Mr. de Rooij,

The specifications used most frequently in the U.S. to qualify a new
manufacturing process are J-STD-001A, Appendix D and F, and MIL-STD-2000A,
Appendix A and C.  Of the two, I recommend the former as sounder, although it
has it's drawbacks.

The real question for you, or anyone validating a new manufacturing process,
what will your customer accept as a valid test document or valid test data?.
 It does no good to qualify to J-STD-001A, if your customers require
MIL-STD-2000.  I have worked with numerous customers who would like to
qualify to J-STD-001A, but the customer is more comfortable with
MIL-STD-2000A.  Going the 2000A route is the path of least resistance.

What contractual obligations do you have?  They can indicate what testing you
need to do to implement a new process.  As above, my customers have been
faced with the J-STD / MIL-STD choice.  The J-STD is more cost effective
testing and can tell more about the process, but if the J-STD data was
generated, the customer had to convince all of his customers to accept J-STD
qualification data.  On the other hand, if the MIL-STD data was generated,
all he had to do was have the data on file and he could make the switch and
need convince no one.  Again, the path of least resistance.

How representative is your qualification vehicle (e.g. the B-36 board) to
your actual product?  The greater the leap between the B-36 and your product,
the murkier the link between qual data and reality.  Generating data on
actual boards and assemblies is more valid than using a standard test
vehicle, but interpretation of the results, relative to spec pass/fail
criteria, becomes an art form.

I must confess ignorance of how the IEC goes about qualifying a new
manufacturing process.  I would really like to know, if anyone can fill me
in.

If you are interested, last summer I wrote a document fondly titled " The
Poor Schmuck's Guide to Qualifying New Fluxes for MIL-STD-2000A or MT-0002".
 As the new process qualification procedures of J-STD-001A are identical to
MT-0002, the Guide would apply to J-STD-1A as well.  I have turned the
document over to Dave Bergman at IPC for IPC use.  Anyone who wants a copy,
ask Dave ([log in to unmask]) or myself ([log in to unmask]).  I also published an
article in the April 95 Circuits Assembly titled "Qualifying A New
Manufacturing Process".  That one must be requested from Circuits Assembly
(Ron Daniels).

I would not go the route of military specifications.  The U.S. Department of
Defense has issued edicts mandating activities to get out of the
specification writing business and have cancelled MIL-STD-2000A without
replacement.  Wherever possible, commercial specs are being adopted.  Field
personnel in the DoD are being converted or shot.  J-STD-001A is the present
front-runner, but the B revision is due out in a month or so, and the process
qualification protocol changes dramatically.

Well, time to get down off my stump and go back to work.  Let me know if I
can be of any further help.

Doug Pauls
Contamination Studies Laboratories
201 East Defenbaugh St.
Kokomo, IN  46902
(317) 457-8095
[log in to unmask]