-----Original Message-----
From:   Brian Ellis [SMTP:[log in to unmask]]
Sent:   Tuesday, May 01, 2001 2:15 AM
To:     TechNet E-Mail Forum.; Lou Hart
Subject:        Re: [TN] Process control for cleaning

Lou

Bad idea!
- Solder mask curing, in practice, is very variable.
- Flux removal, from board type to board type, is very variable
Good idea:
Do systematic tests on every board type and do your SPC on the results
of these, individually. On the instruments I used to make, we had a
separate file for each assembly and the SPC was automatic. Quite often,
we saw clients who had some board types trend up while simultaneously
other types trended down, even though the process was the same. This may
have been due to variations in bare boards, components, air quality or
possibly power voltage and, of course, human variations.

My advice is test all your board types, several times per day, if you
really want to do SPC.

Brian

[Lou Hart]  Brian, thanks for your comments.  I was originally not in favor
of our production manager's standard board approach, thinking that, as the
solder mask went through the wave time and time again, its properties would
change and introduce unknown and unpredictable variability into the
measurements.

A while back, working at our sister PC fab plant, we used SPC on bare board
cleanliness testing, and I was pleased with the results.  It cut down on
cleaning bath maintenance.  It also indicated that boards requiring test
with more than one fixture were getting contaminated from handling and led
to having the testers wear gloves when performing a multiple-fixture test.

Our product mix here is so variable and unpredictable, I hesitate to try to
do anything more complicated than something like test a board at the start
of the first shift, or at the start of every shift, or every hour, etc.

Board variability would be added to process variability, if we used only
one chart.  After reviewing TechNet responses, I am considering performing
a destructive measurement system evaluation as described in a paper by Don
Wheeler.  I've used Wheeler's methods before (his book on Evaluating the
Measurement Process is widely available) and think they are good
supplements, at least, to the standard gage repeatability and
reproducibility techniques.  (The Omegameter qualifies as a destructive
gage, since you can only make one measurement on each sample with it.)
 Such an evaluation would let me estimate gage variability and process
variability.  I'll have to figure out what to consider as part of the
process.  Do I take sets of assemblies as they come in the production
stream, pulling out a couple every day for a week?  Or do I take a
non-urgent set and run 2 of them per day, over a 1-week period, say,
through the wave and cleaner?  I'm inclined to favor the latter.

Thanks to all TechNet commentators.  If anything very interesting comes out
of this work, I will submit a report to IPC Review.

Lou

---------------------------------------------------------------------------------
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
---------------------------------------------------------------------------------