From: firstname.lastname@example.org (Glen Walpert)
Subject: Re: Side effects of potting compound !
X-Newsreader: News Xpress 2.01
Date: Fri, 08 Nov 2002 13:21:18 GMT
NNTP-Posting-Date: Fri, 08 Nov 2002 08:21:18 EST
Organization: Net Access (netaxs.com)
In article , email@example.com (Bob Wilson) wrote:
>In article ,
>>>Exactly. Unfortunately, too many people forgot their Grade 9 science, and
>>>got scared when they heard that normal RTV generates acetic **ACID**
>>Your error is basing your understanding of corrosion on that 9th grade
>>science. Just because copper does not instantly corrode when dunked in
>>acid, does not mean that acetic acid will not contribute to long term
>>corrosion - it will. And you cannot avoid the persence of water, it is
>>in a layer many molecules thick on all surfaces exposed to the atmesphere.
>>Try reading the ASM Metals Handbook Volume 13, Corrosion, if you want to
>>understand how electrolyte aided crevice corrosion occurs, and why the
>>electrolyte in itself does not need to be chemically corrosive to cause
>>corrosion. Any other modern book on corrosion will explain this also.
>>But like I said before, this is a long term effect which occurs over a
>>of years, depending also on the inevitable porosity in the gold plating on
>>the trimpot wiper, and since long term reliability is not an issue here ...
>So how are you going to get RTV to continually produce acetic acid for
>years? After a few days, it is fully polymerized and no more acetic acid is
>produced. RTV is a very permeable material (the most permeable of all common
>polymers), and whatever acetic acid vapours are trapped in cavities, will
>diffuse out long before "years".
>This has little to do with crevice corrosion, which is corrosion
>(particularly of ferrous metals) which is dramatically accelerated in the
>near absence of oxygen. We were talking about simple acetic acid attacking
>copper here, which cannot occur unless there is an oxidizing environment.
What is to make the acetic acid leave, after it has been adsorbed onto the
interior surfaces of the trimpot, and potted in epoxy? This will be a slow
process, the surface adhesion of the acetic acid will be quite good, and the
epoxy is not as permeable as the RTV (which will still contain traces of
acetic acid long after you can no longer smell it).
Your explanation of crevice corrosion is a gross oversimplification - an
electrolytic cell is formed by the oxygen concentration gradient in the
electrolyte, an effect not limited to steel with a salt water electrolyte, the
same effect will occur with copper in an acetic acid electrolyte - the acetic
acid will play the role salt plays in corroding your steel car, enabling
corrosion by making a highly conductive electrolyte, while not chemically
taking part in the corrosion reaction, which is oxidation. This is a well
known cause of failure of electrical contacts - like the wiper of a trimpot.
The question is not if corrosion will occur, or if it will occur faster in the
presence of acetic acid, the question is, how much faster will corrosion
occur, after the initial presence of acetic acid gives the corrosion cells a
"jump start". The answer may well be, not enough faster to be significant, or
even slower than if directly exposed to the atmosphere. Or not. It is an
unknown risk factor, you decide your risk tolerance.
My paranoia about trimpot corrosion comes from seeing *many* failed trimpots
in the field, probably an experience that a factory design engineer without
field experience does not have. On one job alone, a large ship simulator with
12 20 card racks of VME I/O containing hundreds of zero and span trimpots (and
an entertaining 4000 pole double throw switch to select simulation or 1/2 of a
real engineroom), we experienced *dozens* of trimpot failures after a year of
storage in an unheated warehouse. When the adjustment was moved and corrosion
wiped off, the affected channel would jump back to a reasonable value, off the
rails. This will be hard to do in a potted assembly. (There were similar
problems with the card connectors, solved only with lubrication.) And this
was a relatively benign environment, far from salt water - a marine
environment is absolutely murder on trimpots and cheap unsealed connectors
like VME card connectors.
As always, YMMV - none of this is an issue with cheap consumer electronics
with a one year warranty. OTOH, if you are designing for a 20 year life in a
harsh environment, trimpot contact corrosion should be a serious