Whilst I wholeheartedly concur with what Doug stated, I do not agree for
the
need for Ion Chromatography (IC) testing.

*WHAT?!!!!  LIGHT THE TORCHES, BURN THE HERETIC.......  (he turned me into
a newt....I got better...)
I must disagree with your disagreement.

Iain, if you run a correct SIR test regime on your proper proposed
production process, and you have acceptable results, then you can feel
comfortable that you will have a reliable end product. If you have failures
in this test, then you need IC to find out precisely what is present on the
surface that may be causing your problem. Yes you should then use ionic
extract testing (SEC/ROSE) to keep an eye on your process conditions - but
make no attempt to correlate SIR with SEC/ROSE results.

*Graham brings up a valid point that there is more than one tool for
assessing cleanliness or quality, but I don't know that SIR would be any
better than IC at this point.  For both tests, you need someone to help you
interpret the numbers.  Having the information that your assemblies have
2.7 ppm of chloride doesn't do much good unless you know if that is a good
or bad number.  Having the information that your process yields a mean SIR
level of 283 megohms on a B-24 doesn't mean much unless you can relate to
product.  The advantage of SIR is that of these two tests, it is better at
showing if residues have a propensity for electrochemical failures
(leakage, corrosion, metal migration), but you still have to do more
extensive correlation studies on actual product to know whether the values
from that lab test relate to product life in the field.  Then you have to
also do a correlation study with ROSE/SEC to see what levels to use as a
target and upper control limit.  An additional disadvantage of SIR is that
it is a long test.  Brian Ellis wrote a paper a few years ago that long
term SIR testing could be cut down to an 8 hour test, but that is still
much longer than a 2 hour IC test.  Difficult to do as a process control
measure. I have found it easier to correlate IC results with SEC/ROSE than
with SIR.

*You know, I find it really weird that after being and SIR champion for so
long, I argue against it here.  I guess the point being that if you are
going to use ROSE/SEC as a process monitoring tool, which several companies
do successfully, you have to do extensive correlation studies between
product life testing (dozens of ways to do it) and any quality control
measures you implement.  You have to know what your monitoring signal means
in order to properly set the target and upper control limits.  Most
existing ROSE/SEC levels in specifications are both antiquated and bogus.

I would suggest that you retain the use of SIR equipment as a
process/quality monitor. Taking sample coupons at each stage of the
production process and running short tests will (probably) give you the
necessary indicators if the process or process materials are changing.

*This is possible, but I would say that you need a more focused test coupon
than the B-24 and one that is fabricated with your set of materials.
Additional cost in doing this, but the data is more relevant.  Maybe use
the **fanfare** CONCOAT TEST COUPON!!!!  (Look, up in the sky, its a bird,
its a plane...)

To ensure that you have used the correct test protocol, I assume that you
took on board the recommendations as published by the EU project partners
NPL, Siemens, NMRC and Lares Cozzi. If anyone wants a copy of this work,
let
me know.

*What a clever way to do some academic spamming <grin>. Sayyyyyyy, wasn't
that the work where the particpants used **fanfare** THE CONCOAT AUTOSIR?
(Notice how I very slyly slip in some gratuitous advertising for my friend
across the pond.  Ain't I clever?).

*Graham refers to a consortia activity going on for the last three years or
so looking at continous monitoring SIR vs. various flux and processing
conditions.  I wasn't aware that this was a published report yet.  I have
invited Dr. Chris Hunt of NPL in the UK to give a presentation on this work
at the SIR task group meeting at the Fall IPC meeting.

By the way, running IPC-TM-650 did they test to Appendix D of J-STD001 (or
is it now B? Doug??) or did they run to 6.2.3. or 6.2.3.3? What was the
coupon they used? What was the pitch/width they used? What was the test
voltage used?

*I'm sure Graham means methods 2.6.3 (cyclical environment) and 2.6.3.3
(static environment).  The SIR protocol that Graham refers to was found in
Appendix D of J-STD-001, Revisions A and B.  Revision C is now the current
version and the protocol is now Appendix B (lost two appendices somewhere).

*Graham, in his last two sentences brings up another good point - there are
so many possible choices in an SIR test (or IC test for that matter), that
you really need a guide in selecting the best factors for any particular
situation.  And where would you FIND such experts?  Why, at the upcoming
Fall IPC meeting in Orlando, Florida.  (notice the sly advertising for the
IPC...) Usually you can get a few of them to guide you along if ya buy the
beer (well, at least ONE such expert I know of...).

Doug Pauls
Rockwell Collins
(Well into his third Mountain Dew of the day)

---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt delivery of Technet send the following message: SET Technet NOMAIL
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------