I've still got a paper copy of an HP Bench Brief from 1983 that was one
of my first introductions to the dangers of ESD. I've used a wrist
strap and antistatic mat since then. ESD protection in the ICs has
improved since then, but I think that the article is still mostly
applicable today.
http://www.hparchive.com/Bench_Briefs/HP-Bench-Briefs-1983-03-05.pdf
Ed
Lux, Jim (337C) wrote:
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf
Of Charles P. Steinmetz
Sent: Monday, January 25, 2010 10:27 AM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Conducting Bench Top Material
Bruce wrote:
Although over the years the non-conductive top has been an asset in
avoiding short circuits, etc., I am concerned about static discharges when
handling modern semiconductors. Would it make sense to spray the Masonite
with a weak copper sulphate or similar solution so as to make the masonite
slightly conductive, but not so conductive that 155 VAC connections
could not
safely rest upon it? Is there a better-suited material that could be used
to replace the Masonite?
One generally looks for static-dissipative surfaces, rather than conductive
surfaces. 1 Megohm/square, for instance. The idea is to keep everything
isopotential as charge drops onto things, not to rigorously establish a common
voltage.
I notice that many folks who have contributed on this thread use
anti-static benchtops, but I have never found it necessary (and I try
to keep the RH in my house under 45% -- it is generally 20% or less
in the winter). I've been fooling with static-sensitive parts for 35
years and haven't lost one to static yet
You haven't lost one *that you know of*. It also depends on the kinds of parts
you're working with. There are some that are quite sensitive AND which don't
fail outright, but just degrade performance a bit when they take a hit. It also
depends on the energy behind the hit, of course. An example might be the
MiniCircuits ERA-4 or ERA-5 (just because I happen to have the data sheet
handy). Take a look at the later pages in the report, and you can see where
the gain changes slightly as a result of 100V ESD hits (see page 6, where you
can see gain dropping about 1.5 dB over 8 pulses, with about 0.1dB per hit.)
As they say at the end of the report:
The new amplifier ERA-4XSM shows gradual degradation in the gain and the
device voltage. That fact is not so bad. Even with the multiple stress a
customer
would rather have gradual changes then catastrophic failure. The amplifier
withstands a single 100V ESD pulse, or 3 pulses at 50V.
http://www.minicircuits.com/pages/pdfs/an60028.pdf
----
When we (JPL) do site visits to vendors, lackadaisical approaches to ESD handling are one of the common problems. For us, who are building just one or two of something that's going to be going somewhere where repair isn't an option, latent damage and gradual degradation are a big deal.
It's really a "habit" thing that everyone has to get used to. That's why even nuts and bolts come
in ESD packaging (even though they're obviously ESD immune): it gets people in the mindset of "come in
the area, put on the wrist strap". Back in the 70s, when ESD processes started to be used, they would
have multiple categories of parts, some which needed ESD precautions (CMOS parts, DRAMs,etc.) and some which
didn't (resistors, capacitors). It was found that workers would be working with something in one category,
and the habits would carry over to the others, so the industry, in general, went to the "everything is
ESD sensitive" approach.
The *worst* offenders for ESD are the engineers (like those of us reading the list!), because they actually know what parts are sensitive and which aren't, and tend to take shortcuts with the non-sensitive parts. Which works, sort of, until they guess wrong, and cook something. "Hey, why is the NF on this LNA 0.2 dB higher than it was yesterday?"
jim
_______________________________________________
time-nuts mailing list -- [email protected]
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
