I will give you my 5c worth on 10B.3 and 10B.4, which by the way should come
before 10B.2, since typically vibration and mechanical shock is less of a
problem. My suggestions are in italics.
>
>10B.3	Identify reliability issues of plated-through holes and vias.			
Use Appendix B of IPC-D-279, Design Guidelines for Reliable Surface Mount
Technology Printed Board Assemblies, as background material.
>. Identify via characteristics
PTHs and PTVs have the function to electrically connect the various layers in
an MLB in the vertical direction through the MLB.  The greatest risks to
reliability are for PTH/PTVs with small diameter in thick MLBs (high aspect
ratio: thickness/diameter) and come from many excursions to soldering
temperatures particularly with high maximum temperatures, like manual
soldering and rework. The damage mechanism is driven by the high coefficients
of thermal expansion 
(CTE) of the resins in the dielectric layers of the MLB, particularly above
the glass transition temperature, compared to the CTEs of the copper in the
PTH/PTV barrels and the glass reinforcements restraining the thermal
expansion of the resins in the x- and y-directions. Thus, the coefficient of
thermal expansion CTE of the MLB in the z-direction is about 3-times the  CTE
of the un-reinforced resin. This large thermal expansion mismatch causes
large hydrostatic pressures around the PTH/PTVs.  
>. Define barrel characteristics and weaknesses
The PTH/PTV barrels are formed by drilling, smear removal, Cu plating
(catalysis & electroless Cu flash & electrolytic plating; or catalysis &
full-built electroless plating), and the lay-up as well as the lamination
process play a role too. In each of these processes it is important that
quality be maintained. The lay-up should produce laminates that are more
resin-poor than have too much resin, the lamination process needs to assure
complete curing of the B-stage layers, drilling needs to produce smooth hole
walls without the disruption of the bond between the reinforcement glass
fibers and the surrounding resin; smear removal needs to remove all resin
smear on the Cu surfaces to make contact with the CU barrel; all wet
processing needs to have assured removal of any air bubbles in the holes; the
Cu plating parameters and solutions need to fit the hole geometry to produce
uniformly thick, high-strength, high-ductility deposits in the hole; the
minimum plating thickness needs to be adequate for the assembly processes and
the severity of subsequent field use. 

The Cu barrels will be stretched during any increase in temperature, but
particularly to soldering temperatures, which are the highest temperature
excursions experienced by MLBs. If the Cu barrels have stress concentrations
due to non-uniform plating or rough hole surfaces, the plating thickness is
too low, or the Cu deposit has low ductility, fractures in the PTH/PTV
barrels can occur during processing or later in the field with the more
severe use environments. 
>. Detail issues on post separation
The PTH/PTV Cu barrels need to connect to internal signal, power, and ground
layers. These interconnects to the inner layers are subject to large radial
forces towards the center of the hole due to the hydrostatic pressures around
the PTH/PTV, which also stretch the PTH/PTV barrel. The hole drilled for the
PTH/PTV disrupts the continuity of the reinforcement glass fibers; thus the
epoxy is free to compress the PTH/PTV Cu barrel putting the inner layer
interconnects in tensile stress during temperature increases. 

If smear removal is inadequate, the metal interfaces have inadequate bond
strength, or the inner layer copper foil has inadequate ductility, post
separation--the disruption in continuity between the PTH/PTV barrel and the
inner layer--can occur.
>. Determine testing conditions to establish reliability characteristics
Testing is best carried out on coupons built into the product panels, since
effective testing would damage good product. The traditional test method is
temperature cycling mimicking assembly and/or use conditions--this methods
can be quite time-consuming. Newer test methods using resistance heating can
significantly reduce test durations, but are as yet not quantitatively
correlated to temperature cycling and product reliability.
>
>10B.4	Identify solder connection reliability issues.						
Use Appendix A of IPC-D-279, Design Guidelines for Reliable Surface Mount
Technology Printed Board Assemblies, as background material.
>. Detail solder joint formation
Solder joints are formed by creating a metallurgical bond between the
surfaces that need to be electrically (also mechanically, thermally)
connected to an intermediary metal--solder. For tin-lead solders this
metallurgical bond is produced by the formation of intermetallic layers
(solid solutions) of the base metal of the contact areas (copper, nickel,
Alloy 42) with a solder constituent, typically tin. The solder joint
formation requires heat to clean the surfaces to be soldered using flux, to
melt the solder on the soldering surfaces (solder coatings, solder paste), to
bring about the dissolution and dispersal of any surface finishes (gold,
silver, palladium, OSP), to bring about the dissolution of some of the base
metal to form intermetallic layers and a metallurgical bond. The amount of
heat required depends on the thermal mass of the MLB and the components as
well as on the metals of the contact areas--nickel and Alloy 42 surfaces
require more heat than copper surfaces. The amount of heat can be varied by di
fferent peak temperatures and dwells at these temperatures. A good rule of
thumb is 10 seconds at solder liquidus temperature (L)+20C for copper
surfaces and  L+35C for nickel and Alloy 42 surfaces; the 10 seconds need to
be at the solder joints of the most massive component in the assembly--the
solder joints of smaller components will see higher temperatures and longer
dwells. 
>. Explain substrate [CTE] thermal expansion mismatch related to components
The thermal expansion mismatch between substrate and component results from
the difference in the thermal expansion of the component (product of
CTE(component) x deltaT(component)) and the thermal expansion of the
substrate (product of CTE(substrate) x deltaT(substrate)). The coefficients
of thermal expansion (CTE) of the substrate and the components can be quite
different, e.g., a ceramic chip capacitor (CTE~6.5 ppm/C) or TSOP with Alloy
42 lead frame (CTE~5.9 ppm/C) on a FR-4 MLB (CTE(x,y)~17 ppm/C). Components
typically have larger deltaT's because the active components have internal
heat generation. As a rule of thumb, a good design has (CTE(substrate) -
CTE(component)) < 4ppm/C for the larger components.  
>. Define cyclic strain on solder joints
The cyclic strains in solder joints come as the result of temperature
differences during day/night operation, on/off cycles, and load variations
(mini-cycles). These cyclic strains cause fatigue in the solder joints which
can cause premature failures when adequate reliability was not assured with
appropriate design measures.
>. [Determine required solder volume to establish joint reliability] Describe
design measures to improve solder joint reliability (There is no such thing
as a required solder volume to establish joint reliability. However, the
designers need to be aware about the design measures they can take to
increase reliability)
For leadless SMT attachments the fatigue damage term is

deltaD = [component size x deltaCTE x deltaT(effective)] /solder joint height

and for leaded SMT attachments

deltaD = {lead stiffness x [component size x deltaCTE x deltaT(eff.)]
squared} /[solder joint height x solder joint area]

In both cases, the life of the solder joint is roughly proportional  to the
square of the inverse of deltaD. Thus, changes in the first-order importance
parameters in deltaT can have a large impact on the reliability of the solder
joints. 
Reducing deltaCTE, deltaT, lead stiffness, and component size improves
reliability, as does increasing the solder joint height (e.g., column grid
arrays). 

Werner Engelmaier
Engelmaier Associates, Inc.
Electronic Packaging, Interconnection and Reliability Consulting
23 Gunther Street
Mendham, NJ  07945  USA
Phone & Fax: 201-543-2747
E-mail: [log in to unmask]

***************************************************************************
*   The 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.*
***************************************************************************
* If you are having a problem with the DesignerCouncil, please contact    *
* Dmitriy Sklyar at 847-509-9700 ext. 311 or email at [log in to unmask]      *
***************************************************************************