In a message dated 02/22/2000 4:25:54 PM US Eastern Standard Time,
[log in to unmask] writes:

> Probably its not a real problem, because...

Ohhhhhhh yes it is.  But, since I do problem solving in this area, perhaps
this is a self serving answer.

>  *       if you are cleaning your pwa's anyway, you will clean also the
>  components,

And how do YOU determine that both boards and components are sufficiently
clean at the end of the assembly process?  The boards may be fine, but lots
of junk may remain on the components.

>  *       if you use clean-free, you will take accept also all your "normal"
>  internal contaminants (flux, fingerprints ..) and most
>  component-contaminants will be at a similar level/kind.

I will disagree with Brian Ellis (and I thought the Swiss were supposed to be
neutrals) in that there ARE normal contaminants in a process.  We have our
"list of usual suspects" whenever we start to do a residue failure analysis.
You find chlorides, bromides, sulfates, etc., on practically all assemblies.
They are common.  Fingerprints and oils vary depending on the practices in
the assembly area.  Then you get into the process specific residues.

>  I like it, Dougs :"...YOU determine ..". hehe, that's exactely what you
have
>  to do, but instead of to write own standards I would refer to the golden
>  IPC-octuple (you don't know it? so what!: IPC-2221, -2222, -6011, -6012,
>  -4101, -A-600, -A-610, J-STD-001, all class 2, and IPC-D-279 + IPC-HBK-001
>  as guidelines).

I couldn't agree more.  If this were a perfect world, IPC would have all the
answers in their specifications.  But, I have noted, especially on my drive
to work this morning, that we do not live in a perfect world.

The IPC can only put into the specifications what the volunteer companies are
willing to put into it.  Then you have to get all the volunteers, or at least
a significant majority, to agree with the language and the requirements of
the specification drafts.  You need to either have companies donating their
internal research, or participating in IPC sponsored research.  Then you have
to get a majority of the general populous to agree that the requirements you
propose are to the benefit of the industry, and benefit to their company.
And this would be if we focused only on the technical aspects.  There are
political aspects in the process as well.

Lets say our goal is to generate a specification for how clean a hybrid
microcircuit has to be.

Most of my clients, and from some of the responses, many Technet members,
want a number, a single number that they can painlessly drop into their
purchasing specs:  "Hybrid Microcircuit XXX must have a combined chloride and
bromide level under 1.0 micrograms per square inch as determined by
IPC-TM-650, method 2.3.28."  If the components don't meet this level, then
the assemblers can go and do unspeakable things to the supplier.  This is
what everyone, especially those used to MIL-specs, WANT.  It ain't what they
will GET.  Here is why.

Now the fabricator comes back and asks "why do you need it this clean and why
use a method we have never heard of?"  While an answer of "Because Doug Pauls
says so" would be gratifying, it is hardly an answer to be used to control a
significant segment of the industry.

Perplexed by this situation, you start asking yourself other questions:
Do I need this level of cleanliness for my application?
Do I have differing cleanliness requirements depending on which of my
assemblies this part goes on?
Do I have differing cleanliness requirements based on end-use environment?
Can I apply my cleanliness requirements to similar hybrids, but which use
different materials?
Do I have to determine a necessary cleanliness level for all the components
in my inventory?
Do I need to examine all my suppliers to see if they can meet my specs?

This list could go on a long time, and could easily apply to bare boards,
components, assemblies, materials, etc.

Now, let us suppose you are one of the very rare people who have both the
time and the budget to investigate these questions.  You do so.  You do
extensive reliability tests and determine how clean your microcircuits need
to be, and in what applications, and how general you can make your spec.

Or, if you are a real masochist, you volunteer to be an IPC Committee
Chairman dedicated to this specific question and try to get other volunteers
to participate in "free" research with their "free" time.

You bring this information forth in a blaze of trumpets and the adulation of
the crowds.  You propose your cleanliness limits to go into a microcircuits
cleanliness specification.  Get ready for the following questions and
positions:

Who are you to determine cleanliness for me?
Why do I have to be this clean?  None of my other customers are having a
problem.  None of my other customers want this.
There is not enough research data out there, from a broad enough field to
make a specification.
I can't afford the testing you propose.  I vote NO.

Since the IPC works on a one company, one vote basis, one company that does
not agree with you and who does not want to spend any money, can easily
cancel you out.  As a grizzled veteran of the process, I can attest to this
going on.  Unless you can convince that the specification levels you propose
are of general benefit, and worth the pain to use internally (e.g.
VALUE-ADDED), you won't get anywhere.  If you occasionally get a sense of
overall frustration in my responses regarding J-STD-001, this is why.  I
don't fault the hardworking staff of the IPC.  It is inherent to the nature
of a volunteer organization.

About the best that you can expect is for the IPC to draft suggested
protocols or DOE structures on HOW a manufacturer should determine
cleanliness requirements for their products.  There are too many products,
with too many configurations, and too many combinations of materials, to ever
get a specification that covers them all.  So, as Joyce originally asked for
numbers she can use, I responded that it won't happen.

So where is this long diatribe leading?

We have IPC meetings coming up in March and April in Los Angeles and San
Diego, respectively.  If you, the technical community want to be able to use
IPC specifications and standards to improve the reliability, functionality,
and value of your products, you had darn well better be there, to participate
in the process of making the standards better and letting IPC know what you
need from them.  Be willing to offer company research.  Be willing to
participate in the discussions.  Not all can attend, but in this wired world,
you can still work on the committees without being at the semi-annual
meetings.  You can either help drive the industry or have the industry drive
you.  Keep in mind that the driver is the only one with a clear vision of
where he/she is going.

I hope to see you there.

How the hell did I get up on this stump???

Doug Pauls
Grizzled IPC Veteran

##############################################################
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 Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################