Electrostatic Discharge (ESD) - a lost topic?
Mar 6, 2006 at 12:16 AM Post #16 of 25
Quote:

Originally Posted by Twombly
I hope you're not taking off the fleece while working.


Of course not. I would never do that.
 
Mar 6, 2006 at 1:30 AM Post #17 of 25
Echoing some above posts, the only components worth being concerned about are integrated circuits and amps. And of these, the MOSFETs are probably the most sensitive. I usually will touch a grounded instrument chassis before I begin work, but I don't believe I've ever witnessed a component being destroyed by ESD. I have, however, seen a 12V lead arc onto a SMD MOSFET and kill it, leaving a decent pit on it's surface. That was a much lower voltage than humans typically produce, but the current was much higher in that case. If I worked on a production floor where expensive IC's were applied/reworked on PCB's all day long, I'd probably use more industry standard precautions. However, most of the parts I work with cost but a few dollars, and I'm comfortable just taking my chances.
rolleyes.gif


GeekGirl, I'd say the issue is likely understated in the audio community. The basic precautions aren't very cumbersome, and its not much fun trying to troubleshoot an improperly functioning component. That said, industry has very little tolerance for poor quality, especially with expensive and safety-critical devices like communication hardware. I'd wager this is the reason why they harp on you guys so much about the issue. Anyway, you have my respect for your proficiency in the ways of RF. Mastery of the art has thus far eluded me.
 
Mar 6, 2006 at 2:19 AM Post #18 of 25
Quote:

Originally Posted by Whitebread
Of course not. I would never do that.


Lighten up, guy. I was joking. :/
 
Mar 6, 2006 at 2:30 AM Post #19 of 25
Quote:

Originally Posted by Twombly
Lighten up, guy. I was joking. :/


confused.gif

I don't think my comment was too serious. I'm fine.
 
Mar 6, 2006 at 3:28 AM Post #20 of 25
Quote:

Originally Posted by michaelconnor
...the MOSFETs are probably the most sensitive ...
GeekGirl...you have my respect for your proficiency in the ways of RF. Mastery of the art has thus far eluded me.



michaelconnor- thanks for your support and understanding of the problem (perhaps we have crossed paths in the AVS forum Philly HDTV thread?). RF thrives on the fundamental properties of electromagnetic waves. PCB substrate material and trace widths are part of the equation. At audio frequencies, these effects are neglible- a difference of 50 mils in the trace width isn't important whereas at RF frequencies (> 1 GHz or so) it's a disaster.

I'm happy with the responses to this thread. Hopefully it's an incentive for those who help the beginners to remember to add a few pointers about ESD.
 
Mar 6, 2006 at 3:47 AM Post #21 of 25
I have the habit of touching something grounded every time I have to manipulate sensitive electronics.

Proper equipment like what is used in an industrial environment against ESD also costs a lot of money.
 
Mar 6, 2006 at 5:18 AM Post #22 of 25
I'll state what should be obvious.

ESD is not discussed on the forums here because it is essentially a nonissue.

The major semiconductor companies have seen to that, with rigorous ESD qualification, typically 2kV minimum, before a product launches. ESD protection simply entails nominally reversed biased pn junctions with some series resistance between the pin in question and another, generally a supply pin. The more complicated the chip, the larger the ESD matrix. The level of protection is set by the area of the diodes and the current limiting resistor value. FET gates, such as the inputs to JFET opamps or CMOS logic inputs, are most succeptible because the relatively low oxide breakdown voltages. Conversely the outputs are less sensitive, since they, whether for an inverter or a push-pull opamp output stage, present a similar circuit as ESD protection does. That does not mean that they are immune, just less sensitive.

My point is that ESD protection has been, in general, adequately addressed by the semiconductor companies to the point where hobbyist need not be overly concerned. If this were not the case, there would be plenty more help me threads and posts due to mysteriously dead chips. That is not to say be stupid about and rub your bare feet on a carpet before handling a chip. It's not like readers here handle bare wafers or chips being qualified, perhaps even without any ESD protection whatsoever (this is nerve wracking when it's your chip).

So don't worry about it and get back to the solder fumes, burning silicon, and eventual pride in your work when it finally works, in spite of your best efforts of sabotage things inadertently.
biggrin.gif
 
Mar 6, 2006 at 6:09 AM Post #23 of 25
Quote:

Originally Posted by SnoopyRocks
The major semiconductor companies have seen to that, with rigorous ESD qualification, typically 2kV minimum, before a product launches.


Static discharge voltages can be an order of magnitude higher than that. Also, ESD protection isn't meant to handle an arbitrarily large number of shocks. Every time you "hit" the protection device, it breaks down a little. Eventually, you break through it and start breaking something else further down the line.

I agree with your overall position. Basic precautions like grounding yourself after moving around are sufficient. Just don't depend on the device's ESD protection. Not only can you break through it, you really should save the ESD protection for the device's field use. You'll be zapping the device regularly as you use it over the coming years.

Quote:

That is not to say be stupid about and rub your bare feet on a carpet before handling a chip.


Actually, I've found that taking off my shoes and socks will help prevent static buildup more than any other simple technique.
 
Mar 6, 2006 at 12:46 PM Post #24 of 25
Quote:

Originally Posted by GeekGirl
michaelconnor- thanks for your support and understanding of the problem (perhaps we have crossed paths in the AVS forum Philly HDTV thread?). RF thrives on the fundamental properties of electromagnetic waves. PCB substrate material and trace widths are part of the equation. At audio frequencies, these effects are neglible- a difference of 50 mils in the trace width isn't important whereas at RF frequencies (> 1 GHz or so) it's a disaster.

I'm happy with the responses to this thread. Hopefully it's an incentive for those who help the beginners to remember to add a few pointers about ESD.



I too work in RF (currently working on a broadband PA that tops out at 2 GHz), and it's been a real education. In the GHz range, everything matters, and making measurements can be very trying.

I've worked on digital stuff as well, and ESD is a pretty big deal there. I have seen flash memory chips fried through mishandling. On the manufacturing side, ESD is a huge issue, and since we are an ISO 9000 company, we have to comply with industry standards - that means anti-static flooring, mats on the benches, etc. Wrist/heel straps must be checked daily and logged for compliance. We've recently added devices to benches that automatically monitor and verify the continunity of your wrist strap. It's harder to enforce this attitude in Engineering, though
rolleyes.gif


A company I worked for years ago produced a video on the effects of ESD on electronic components, not just killing them, but changing device characteristics. They basically took a part (I think it was a MOSFET, but I don't recall for sure now). They measured the device on a curve tracer, and zapped it with various levels of ESD and remeasured it after each zap. They also demonstrated how different packaging protects (or doesn't protect) sensitive parts.

I'm a bit of an ESD nag, since I'm the guy who usually has to troubleshoot and repair the stuff that gets fried.
frown.gif
Software guys are especially bad IMO, and I've had to lay down the law regarding handling of PWB's and wrist strap usage. And I do have a wrist strap at home, and use it when I have to get inside my PC.

The best way is to develop good work habits and use ESD precautions, whether it's a wrist strap or just being careful to ground yourself before handling sensitive devices.
 
Mar 7, 2006 at 2:07 AM Post #25 of 25
Quote:

Originally Posted by GeekGirl
michaelconnor- thanks for your support and understanding of the problem (perhaps we have crossed paths in the AVS forum Philly HDTV thread?). RF thrives on the fundamental properties of electromagnetic waves. PCB substrate material and trace widths are part of the equation. At audio frequencies, these effects are neglible- a difference of 50 mils in the trace width isn't important whereas at RF frequencies (> 1 GHz or so) it's a disaster.
.



Unfortunately I don't watch much TV, so I don't think I've been there, but it's always cool to meet other Philly people on Head-Fi.

That's a good point you've raised about EMI and audio, and one that's been tugging at me since I first got into this hobby. Often, it appears that EMI reduction principles for high-freq applications are implemented with regular analog audio equipment without a second thought. Shielding methods aren't always the same when dealing with different types of interference, but I don't always see this factored into audio hardware design decisions. I'm still not too sure how relevant EMI is to audio design, but looking at the triple-shielded cables I occasionally see floating around, you'd almost suspect otherwise.
 

Users who are viewing this thread

Back
Top