AN-652: Considerations for Soldering Accelerometers in LCC-8 Packages onto Printed Circuit Boards (Rev. 0) PDF

AN-652: Considerations for Soldering Accelerometers in LCC-8 Packages onto Printed Circuit Boards (Rev. 0) PDF
AN-652
APPLICATION NOTE
One Technology Way • P.O. Box 9106 • Norwood, MA 02062-9106 • Tel: 781/329-4700 • Fax: 781/326-8703 • www.analog.com
Considerations for Soldering Accelerometers
in LCC-8 Packages onto Printed Circuit Boards
By Hubert Geitner
INTRODUCTION
The ADXL78/ADXL278/ADXL193 are low cost, complete
single-axis or dual-axis accelerometers that can measure
both static acceleration like tilt or gravity and dynamic
acceleration like shock and vibration.They are available in
an ultrasmall ceramic leadless chip carrier package (LCC-8).
The ceramic package is hermetically sealed with a metal
lid and can be soldered onto a printed circuit board (PCB).
This application note discusses what to consider when
soldering an LCC-8 package to a PCB.
PCB DESIGN AND LAYOUT
Land pattern and pad sizes should be tested and qualified
in the manufacturing process. IPC, The Association Connecting Electronics Industries, created the Surface Mount
Design and Land Pattern Standard (IPC-SM-782). A recommended pad design is shown in Figure 1.
Symmetry and balance are important considerations for
the PCB layout and design. Traces should be connected
to the pads in a symmetric fashion. Asymmetric pads
or several connections to one pad would lead to solder
migration. Balanced trace entry to the pad minimizes
component rotation. Vias and through holes should not
be part of the pad.
Additionally, solder mask thickness has to be monitored if
the component is not placed into a solder mask window.
�
�
�
����������������
����������
����������
�����������������
����������
���������������
During soldering, there are several potential error sources
to consider. To minimize measurement errors, accelerometers need to point exactly to a known direction. Using the
recommended land patterns (Figure 1) will give a very close
alignment of the package to the given in-plane direction
from the layout of the land pattern.
Experiments conducted by Analog Devices showed that all
160 units of a population correctly self-aligned to the pads
after reflow even if the placement of the component was
fairly coarse (but within the limits of the castellation). The
surface tension of the liquid solder pulls the component into
the pad-given position. No visual rotational misalignment
was evident and measurements showed very good alignment performance of the LCC package (Figures 2 to 5).
RELIABILITY
An LCC-8 package has a ceramic base, so its coefficient of
thermal expansion (CTE) is different from that of FR4, a
commonly used PCB material. Leadless packages do not
have leads to take the stress of the CTE mismatch, so the
solder has to handle this stress. The larger the LCC package, the higher the stress. An LCC-8 package, however, is
very small (5 mm  5 mm). The effects of CTE mismatch
are much smaller than with larger LCC packages and are
more similar to other small ceramic components, such as
passive chip arrays.
Internal and external research (temperature cycling
tests with various temperatures, dwell times, and solder heights) showed that the reliability depends on the
solder thickness. A thickness of 4 mils is more reliable
than a thickness of 1 mil, but even the 1 mil solder height
showed very good performance. For improved reliability
in extreme applications, solder heights greater than 1 mil
are recommended. Uniform solder distribution contributes
to higher reliability.
�
�
�
���
���
���
���
���
���
��������������
����
����
����
����
����
����
����������������
����
���
���
����
���
����
Figure 1. LCC-8 IPC Land Pattern (IPC-SM-782)
REV. 0
SOLDERING
The LCC-8 is a leadless SMD package and is soldered
predominantly using reflow techniques. To avoid deterioration of the hermetic solder seal, the LCC-8 should not
be wave soldered.
Table I shows the number of failure-free cycles at different
solder heights and temperature cycles. A total of 128 sensors were used for temperature cycling tests.
AN-652
Table I. Failure-Free Cycles
Number of Cycles
Thermal Cycle
(°C)
Dwell Time
(Minutes)
1 mil
Solder Height
2 mil
Solder Height
3 mil
Solder Height
4 mil
Solder Height
+25/+35
60/60
624400
647900
668600
687700
–40/+105
60/60
833
1114
1399
1698
–40/+125
10/10
430.5
651
884
1131
–40/+85
60/60
1731
2109
2486
2875
–40/+105
20/20
950
1246
1546
1858
–40/+105
30/30
914
1206
1502
1810
0/+100
30/30
952
1335
1743
2179
Another reliability consideration is gold embrittlement.
The LCC-8 castellation (contact area) has tungsten as a
base material, nickel as the middle layer, and gold as the
top layer.
A mass of more than 2.5% gold inside the solder will
significantly change the Young’s modulus (C.J. Thwaites,
1973). Tests with a standard PbSn paste soldering profile
(Figure 2) showed that the mass contribution of gold
in the solder is 0.47% (Figure 6) at the given geometries.
The solder volume is based on an 8 mil stencil with
5% aperture reduction. This amount of gold does not
raise a concern for the reliability of the solder joint of an
LCC-8 package. Results from testing confirmed that there
is no significant difference between gold-plated and
pre-tinned units.
Figure 3. Package-to-Board Alignment (before Reflow)
SPEED ZONE 1 ZONE 2 ZONE 3 ZONE 4 ZONE 5
27.00
185
140
140
155
255
IN/MIN 185
140
140
155
255
250
DEGREES (C)
43C/sec
187
175
100
25
0000
0100
0200
0300
0400
0500
Figure 4. Package-to-Board Alignment (after Reflow)
Figure 2. Tests with Standard PbSn Paste Soldering;
The LCC Package Supports Standard Reflow Profiles
–2–
REV. 0
AN-652
90
25
WEIGHT % Au VS. IZOD IMPACT
STRENGTH FOR A 63% TIN-37%
LEAD SOLDER
80
IMPACT STRENGTH (in-lb)
FREQUENCY
20
15
10
5
70
60
GOLD % MAX CUTOFF POINT
50
40
GOLD % CALCULATED FOR
LCC WITH 8 mil STENCIL
AND 0.75 MICRON GOLD
THICKNESS
30
20
0
90.7
90.6
90.5
90.4
90.3
90.2
90.1
90.0
89.9
89.8
89.7
89.6
0
89.5
10
0
1
2
3
4
5
6
WEIGHT (% Au)
7
8
9
10
ANGLE (degrees)
Figure 7. Gold Embrittlement of Solder Joint
(Source: “Some Aspects of Soldering Gold Surfaces,”
C.J. Thwaites, Tin Research Institute, 1973)
Figure 5. LCC Angle Alignment after Solder;
The LCC Package Aligns Very Well to the Pads
of the Layout after Reflow Soldering
WEIGHT % GOLD =
WEIGHTAu
(WEIGHTSOLDER) + (WEIGHTAu)
=
1.58 ⴛ 10–5g
(3.30 ⴛ 10–3g) + (1.58 ⴛ 10–5g)
CONCLUSION
The assembly of the ceramic LCC-8 leadless chip carrier
package onto the FR4 printed circuit board has excellent
self-alignment capabilities. With the recommended land
pattern and solder height, the CTE mismatch of the
small package on the PCB has proven to be noncritical.
Gold-plated castellation did not lead to a level of gold
embrit tle ment that would change the Young’s modu lus and, therefore, the reliability of the solder joints is
very high.
ⴛ 100%
ⴛ 100%
= 0.47% GOLD
LCC-8 AFTER SOLDER REFLOW
ACKNOWLEDGEMENT
This application note is based on research and development results of a multidisciplined group around Allyson
Hartzell and Brad Workman of Analog Devices Micromachined Products Division.
SOLDER THICKNESS
X
X
SIDE VIEW
FRONT VIEW
Figure 6. LCC-8 Percent Gold by Weight Calculation
REFERENCES
1. “Solder Pad Geometry Studies for Surface Mount of
Chip Capacitors,” K. Wicker, J. Maxwell, Technical
Articles, Surface Mount Capacitors, AVX Corporation,
EEC 1985
2. “Assembly Induced Defects,” J. Maxwell, Technical
Articles, Surface Mount Capacitors, AVX Corporation
3. “Surface Mount Zero Defect Design Check List,”
J. Maxwell, Technical Articles, Surface Mount Capacitors, AVX Corporation, 1993
4. “Some Aspects of Soldering Gold Surfaces,” C.J.
Thwaites, Tin Research Institute, Electroplating and
Metal Finishing, Sept. 1973
REV. 0
–3–
E03707–0–5/03(0)
© 2003 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective companies.
–4–
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement