Wilcox Tom wrote:
> 
> We are experiencing a high degree of warpage on several different
> multi-layer PWBs. The PWBs are typically 8 layer, .063 thick GFN material
> approximate size of 8"x10".  Warpage is on the order of > .20" across a 13"
> diagonal. The PWBs have 1" wide break-aways on two opposite outside edges.
> Break-aways are GFN with no copper layers. The copper layers in the PWB are
> 1/2 or 1 oz and are fairly well balanced around the mid-plane of the board.
> Numerous interstitial vias are used and all are located in the bottom half
> layers of the PWBs.
> 
> Not all PWBs we receive are warped. Recent runs show a 10% reject rate for
> warpage. It appears that some boards also warp after we receive them (i.e.
> Manufacturer says they past his inspection.) Some boards warp after reflow.
> 
> We are looking for any possible causes of the warpage. There is some thought
> that the unbalanced interstitial vias may be causing the warpage. Any
> suggestions as to cause of corrective action would be appreciated.
> 

While attending the PWB Design Conference in Santa Clara in March, this
subject came up during the High Speed Design program during a discussion
of the advantages and disadvantages of various board stackups. One of
the instructors (Lee Ritchey, 3 Com) indicated that in his experiences
in tracking board warpage problems, it always turned out to be a
supplier and/or vendor quality control problem and not a design problem
due to "asymmetrical stackup". 

If we look at this in more detail, there seem to be many "gotchas" in
the PWB manufacturing process which need to be effectively addressed to
ensure your PWB isn't warped.

Due to the nature of their construction and the processes used in
manufacture (heat and pressure), laminates exhibit a different expansion
factor in each of the three dimensions. Typical CTE values (PPM / deg C)
listed for an FR-4 epoxy fiber-glass laminate are: X (length) 12, Y
(width) 16, Z (thickness) 60. The linear expansion of a material is
calculated as the board dimension, times the CTE, times the temperature
change. Laminated materials undergo size changes during the board
manufacturing processes, some more than others. Excessive size change
contributes to board warpage and layer-to-layer registration problems.
The laminate size change is caused by the difference in CTE between the
base material and copper. In the laminate manufacturing process, the
materials are subjected to high temperatures and pressure in the
laminating press. As the copper clad laminates cool down, the resin
wants to shrink. However, the copper foil laminated to the surface has a
lower CTE and does not want to shrink as much.  The result is a locked
in stress between the base material and the copper foil. When the copper
foil is etched away, the residual stresses are relieved and the laminate
material shrinks.

To make up the various dielectric thicknesses required, multiple layers
of standard thickness prepreg have to be built up into copper-clad
laminates, with the right materials, in the right sequence, with the
right recipe. Without knowing any of the details of what your vendor is
doing, several things to look at would be:

Differential CTE mismatch in X and y dimensions, CTE mismatch due to
misorientation of prepreg materials, different and/or variable
prepreg/laminate suppliers, different moisture absorptions, uncontrolled
prepreg/laminate storage environment (temperature and relative
humidity), prepreg/laminate shelf life in uncontrolled environment,
incompatible prepreg formulations, etc.  

Due to the wide range of pathologic behavior you are expereincing, my
guess is that you are seeing some combination of the above factors. 

Hope this helps.


-- 

============================================================
John Nieznanski                     Eastman Kodak Co.       
[log in to unmask]                     2/4/EP MC35313          
PH: 716-726-2520                    901 Elmgrove Rd.        
FX: 716-726-0745                    Rochester, NY 14653-5313
============================================================

***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To subscribe/unsubscribe send a message <to: [log in to unmask]>   *
* with <subject: subscribe/unsubscribe> and no text in the body.          *
***************************************************************************
* If you are having a problem with the IPC TechNet forum please contact   *
* Dmitriy Sklyar at 847-509-9700 ext. 311 or email at [log in to unmask]      *
***************************************************************************