"Hollandsworth, Ron" <[log in to unmask]> wrote.......
> We are manufacturing PWBs using SMD, PTH, and Mixed
> Technology. Run rates are relatively high and cycle
> time of product flow is relatively fast. Boards range
> from double sided (minimal) to multilayer (most) in
> single, half panel, and full panel configuration. Some
> PWBs, obviously see IR, some flow solder, and some
> mixed technology both IR and flow solder. Some parts
> are commercialized some MIL spec. Our plant
> environment is somewhat controlled (air conditioned).
> Humidity is not high, yet somewhat humidity controlled
> (ESD reasons).
>
> We are currently observing no adverse affects at IR as
> a result of not baking PWBs prior to IR. We are
> considering eliminating pre-wave bake. The question is
> "What experience can you share from eliminating your
> bake cycle at pre-wave." Positive and negative will be
> well accepted and appreciated.
I have found three reasons why people bake PWBs. They are
1. Explosive delamination of multilayer PWBs
2. Measles, particularly on double sided PWBs
3. JiffyPop solder joints, and solder voids / pinholes
The first item is rarely seen in my factory, and I have not seen
more than a single incidence from a given lot of boards in
many years. This is usually a vendor lamination problem.
The second item seems to be associated with some of the
lesser expensive laminate, and might be avoided by ordering
higher Tg material. I have only a few double sided boards,
and the quality varies by supplier a great deal. I have found
that a lot of problems on the manufacturing floor correlate
inversely with the relative cost. I.e., the cheaper it is to buy,
the more costly it is to build with.
The third item is the most prevalent in my factory today. It seems
that plated through hole wall voids are still considered
acceptable by the PWB industry. At the same time, my
customer is looking for defect rates measured in ppm or ppb.
The classic solder joint created by out gassing associated with
pth voids is the JiffyPop joint, named after the pre-microwave
popcorn. I love finding either an un-broken dome of solder or
a dome that has barely opened to reveal the cavernous void
inside. I just hand the problem over to Product Quality Assurance
while sending the balance of the PWBs of to bake. I must say in
defense of the PWB vendors that the problem is less than 5 years
ago, but they are still behind what is consistent with what I need.
I have made elimination of pre-baking a part of my continuous
process improvement program. I have been at companies that
practiced a 16 hour overnite bake at 225 F dehumidifications.
These bakeouts added to poor wetting defects, particularly at
power and ground pins. I have tried to eliminate all avoidable
heatings of the PWB prior to soldering. This includes part
marking, dehumidification, and ruggedization. I try to bond
heatsinks with two-part epoxies that do not require long or hot
cures. Baked epoxy ink serialization is moved to pre-conformal
coat by using raw board serialization for tracking, or replaced
with stick-on labels.
Our PWB suppliers are required to package all product in
moisture barrier packaging with desiccant or in an otherwise
suitable, mutually agreeable manner. This has proven to be
a vital part of keeping our PWBs solderable.
Good Luck
Steve Mikell
SCI Systems Plant 13
[log in to unmask]
Steve Mikell, Process Engineer, Soldering & cleaning processes
SCI Systems Plant 13
[log in to unmask]
|