strap.

 

[Speedskater]

Back in the days of printed catalogs, I had at least one catalog that had a section on ESD reducing devices, materials and supplies.
But now i don't remember what vendor it was.

 

[solderdude]

Electrostatic charges come into existence when 2 non conductive materials are rubbed together of pulled apart.
That charge can't go anywhere when walking on carpets or something with isolation between feet and ground.
That charge can accumulate on the body (which is conductive).
Once you touch something that is grounded (other potential) the electrons will find a way though your body to the ground potential.
Spark flies and you can feel a jolt if enough electrons leave your body at once.

Your non ESD chair has isolating materials, isolating feet and and the floor, most likely, is also non conductive.
Riding/sliding on the floor = friction = charge building up. Getting on/off the chair or 'reseating' is friction = charge building up.
Touch something grounded and it flows away.

So... to get your chair to become ESD safe requires a slightly conductive (ESD) floor.
Wheels that are conductive (like car tires are)
Cloth on the chair that has wires in it or materials that conduct charges and do not create much when there is friction.

What you could do is wear an ESD wristband while on the chair that is tied (via a 1M resistor which usually is the wristband) which is connected to safety ground.
Once you got of the chair take it off or disconnect and you should have no charge to speak of.
Walk with friction on a carpet and its there again.