IPC-D-279 EN.pdf - 第25页

dif ferent locations within an electronic assembly are typi- cally at dif ferent temperatures. Matching the CTEs at a joint is no guarantee of freedom from the problem because electronic components are their own heat sou…

inside the land area. If a ‘‘pads-only’’ approach is not fea-

sible, then it is crucial that solder mask be applied for sur-

face mount designs to act as a dam to solder migration.

The ‘‘pads-only’’ approach has several advantages worth

noting:

• the construction is compatible with soldering pro-

cesses, conformal coatings and common cleaning sol-

vents.

• The electrical and dielectric performance is no differ-

ent than the remainder of the P/I structure. Stenciling

of solder paste is improved.

• Application of solder mask to ﬂexible and ﬂex-rigid

P/I structures may be difficult.

• Resolution of this approach is the same as the resolu-

tion of copper etching.

3.3.11 Components with Reduced Clearances (Traces

Under)

Excessively thick solder mask, particularly dry

ﬁlm over traces under components with small clearance,

can contribute to the formation of crevices which entrap

ﬂux. In the case where the solder mask touches the bottom

of the component, if insufficient solder paste is used, it may

result in chip component draw-bridging (tombstoning),

insufficient solder ﬁllet or lack of solder joint. (See IPC-

SM-782 and IPC-D-275.)

3.3.12 Components with Reduced Clearance and Open

Vias

Components with reduced clearance should not be

over open vias, particularly if wave soldering is used. The

vias allow ﬂux up under the component to be entrapped in

the reduced clearance space.

Heat dissipaters should generally be ‘‘upstream’’ of tall

components to avoid recirculation cells.

Where lower junction temperature T

is required, consider

the following avenues particularly applicable to SMT: ther-

mal vias, thermally conductive adhesives from component

to printed board, and power and ground planes included in

the thermal design. In addition, consider the following

moves: sensitive components ‘‘upstream’’ of power dissi-

paters, power dissipaters further apart to reduce power den-

sity, power dissipaters closer to cold wall (edge of card if

card clamps are used), and power dissipaters ‘‘upstream’’

of tall components.

Long, tall components such as connectors are ideally

placed parallel to the airﬂow. Placement of these compo-

nents perpendicular to the airﬂow results in the generation

of recirculation cells which reduce the heat transfer from

heat dissipaters or which increase heat transfer to heat sen-

sitive components.

3.4 Coefficient of Thermal Expansion (CTE) and CTE-

Mismatch

Electronic assemblies consist of a multitude of

elements of different materials. These materials have differ-

ent coefficients of thermal expansion (CTE), of which some

are listed in Appendix G. Some CTEs also vary with tem-

perature. Because of external temperature variations and

internal heat generation and different thermal resistances,

IPC-279-05

Figure 3−3 Spacing Between Parts

1.25 mm

0.63 mm

1.5 mm

0.63 mm

1.0 mm 1.0 mm

1.25 mm

1.5 mm

July 1996 IPC-D-279

different locations within an electronic assembly are typi-

cally at different temperatures.

Matching the CTEs at a joint is no guarantee of freedom

from the problem because electronic components are their

own heat sources and because there is a temperature differ-

ence between component and substrate. The problem is

directly related to the size of the component, the thickness

of the solder joint and the compliance of the lead. The

major problem arises from thermally induced cyclic stress

in the solder joint of the leadless ceramic chip carrier

(LLCCC) components. Some components are very easy to

assemble, test and repair; others are not. The availability of

a component, its performance data, its degree of testability,

its reliability and its compatibility with manufacturing and

assembly processes and equipment inﬂuences component

selection.

3.5 Solder Joint Reliability The reliability of the surface

mount solder attachment of components to the printed cir-

cuit boards depends on a number of different parameters.

Some of these parameters are under the direct control of

the designer, some can be inﬂuenced by the designer, and

some are beyond the control of the designer.

Also, some of these parameters have a very strong inﬂu-

ence on reliability. Relatively small changes or inaccura-

cies in these primary parameters have a large impact on the

reliability. The impact of these parameters can be seen

from the reliability prediction Equations A-3 and A-4 in

Appendix A, Solder Joint Reliability.

3.5.1 Primary Design Parameters The following are

design parameters that have been identiﬁed as having a

primary (order of magnitude) inﬂuence on SM solder

attachment fatigue reliability.

3.5.1.1 Component Size The physical size of the com-

ponent determines the amount of displacement a solder

joint experiences during thermal expansion/contraction of

the component and the substrate to which it is soldered.

Larger components are larger threats to reliability. The

component size is determined by the number of I/Os and

the pitch.

3.5.1.2 Attachment Type The choice of attachment type

(leadless or leaded) determines the maximum stress level

that can be experienced in a solder joint during thermal

cycling. The stiff leadless attachments typically stress the

solder beyond the yield strength, whereas compliant leaded

attachments typically do not. This choice determines the

reliability model that needs to be applied and affects the

statistical failure distribution (Weibull slope). Leaded

attachments provide larger reliability margins, which

increase with decreasing lead stiffness. It needs to be noted

that lead compliance can vary greatly depending on the

lead geometry. Lead stiffnesses as high as 1100 N/mm for

SM connector header leads and as low as 1.5 N/mm for

ﬁne-pitch S-leads have been determined.

3.5.1.3 Solder Joint Height The solder joint height

determines the strain level experienced in the solder joint

for a given component/substrate displacement. It results

from the solder-ﬁlled gap between the component metalli-

zation or component lead and the substrate pad. Higher

solder columns reduce the strains in the solder joints and

increase reliability. Solder joint height is not the height of

the ﬁllet. For leaded attachments the solder joint height is

not a measurable quantity; it has been deﬁned in terms of

the solder paste stencil thickness.

3.5.1.4 Solder Joint Area Solder joint area determines

the stresses in a solder joint resulting from a given compo-

nent/substrate displacement. It is of importance primarily

for compliant leaded attachments. Larger solder joint areas

reduce the applied stresses and increase reliability; how-

ever, the possible range of effective increase in area is very

limited.

3.5.1.5 Lead Stiffness The lead stiffness determines the

forces resulting from a given component/substrate dis-

placement. Because the corner solder joints experience the

largest displacements and the displacements are in the

direction of the component center (neutral point), it is the

diagonal lead stiffness which is of primary importance.

Lower lead stiffness results in increased reliability.

3.5.1.6 Coefficient of Thermal Expansion The linear

coefficient of thermal expansion (CTE or α) represents the

change in linear dimension of a material due to a change in

its temperature. Components and substrates consist typi-

cally of a variety of materials all having different CTEs; the

effective CTEs are a combination of the individual material

CTEs and typically are different in different directions of

components and substrates. CTEs need to be measured to

avoid possible large errors in the reliability predictions. See

Appendix G for CTE values of different materials.

3.5.1.7 CTE - Mismatch The CTE-mismatch (∆CTE or

∆α) is the difference between the coefficients of thermal

expansion (CTE) of two materials or parts joined together;

in most instances it is the CTE-mismatch between compo-

nent and substrate, the global expansion mismatch, that is

most important, while the CTE-mismatch between the sol-

der and the materials to which it is bonded (ceramic, alloy

42, Kovar ), the local expansion mismatch, plays a smaller,

but not negligible role. In some designs (ceramic compo-

nent on ceramic or silicone substrate), this local CTE-

mismatch assumes primary importance. Large CTE mis-

matches pose large reliability threats; the effect of power

dissipation within the component makes CTE matching not

IPC-D-279 July 1996

the optimum solution (see A-2.3 and A-3.6 for explana-

tions).

3.5.1.8 Cyclic Temperature Swing The cyclic tempera-

ture swing (∆T) of components and substrate is the differ-

ence in the maximum and minimum steady-state tempera-

tures experienced during either externally (daily) imposed

temperature variations or operationally (on/off, load ﬂuc-

tuations) imposed variations. It needs to be realized that the

temperature swing of the components is typically not the

same as the temperature swing of the substrate due to the

power dissipated in active devices. Smaller ∆Ts result in

improved reliability. It needs to be noted that for some

applications the temperature swings during transport and

storage prior to operation can be more severe and a bigger

threat to reliability than the operating conditions.

3.5.1.9 Cyclic Expansion Mismatch The cyclic expan-

sion mismatch, ∆(α∆T) results from the difference in the

thermal expansion of components and substrate which are

determined by the respective thermal expansion coeffi-

cients (CTE) and cyclic temperature swings (∆T). Smaller

expansion mismatches result in improved reliability.

3.5.2 Secondary Design Parameters While the effects

of secondary design parameters are, by themselves, of

second-order importance, their additional contribution to

the effects of the ﬁrst-order parameters can be signiﬁcant.

The effect of some of these second-order parameters might

be different in accelerated testing and actual operational

use. The effect of these secondary parameters is indirectly

included in the reliability predictions of Equations A-3 and

A-4 in Appendix A, Solder Joint Reliability, is the ‘‘non-

ideal’’ factor, F. This F-factor is empirically determined.

Design parameters having second-order effects on solder

joint reliability are as follows:

3.5.2.1 Solder/Base-Material CTE-Mismatch The large

CTE-mismatch (∆α) between the solder and some base

materials (ceramic, Alloy 42, Kovar, silicon) can make sub-

stantial contributions to the cyclic fatigue damage (see

A-2.3 and A-3.6).

3.5.2.2 Solder Joint Shape/Fillet/Volume Experimental

evidence indicates that solder joint shape/ﬁllet/volume

have only secondary importance for reliability. In some

highly accelerated tests cyclic life improvements of about

a factor of two have been achieved, but it is not clear

whether even these small beneﬁts would result for the

slower conditions prevalent in most product operations.

The improvements result from the time necessary for crack

propagation through the ﬁllet.

Stress concentrations, e.g. from solder-mask-deﬁned sol-

dering lands for ball grid arrays (BGAs), can reduce the

solder joint fatigue life by as much as a factor of three

depending upon the severity of the loading conditions.

3.5.2.3 Solder Joint Uniformity Some experiments in

which solder joints were loaded primarily in a stress-driven

mode (high cyclic frequencies, no hold times, very large

temperature swings with fast transitions) showed the need

for extreme uniformity of all the solder joints of a compo-

nent to avoid unequal stressing; accelerated tests utilizing

test conditions more closely resembling product use condi-

tions did not reveal a need for extraordinary solder joint

uniformity.

3.5.2.4 Initial Solder Joint Grain Structure Aﬁne initial

grain structure in solder joints results in cyclic life

improvements of about a factor of two in highly acceler-

ated tests. The grain structure of solder is inherently

unstable and will grow with time. Higher temperatures and

cyclic loading accelerate the grain growth. Thus, for most

product applications a ﬁne initial grain structure will not

result in a signiﬁcant improvement of fatigue life; the sol-

der joints of accelerated test vehicles should be artiﬁcially

aged to start the tests with more product-related grain

structures.

3.5.2.5 Conformal Coating Conformal coating can have

different effects on solder joint life during thermal cycling

depending on the type of material, thickness, and location.

The advantage of conformal coating is that it slows the

absorption of water and oxygen into surface cracks. The

presence of oxides on the cracked surfaces may accelerate

the crack propagation. Oxidation layers prevent ‘‘re-

welding’’ of the solder during crack closure.

On the negative side, conformal coating may add another

material with a very high thermal coefficient of expansion

which may inﬂuence reliability. This addition can be sig-

niﬁcant especially if the coating wicks under components,

ﬁlling the gap between the printed board and component.

In addition, conformal coating can become rigid below the

glass transition temperature. This condition can exert con-

siderable stress on the components and solder joints during

the thermal cycles.

Because of the large variation in conformal coating mate-

rial properties, thickness applied, methods of application,

etc., the effect of conformal coating, in general, needs to be

evaluated empirically for each application.

3.5.2.6 Compliant Substrate Surface Layers Compliant

layers at substrate surfaces can provide additional reliabil-

ity margins, but by themselves are not adequate to counter-

act the effects of large expansion mismatches.

3.5.2.7 Solder Composition The most widely used sol-

der compositions are eutectic (63/37) and 60/40 tin-lead

solder. Solder compositions other than these can have

July 1996 IPC-D-279