In a message dated 06/27/2000 3:26:26 PM US Eastern Standard Time,
[log in to unmask] writes:

> Doug, thanks for the information

You're welcome.

>
>  Answer to your questions
>  The Chlorine spikes were relatively small, what is a better test to
>  ascertain chlorine contamination

Well, there are a number of different residue analysis techniques.  Our
preferred method is ion chromatography (IPC-TM-650, method 2.3.28).  It is an
extraction based test, but a fairly rigorous one.  Like any technique or
tool, it has its plusses and minuses.

>
>  Your last comment about the corrosion problem when mixed with moisture was
>  like you looked into crystal ball and know the problem that we are
having!!!

Seen it many times before.

>
>  How much moisture is required?

That depends (I irritate a lot of people with this response).  The greater
the level of ionic contamination, the lower the moisture threshold required.
The greater the electronegativity of the residue, the greater the threat.
You would need less water at a higher driving voltage.  Think of
electrochemical failure as a three ring Venn diagram, with the three rings
being moisture, ionic contamination, and voltage.  The diameter of the
circles relates to the magnitude of the effect.  The intersection of the
three areas is the risk of electrochemical failure.

>
>  Have you ever seen a glass diode that loses the "normal" orange color and
>  turns a pale color, it seems that the copper oxide layer between the diode
>  slug and glass is being etched and then causing the diode to become leaky,
>  the diode is surrounded by polyurethane but there is a chance that moisture
>  could have a path onto the pcb. Also the negative side of the diode
>  discolors first.

Well, not having seen the diode in question, but having seen this phenomena
before, I suspect that a corrosive element exists on the copper oxide layer.
The conversion to a lighter color is indictive of oxide stripping.  We have
all seen oxided copper come out nice and shiny after a sulfuric acid dip.
How that corrosive material got there is another issue.  Either the component
manufacturer did not rinse properly before the application of the glass shell
(or it may be an epoxy shell), or the seal between the lead and the glass
body is not hermetic and a corrosive fluid can flow into the body of the
diode by capillary force.  The discoloration at the negative side of the
diode is not unusual.  I see it in SIR testing all the time.  Alternate
fingers of a comb pattern darkening.  It is an electrochemical effect, but I
don't presume to be electrochemist enough to explain it.

You talk about the diode becoming "leaky".  In most of the situations I have
seen, the corrosive material is attacking the internal metal of a component,
corroding it open.  In the case of a diode, from your description, I would
expect the metal to be corroding open.  If you are seeing electrical leakage
across that part, then you may have an alternate conductivity path, such as
across the board surface, or through the urethane coating, rather than
through the diode.  You may have two failure mechanisms acting in tandem.

Try this.  Remove the faulty diode and test it independent from the board.
It the failure mechanism remains, the fault is the diode itself.  If it
doesn't, replace the diode with a known good diode and test the assembly.  If
it works, then you can conclusively indict the bad diode.  If not, then you
likely have a board level problem.

Doug Pauls
Technical Director
CSL

##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in
the body:
To subscribe:   SUBSCRIBE TECHNET <your full name>
To unsubscribe:   SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Keach Sasamori at [log in to unmask] or
847-509-9700 ext.5315
##############################################################