Page 24 - Chip Scale Review_November December_2021-digital

that matter), systems equipped with camera is not able to do so. to QFPs. This smaller body size results in
side-view cameras are required. An Thermal performance study. As with less area from which to conduct or convect
example of such AOI images are shown any semiconductor package, thermal heat. The thermal resistance from junction
in Figure 8. The side camera of this AOI performance is a critical characteristic. For to ambient air (θ JA ) is commonly used to
system is able to capture the light reflection MaxQFP, this is particularly so because it determine thermal performance of each
of J-lead solder joints, where a top-view effectively shrinks the body size compared package type and can be calculated by a
thermal simulation model. The procedure
is defined in the JEDEC JESD51-2A
specification [10] for a natural convection
environment. Following this method, the
θ JA values of a 172-lead MaxQFP, a 172-
lead MaxQFP_EP, a 176-lead QFP, and
a 176-lead QFP_EP are calculated and
compared in Table 4. These packages are
all modeled assuming a two layers of signal
and two layers of power (2s2p) thermal test
board consistent with that described in the
relevant JEDEC specification (JESD51-
7) [11]. These θ JA values are not meant to
predict the performance of a package in an
application environment on a product PCB.
2
Three die sizes (45, 50 and 55mm )
were modeled for all four package types
to calculate the θ JA values. As expected,
a larger die size can result in a relatively
Figure 8: An AOI image of a good solder joint where both J- and gull-wing lead joints are visible. This PCB used smaller θ JA value at the same body size.
a 0.125mm solder paste stencil thickness. This image required a side camera to capture the light reflection. Also expected, a package with an exposed
pad has much better thermal performance
than one without, because the additional
heat transfer area is considerable. The
surprising result here is that the 55%
smaller MaxQFP package performs better
thermally than the larger QFP package at
similar pin counts. Comparing same die
sizes, the improvement is about 16% for
MaxQFP and about 10% for MaxQFP_
EP. The reason for this performance
improvement is likely the high lead
count density (i.e., leads per area) of
MaxQFP compared to QFP. Again, this
Table 4: Thermal resistance value (θ JA ) comparison between MaxQFP and QFP at similar pin counts. was a nonintuitive and surprising result,
especially for the exposed pad package.
But, these results clearly indicate that
MaxQFP can fully replace QFP from
a thermal perspective and even offer
an advantage.
Electrical performance study. The
100-lead MaxQFP package (body size
10x10mm) was chosen for an electrical
performance study and its characteristic
results were compared with those of a
100-lead QFP (body size 14x14mm). Like
the thermal performance study discussed
above, this electrical performance study
was also conducted by simulation. The
die size used in the simulation was 4.25
x 3.91mm. The wire lengths ranged from
1.71 to 2.36mm for the 100-lead MaxQFP
part, and from 1.84 to 2.35mm for the 100-
lead QFP part. The direct current (DC)
Figure 9: DC resistance comparison between a 100-lead MaxQFP and a 100-lead QFP.

22
22 Chip Scale Review November • December • 2021 [ChipScaleReview.com]