-- [ From: Doug Jeffery * EMC.Ver #2.10P ] --

>On 7/26/96 Klaus Giese wrote:

"PWB stretch or shrinkage seems to be the limiting factor for stencil 
printing of larger PWB's with fine pitch pad's. We use laser cut
stencils  and stencil printers with visual alignment but still have
problems with  larger PWB's."

Klaus, A few questions:

1.  Is the problem that you can line up the one end of the part but the
other end is off or is it the device pads are all over the place? 

2. How is the registration of the circuit pattern to the hole pattern,
is it within tolerance? 

3. Do you specify a tooling hole to pattern dimension on your drawing? 

4.  Is the hole to hole across the part distorted the same as the
artwork pattern?


Q. What is reasonable to expect from PWB suppliers?

The artwork image should be within size control of +/-.00025 in/in
average based on typical FR4 shrinkage information. However, this
number will vary based on glass style, mtl thickness, grain direction
and the copper pattern.  So, finding a good number for reasonable image
change from original data is a bit complex. However, over the length of
the part it should not be very much more than .0005 in/in.  Most PWB
fabricators understand this dynamic and compensate interanal layers for
it and should be aware if its effects external layers.  All PWB
fabricators probably verify that the artwork is within tolerances to
the drilling pattern.  (est. +/-.005") and this is done usually by
annular ring evaluation.

Does your supplier have Photo tool size controls in place?  This would
be the most likely source of distortion if the photo tool grew or
shrunk due to temperature and humidity changes.  If it did the artwork
to drill pattern should show this change in size.  If the fabricator
could not hold the registration they should have looked into the
artwork size.  

Is the original design data accurate enough for your design? I have
recently run into 2 casses where an assembler was have problems pasting
and placing fine pitch devices (.5mm, .0197").  The complaint was that
the pad to pad spacing wasn't equal.  When we reviewed the design data
we found that it was generated in 2.3 format. This would make the data
points for pad centers alternate between .019 and .020 and if the
plotting was done in that format  it could be off by another factor of
the plotting resolution used.   (usually .0005").  Any design using
millimeter to inch conversions for data generation should be taken to
the fourth decimal place (2.4 format gerber) to assure that the feature
locations are accurate.   When plotting designs using loaction
intervals of .0197" it is important that the photo tool be plotted in
.00025" resolution.  This will give you the least amount of variation
in the orignal tool.  Plotting in .0005" resolution will round between
.0195 and .020, plotting in .00025" resolution will round between .0195
and. 01975.

Q.  Does anybody know about a specification for PWB stretch?  

A.  Some of the IPC specifications give information about allowable
tolerances for the boards longest dimension, but I believe these are
for mechanical features.  You should be able to get a number from IPC-D-
275 for accuracy of circuit pattern to drill pattern.  

Q. I would like to ask  for a stretch / shrinkage tolerance depending
on PWB size and smallest pad's  on the PWB. What can be expected on FR4
boards without paying a premium?   What can be supplied at higher cost?

Let's find a root cause before we answer these questions.


Doug Jeffery
Electrotek, Inc
[log in to unmask]

***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to:           *
* [log in to unmask] with <subject: unsubscribe> and no text.        *
***************************************************************************