Mr. Dore,
You ask some very broad and complex questions. I'll give you my slant on
some of them. It sounds as if you are my Irish counterpart. CSL does much
the same work as you describe, helping customers with process related
problems, often in Frantic Mode. Ok, Ok, no more blatant commercials!
In the US, during the downsizing frenzy of the 80's and 90's, process
engineers or process scientists were seen as "overhead" since they supported
all programs, rather than being tied to any one program. Consequently, the
bean counters (the natural prey of the engineer in the wild) viewed them as
unnecessary and they got the axe. When the sweeping environmental
regulations came into place, companies did not have the seasoned process
professionals in place to help them, which was a general boon to consultants.
Couple that to the normal turnover in professional personnel, with seasoned
vets retiring and new engineers coming in, companies not spending as much on
process related training, etc., and you can see why many manufacturing
process are reactive rather than proactive. This is especially the case for
medium to small size companies that cannot afford the large overhead of a
process dedicated staff. In addition, many companies do not budget monies
for process troubleshooting or process support, so when an emergency does
arise, you may have the situation of undertrained personnel trying to fix an
undetermined problem with no money. I have talked to some pretty frantic
engineers in my time, although these calls are often the most gratifying for
me.
>We offer prompt attention to companies who panic in the face of emerging
>defects and perform the required tests in our Electronics Manufacturing
>Centre.
>The main tests include;
>SIR testing, Ionic Contamination testing, Electromigration testing,
Microsectioning
>Flux corrosivity testing, etc
>which are all implemented into helping develop new processes, such as
>No-Clean, Paste Evaluation, Cleanliness, BGA Analysis etc
You have an impressive array of services to offer. I am curious about your
Ionic Contamination Testing. By this, do you mean Resistivity of Solvent
Extract (ROSE) or to use the European equivalent, Solvent Extract
Conductivity (SEC)? In my view, this form of ionic testing has limited
utility, except in the hands of a very experienced operator, when trying to
troubleshoot a process problem. We use ion chromatography extensively for
our analysis of ionic residues. The science is costlier, but you get a
wealth of information.
>
>I would appreciate if you could forward your opinion on whether companies
>should monitor key elements within the process on a continuous basis instead
>of waiting for the defect to occur before taking action.
Should they? Absolutely. Do they? Rarely. Sometimes due to budget
constraints, sometimes due to "managerial" influences. There's never enough
time to do it right, but always enough time to fix it. Process monitoring is
the heart of the statistical process control concept, and commercial assembly
level documents are going to rely on SPC more and more as time progresses.
So companies will have to implement process monitoring actions, or be left
behind.
>Are companies ignoring the importance of such process monitoring tools in
>your experience?
Not so much ignoring process monitoring tools as the lack of adequate tools
for them to use. Consider the cost of AOI. Cost effective mostly for large,
high throughput assemblers. Consider the venerable Omegameter or Zero Ion.
Many people are now aware of the pitfalls of using these class of
instruments as analytical tools or using them on non-rosin flux technologies.
For those who are not aware, they should get a copy of EMPF's Report
RR00013, "An In-Depth Look at Ionic Cleanliness Testing". To my knowledge,
there are very few process monitoring tools that are amenable to the shop
floor environment. There are lab analytical tests, such as ion
chromatography, UV-Vis, FT-IR, etc., that can be used to analyze process
residues, but they usually take at least an hour for extractions, and more
for adequate analysis. Add to this the cost of the equipment, the need for
highly trained people to interpret the results, and you can see why the tests
are only in labs or affordable in house by the larger manufacturers. There
are efforts under way by a number of individuals and companies (some
confidential) who are working on developing tools for process monitoring at
the shop floor level, but these are probably a year away, at best.
We have counseled our customers, who are transitioning to a no-clean
technology, to benchmark the cleanliness of the bare boards, and to monitor
the bare board cleanliness. In the no-clean world, the quality and
reliability of your end product are held hostage to the incoming cleanliness
and quality of the bare board. You have no opportunity to remove fabrication
residues in a no-clean assembly operation. In some cases, our no-clean
customers have moved their big in-line aqueous cleaner to the start of the
assembly process, to remove that big unknown. Some customers have implemented
bare board monitoring using ion chromatography, others have not, but usually
from a cost standpoint, rather than ignoring the risks. We have one high
volume no-clean assembler with a one-year monitoring program in place and it
has furnished us with a good idea of the variations expected from a good fab
house and a poor fab house.
>We have a few companies who control their process using our services. Why in
>your mind are the rest not?
The reasons are probably as varied as the companies you might poll. Some
cannot absorb the cost, some are not aware of the issues, some (foolish
people) do not believe it necessary, some are wary of dealing with screwballs
from central Indiana, etc. I spend a good portion of each week talking with
process engineers, discussing residue issues, cleanliness issues, reliability
issues. Many "process" engineers are really electrical / mechanical /
industrial engineers thrown into a position. Only a handful of Universities
in the US offer true "process engineering". Carnagie-Mellon and Georgia Tech
offer such programs. Most engineers have to sort out the fine points of the
process by reading the trade journals, attending conferences, etc. Most of
these data sources do not really address troubleshooting issues, because, as
we all know, YOU might have problems, but here at XYZ Corporation, WE have a
0% scrap rate and NEVER have problems. (Wanna buy a bridge?!)
On the other hand, some companies have worked very hard to qualify their
board vendors, to control their internal manufacturing processes, using JIT
and SPC concepts, and by proactively working with suppliers and their own
personnel, have a good process that is in control. There is less need for
monitoring in such a case.
>Would the downtime in production be significantly decreased if process
>checks were carried out on a continuous basis.
That is a hard one to answer, from my perspective. I think that the impact
would be greater from a product quality standpoint than from a production
down-time. An analogy would be an automobile. Preventative maintenance,
done when convenient to you, is less expensive and less stressful than having
emergency work done when you are out in the middle of a dark nowhere, with
the rain / sleet / snow falling (typical Indiana March). And you never break
down in front of a garage with an honorable, gifted mechanic!
Process monitoring, and all the hours and $$$ it represents, must be a
proactive action, requiring some foresight and planning. How likely is that
in your situation? Otherwise, you set yourself up for the reactive case,
with a downed production line and the high stress and large loss of money it
represents. In this case, you may have faulty products that have already
been completely assembled. Stress factor doubles. You may have products
which have already shipped. Stress factor triples.
Well, time to get down off my stump and go do some work for my employer. He
can be so unreasonable at times!. If you (or anyone else for that matter)
would like to discuss any of these matters in greater depth, please feel free
to contact me.
Doug Pauls
Contamination Studies Laboratories
201 East Defenbaugh St.
Kokomo, IN 46902
(317-457-8095)
[log in to unmask]
|
