Infineon-Package Top Side Cooled Package For High Voltage application-ApplicationNotes-v01 00-EN

AN_2101_PL52_2103_112902
Innovative top-side cooled package solution for

high-voltage applications
Application considerations for best performance
About this document

Scope and purpose
This application note shows the benefits of using top-side cooled (TSC) power devices in high-voltage (HV)
applications. In addition, it should help designers of such applications to understand how the device can be
used and how an efficient and easy-to-assemble approach can be chosen to integrate TSC devices into the
system. This application note describes Infineon’s new heat-spreader dual small outline package (HDSOP)
family, which consists of TSC surface-mount devices (SMDs) designed for HV applications.
The document explains the advantages of TSC and shows the different assembly methods and thermal stacks
and their influence on thermal performance. It also addresses the different manufacturing challenges a
hardware designer must consider when using TSC packages.
The following information is a guideline for use and should not be regarded as a description or warranty of a
certain functionality, condition or quality of the device.
Intended audience
System and hardware development engineers in the field of high-power/HV applications such as onboard
chargers, SMPS, drive applications, etc. who want to understand the advantages and challenges of the TSC
approach or who already use TSC in their applications. The application note is aimed at the automotive and
industrial space.
Application Note Please read the Important Notice and Warnings at the end of this document V 1.0
www.infineon.com page 1 of 30 2021-09-30
Innovative top-side cooled package solution for high-voltage
applications
Table of contents
Table of contents
About this document ....................................................................................................................... 1
1 Introduction .......................................................................................................................... 3
2 Advantages of top-side cooled devices ..................................................................................... 4
2.1 Definition of “thermal performance” ..................................................................................................... 4
2.2 Comparison to through-hole devices ..................................................................................................... 5
2.3 Comparison to bottom-side cooled devices .......................................................................................... 5
3 HDSOP package family............................................................................................................ 8
3.1 Package family overview......................................................................................................................... 8
3.2 Positive standoff...................................................................................................................................... 8
3.3 Reliability of HDSOP packages ............................................................................................................... 9
4 Assembly methods ................................................................................................................ 11
4.1 Thermal interface materials ................................................................................................................. 11
4.1.1 Gap pad ............................................................................................................................................ 11
4.1.2 Liquid gap filler................................................................................................................................. 11
4.1.3 Phase-change material .................................................................................................................... 12
4.2 Experimental results ............................................................................................................................. 12
4.2.1 Description of the test PCBs ............................................................................................................ 12
4.2.2 Thermal performance results .......................................................................................................... 14
4.2.3 Interface thickness ........................................................................................................................... 16
4.3 Interpretation of the measurement results.......................................................................................... 18
4.4 Improve thermal performance by reducing mechanical tolerances .................................................. 19
4.4.1 Applying a permanent force with pre-cured gap pads ................................................................... 19
4.4.2 Reduce warpage of PCBs ................................................................................................................. 19
4.4.3 Reduce tilting of packages ............................................................................................................... 20
4.5 Insulation considerations ..................................................................................................................... 21
4.5.1 Maintain a minimum vertical distance ............................................................................................ 23
4.5.2 Reinforced insulation – use of additional insulation foil ................................................................ 24
4.5.3 Creepage and clearance optimization ............................................................................................ 25
5 Conclusion ........................................................................................................................... 27
6 References ........................................................................................................................... 28
Revision history............................................................................................................................. 29
Application Note 2 of 30 V 1.0

2021-09-30
applications
Introduction
1 Introduction
TSC devices are surface-mounted power devices that are soldered onto a printed circuit board (PCB). The
generated heat of the semiconductor die is extracted to the top side of the package to the attached coldplate.
Thereby the thermal path is decoupled from the electrical connections and the PCB itself. The top side of the
package has an exposed pad to dissipate the generated heat. Figure 1 shows schematically the TSC package
approach assembled on a PCB and attached to a coldplate.
Applying the TSC approach also brings various challenges to be solved on a system level. If the mechanical
challenges are addressed, the TSC system is a superior device concept that can help to solve system problems.
With TSC designs the power density is significantly increased, power dissipation is improved, manufacturing is
simplified and electrical performance is improved.
In reality, not all TSC packages will have the same distance from package top to coldplate due to
manufacturing decisions. Therefore, the height tolerances must be balanced with some kind of thermal
interface material (TIM).
In the following chapters the TSC package family will be introduced and possible assembly methods are
explained. An experiment was done to show the thermal performance of different assembled boards.
Figure 1 TSC package concept

2021-09-30
applications
Advantages
Applicationofconsiderations
top-side cooledfor
devices
best performance
2 Advantages of top-side cooled devices

2.1 Definition of “thermal performance”
In the context of this application note, “thermal performance” is measured as thermal resistance (Rth) in Kelvin
per Watt (K/W). The lower this value is, the better the thermal performance.
Generally, the thermal resistance per area of the interface material depends on its thermal conductivity and the
thickness of the material. The contact resistance is an additional contributor, which is material and surface
dependent.
Consequently, the equivalent Rth can be summarized as:
𝑅𝑡ℎ = 𝑅𝑐𝑜𝑛𝑡𝑎𝑐𝑡 + 𝑅𝑡ℎ_𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙_𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 (1)
The contact resistance (Rcontact) is material dependent and can be reduced by applying a higher force between
the coldplate and the power package. This will improve the conformity of the TIM to the contact surfaces and
thus reduce the contact resistance. Nonetheless, a high force on the semiconductor could stress the solder
joints of the semiconductor components and therefore deteriorate the reliability of the joints.
Note: For the Infineon TSC package the effect of long-term force stress on the package has been
analyzed thoroughly – the results of force stress vs. temperature-cycling onboard show that a
force of up to 100 N has no visible impact on the reliability up to 3700 temperature cycles. Details
of this investigation are available in Chapter 3.3.
The second part of the equation (Rth_material_thickness) depends purely on the conductivity and the thickness of the
material, and can be represented as:
𝑇𝐼𝑀_𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠
𝑅𝑡ℎ_𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙_𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 = (2)
𝑇𝐼𝑀_𝑡ℎ𝑒𝑟𝑚_𝑐𝑜𝑛𝑑
A graphical representation of these formulas is given in Figure 2.
Rth vs. thickness of TIM

3
Rth in K/W
1
contact resistance
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Thickness of thermal interface material in mm
Figure 2 Thermal resistance (Rth) vs. thickness (with constant contact resistance)

2021-09-30
applications
Advantages
top-side cooledfor
devices
best performance
2.2 Comparison to through-hole devices

Through-hole devices (THDs) like the TO-247 or TO-220 package are still dominant in a number of applications.
However, these devices suffer from high manufacturing costs, as they need to be manually inserted into the
PCB and soldered onto the bottom side via wave soldering or manually. For this reason, through-hole
technology is increasingly being replaced with surface-mount technology (SMT), a standard process in the
electronics industry that has several advantages. The automatized placement of SMDs results in higher
throughput combined with increased reliability, which in turn reduces system cost. More details on the
differences of THT and SMT are described in [3] and are therefore not investigated further in this application
note.
2.3 Comparison to bottom-side cooled devices

Both TSC and bottom-side cooled (BSC) devices can be assembled with automated pick-and-place machinery.
The TSC package family has many advantages compared to a BSC SMD package variant (like D2PAK or DPAK).
But first, it is important to distinguish between BSC on FR4 PCB and insulated metal substrate (IMS) board.
Common cooling approaches are shown in Figure 3.
heatsink / coldplate
a) b) c)
Figure 3 Common cooling concepts for SMD devices: a) BSC approach on FR4 PCB with thermal vias; b)
BSC design on IMS board; c) TSC with double insulation approach (gap filler and insulation foil)
When comparing the performance of the TSC vs. BSC on FR4 PCB, the thermal performance of the TSC is
improved, as the thermal resistance of the thermal stack from chip to coldplate is reduced. PCBs are not
optimized for heat conduction and create a substantial thermal barrier for a SMD device placed on such a
board. The designer must improve the thermal dissipation by placing thermal vias in the PCB. But even with
optimized thermal design, the thermal resistance is higher in comparison to TSC, as the results from 3D FEM
simulations indicate in Figure 4. The graph shows that the thermal resistance from junction to coldplate is
reduced by 30 percent when comparing BSC on FR4 PBC to TSC. In addition, the routing on the PCB is much
more challenging for BSC, as large areas of the PCB are lost to thermal dissipation design elements. By
decoupling the electrical and thermal paths, as is done with TSC, the PCB design becomes much easier and
more flexible. Potential gains as a result of PCB area reduction and EMC noise mitigation can be expected.

2021-09-30
applications
Advantages
top-side cooledfor
devices
best performance
8
6.5
6
Rthj-c [K/W]
4
2.5 2.3
2
0
BSC on FR4 - D2PAK-7L
BSC on IMS - D2PAK-7L
TSC - double isolation - QDPAK
Figure 4 Thermal simulation of Rthj-c (junction to coldplate) for different cooling concepts: BSC on Fr4
PCB with D2PAK-7L, BSC on IMS with D2PAK-7L and TSC with double insulation (gap filler: 5.1
W/(m·K), 200 µm, insulation foil: 0.46 W/(m·K), 70 µm)
When comparing BSC on IMS to a TSC solution, the difference in thermal performance is not the key
differentiator. The simulation also shows that for a single-layer IMS board with 50 µm copper similar thermal
results can be achieved as for TSC with gap filler material and insulation foil. The main advantage when
compared to BSC on IMS is the reduced PCB cost and reduced space required. The IMS board is thermally very
accommodating, but also quite costly. With the TSC solution, the expensive IMS can be replaced by a lower-
cost FR4 PCB solution while maintaining the very low thermal resistance path from junction to coldplate.
As the IMS board is so costly, only the high-power components are placed on the IMS board, whereas all other
electrical components are customarily placed on standard FR4 PCBs and then connected to the IMS. Because
these connections are unnecessary when using TSC, the switching cell-loop design can be improved, as both
sides of the PCB can be used for routing. The gate driver, for example, can be placed directly above the switch
on the other side of the PCB. Therefore, the gate- and power-loop parasitics are reduced, resulting in better and
more precise gate control and, in turn, less ringing, higher performance and a smaller risk of failures.
One major advantage of TSC compared to BSC on IMS or FR4 PCB is the increased power density. The usable
PCB area is doubled for TSC as the coldplate is not directly attached to the PCB and the heat is not conducted
through the PCB. This is indicated in Figure 5, where the increased power density is shown schematically. By
using TSC packages, one additional layer of the PCB can be used to populate devices, virtually doubling the
power density of the assembly.

2021-09-30
applications
Advantages
top-side cooledfor
devices
best performance
Figure 5 Power density increase of TSC approach

2021-09-30
applications
HDSOP package
Application family
considerations for best performance
3 HDSOP package family

3.1 Package family overview
Infineon is introducing the HDSOP package family for HV devices above 600 V. In Table 1, the already released
packages in this family are summarized, with the key package information. One parameter they all have in
common is the package height of 2.3 mm. This standard height makes it possible to mix different discrete
devices, such as a 650 V SiC diode and LV and HV MOSFETs and attach them to one common coldplate. The TSC
packages of Infineon’s LV MOSFETs are not included in the overview in Table 1. The assembly process is
optimized and the distance from the coldplate to the package top side is the same for all power devices
populated on the PCB. This reduces the efforts of assembly, and a special structuring of the coldplate is not
necessary.
Table 1 HDSOP package family details

Package identifier Package name Drawing
PG-HDSOP-10-1 DDPAK
PG-HDSOP-22-1 QDPAK
PG-HDSOP-22-2 QDPAK
3.2 Positive standoff

The TSC HDSOP family for HV applications comes with the industry standard of a positive standoff, shown in
Figure 6. The package body is up to 150 µm higher than the lower edge of the package leads. One benefit of the
positive standoff is that a standard stencil height for reflow soldering can be used. The device only needs to be
picked up and placed on the PCB, before the devices can be soldered via reflow soldering. With a positive
standoff, no extra board cleaning is required before the SMD device is placed on the PCB for reflow soldering.
This enables direct contact between the pins and the solder stencils.

2021-09-30
applications
HDSOP package
Application family
The package can also be fixed with an adhesive material (e.g., SMD glue) to the board, enabling a rigid assembly
and double-sided component soldering.
Figure 6 The QDPAK (HDSOP-22-1 package) offers a positive standoff of 150 µm (nominal)
3.3 Reliability of HDSOP packages
Infineon’s application note “600 V CoolMOS™ G7 and 650 V CoolSiC™ G6 come in a new top-side cooling
package – the DDPAK” [1] shows the results of reliability investigations done for the HDSOP-10-1 (DDPAK),
which is one member of the HDSOP package family. In Chapter 4.3 of this application note, the temperature-
cycling on board (TCoB) was studied for the DDPAK. TCoB tests were performed for devices which were
soldered on four-layer PCBs. No electrical failure or optical abnormalities were observed for over 2000
temperature cycles [1]. In addition to the TCoB tests, some mechanical tests were also performed to
investigate the compression reliability of the DDPAK (Chapter 4.4 of [1]). Devices were soldered onto a PCB and
pressure was applied and successively increased until device failure. With this test the maximum allowed force
can be achieved. No failures were observed for forces up to 2500 N.
By applying a pressure force on top of the device, the leads act as springs and the package body is pressed
down to the PCB level. This unique feature of the device results in very high mechanical stress resistance [1].
Similar investigations were also performed for the QDPAK package to prove the reliability for other members of
the package family. Two different investigations were performed in parallel. One set of boards was prepared
with devices soldered onto a standard FR4 PCB. Another set of boards was additionally stressed with 100 N per
package. A pressure force was applied on top of the devices after soldering onto a PCB. Afterward, TCoB
investigations were conducted for both setups to identify the influence of an applied force on the TCoB
robustness of the packages. Depending on the chosen coldplate, a force can be necessary to press the device
against the coldplate. Detailed information on the different thermal stacks is given in Chapter 4.1. In Figure 7,
the solder joints of both variants are shown. It is clearly visible that the additional force on the package does
not change the solder joint connection, and no stress is directly imprinted. The leads of the package act like
springs and take the applied force. The lead is deformed in the shoulder region of the package. The subsequent
device investigations showed that the solder joints after temperature-cycling are more stressed than the
devices without any load applied. Nevertheless, no failures were observed for either setup for TCoB cycles for
far more than 2000 cycles, proving very high reliability and mechanical stability for both; devices without load
but also with 100 N applied. This very good TCoB reliability can be attributed to the spring-like behavior of the
package.

2021-09-30
applications
HDSOP package
Application family
Figure 7 Solder joint without and with applied force of 100 N: no impact on solder joint is visible

2021-09-30
applications
Assembly methods
4 Assembly methods
This chapter shares best practice examples of how TSC packages can be attached to a common coldplate. The
design goal is to maximize the thermal performance while ensuring proper electrical insulation. Therefore, an
interface material between the exposed metal pad of the HDSOP power packages and the heatsink is needed.
This TIM has the task to conform well to the power package and the coldplate and thus avoid any air voids.
These air voids cause several problems. First, air is a very bad thermal conductor and thus the thermal
performance will deteriorate. Second, air voids are problematic in terms of insulation and creepage
requirements and reliability. Therefore, voids must be avoided as much as possible to achieve high
performance and long-term reliability.
Note: Generally, attaching small TSC packages to individual heatsinks is relatively easy, because
mechanical tolerances do not need to be considered. Therefore, the individual heatsink mounting
for single packages is not covered here. Further information about individual heatsink mounting is
available in [1].
4.1 Thermal interface materials

Table 2 gives an overview of common classes of electric non-conductive TIMs.
Table 2 Thermal interface materials

Identifier Consistency Contact pressure Comment
Gap filler pad Solid (silicone or acrylic) High Can be fiberglass enforced
Liquid gap filler Liquid paste that becomes No pressure needed Often messy, but
solid after curing dispensation can be
automated
Phase-change materials Solid (at room Low Sticky material, changes
temperature) from solid to half-liquid
state at the specified
temperature to fill air gaps
Typical electric and thermal conductive materials are greases, which are excluded in this research as their
capacity for providing isolation is very limited.
4.1.1 Gap pad

Gap pads are pre-cured silicone and ceramic powder materials.
Gap pads have a high contact resistance because they do not conform well to their environment without
significant force being applied. Therefore, a permanent force needs to be guaranteed over their operating life
to achieve good thermal performance.
Gap pads exhibit higher operating temperatures, and they are also more stable than phase-change products.
4.1.2 Liquid gap filler

Liquid gap filler materials can be applied by imparting nearly zero force to the connected component.
They conform very well to different geometries while filling grooves, air voids and surface roughnesses due to
their liquid state, and thus achieve excellent thermal performance. These materials are cured in place to
become solid. State-of-the-art liquid gap fillers offer thermal conductivities of up to 7 W/(m·K) [14].

2021-09-30
applications
Assembly methods
4.1.3 Phase-change material

Phase-change materials are solid at around room temperature and become liquid as the temperature increases
during operation. Their advantage is that they offer much easier handling in production (because they are
solids when they are installed). They are available as isolating or non-isolating materials and need only a low
pressure for good thermal performance. The drawback of these materials is that they can only compensate
narrow mechanical tolerances (up to 100 µm).
4.2 Experimental results

The goal of the experimental measurements was to benchmark different assembly methods and different
materials against each other. The thermal resistance from the junction to the coldplate is used as a
benchmarking parameter.
4.2.1 Description of the test PCBs

The test boards contain eight QDPAKs under a common coldplate fixed with spring-loaded push-pins with
different interfaces in between (Figures 9 to 12). Electrically, each MOSFET can be driven separately via jumper
settings, as shown in Figure 8.
To have a defined gap between the component and the coldplate, distance spacers were plugged on the push-
pins for some boards, as shown in Figure 9. The vertical force on the QDPAK is a result of the board stack and
the springs chosen. The force in this setup was defined at 20 N per package and is calculated considering the
free height and the spring force.
Figure 8 QDPAK test board
Table 3 Test board configurations

Board Interface Manufacturer Type Thermal Comment
no. conductivity
4 Liquid gap Shin-Etsu SDP- 5.1 W/(m·K) Spring-mounted without distance
filler 5040A/B holders
18 Gap pad Henkel TGP HC 5000 5.0 W/(m·K) Spring-mounted without distance
holders

2021-09-30
applications
Assembly methods
Board Interface Manufacturer Type Thermal Comment

no. conductivity
24 Liquid gap Kerafol GFL 3020 1.8 W/(m·K) Spring-mounted without distance
filler holders
26 Liquid gap Hala TDG-U-SI-2C 3.6 W/(m·K) Additional mounting of isolation foil:
filler Hala dielectric film TFI-KMT-A2
(0.46 W/(m·K); 70 µm thickness)
Additional distance holder at
coldplate to guarantee minimum
distance
40 Liquid gap Henkel TGF 3600 3.6 W/(m·K) Spring-mounted without distance
filler holders
Figure 9 Mechanical mounting method with distance holders (TIM not applied yet)
Figure 10 Dispensing of liquid gap filler

2021-09-30
applications
Assembly methods
Figure 11 Assembly of board #26 (liquid gap filler with additional isolation foil and distance holders)
Figure 12 Final assembly of board #24 (liquid gap filler as TIM, spring-mounted coldplate without
additional distance holders)
4.2.2 Thermal performance results

Experiments were performed to obtain the indicator for the thermal performance from chip junction to
coldplate (Rth_j_cp) for each solution.
The temperature behavior of a P-N junction to determine the junction temperature of the semiconductor
switch [7] is used for the MOSFET body diode.
The first step is to calibrate the forward voltage of the body diode over the temperature in steady-state
conditions with a constant current source. This was performed by using a climate chamber to control the
ambient temperature while measuring the diode forward voltage. The forward voltage was measured for four
switches on each PCB to evaluate the variation of the Rth results.
Exemplary results for switch Q1 on PCB #40 are shown in Figure 13.

2021-09-30
applications
Assembly methods
450
400
350
300
Vf [mV]
250
200 y = -2,1538x + 463,57
150
100
50
0
20 30 40 50 60 70 80 90
Tamb [°C]
Figure 13 Diode forward voltage vs. temperature on test board #40
The second step was to measure the actual Rth_j_cp value. Therefore, a known power was dissipated within the
semiconductor switch while the coldplate temperature was measured in thermal steady-state. To ensure
constant environmental conditions of the coldplate, an enclosure was used to prevent significant convection.
Subsequently, the power was turned off and the voltage drop of the diode was measured immediately
afterward.
Next, the junction temperature of the semiconductor chip was determined using the calibration data.
The next step was to calculate the actual Rth_j_cp values. Table 4 and Figure 14 summarize these results. The
results are explained in Section 4.3.
Table 4 Rth [K/W] results – test boards

Board no. Rth_j_cp_averaged Rth_j_cp_Q1 Rth_j_cp_Q4 Rth_j_cp_Q7 Rth_j_cp_Q8
4 1.5 1.5 1.5 1.5 1.5
18 2.2 3.2 2.5 2.0 2.5
24 1.7 1.8 1.9 1.8 1.8
26 4.3 4.2 4.4 4.5 4.4
40 1.6 1.6 1.6 1.6 1.6

2021-09-30
applications
Assembly methods
Figure 14 Results of the lab experiment showing the average thermal resistance from semiconductor
junction to coldplate for different boards, and the Rth_j_cp measured for the different power
switches (Q1, Q4, Q7 and Q8)
4.2.3 Interface thickness

The thickness of the TIM is a key design parameter. The thinner it is, the better the thermal performance.
Therefore, it is important to minimize all mechanical tolerances in the Z-direction (which must be compensated
by the TIM eventually). On the other hand, the thinner the TIM gets, the less isolation it will provide.
Consequently, it is important to guarantee a minimum distance of the TIM by mechanical means (such as
distance holders) because this will also guarantee the minimum electrical isolation strength.
The thicknesses of the individual TIM layers of the test boards were analyzed via a 3D optical scan of the
surfaces. These results are shown in Figure 15, Figure 16 and Figure 17. Based on this measurement, the
maximum thickness of the TIM for the different test boards was extracted, yielding ~160 µm for PCB #24 and
~220 µm for PCB #40, ~270 µm for PCB #4 and 1.5 mm for PCB #26.

2021-09-30
applications
Assembly methods
Figure 15 3D optical surface measurement of the TIM of board #24
Figure 16 3D optical surface measurement of the TIM of board #40

2021-09-30
applications
Assembly methods
Figure 17 3D optical surface measurement of the residues of the TIM on the coldplate of board #40
4.3 Interpretation of the measurement results

The measurement results shown in Figure 14 illustrate that the best Rth_j_cp_avg values were achieved with board
#4, followed by #40 and #24, all using a liquid gap filler as the TIM. The height measurements showed that
board #4 has a gap filler thickness of about 270 µm. The thermal conductivity of the gap filler is the highest with
5.1 W/(m·K). Even though the gap filler thickness is not the thinnest measured in the experiment, the overall
thermal performance is best, which can be attributed to the very good thermal conductivity of the gap filler.
Aside from the absolute value it is worth mentioning that the spread between the cohabitating packages is
negligible.
Board #26 combines a gap filler with a dielectric foil. The height measurements revealed that the gap filler was
assembled to be very thick, about 1.5 mm. The thermal resistance of the thick gap filler and the additional
dielectric foil results in a total Rth_j_cp_avg of 4.4 K/W. The double isolation can be used to guarantee isolation
independent of void rate or gap filler height. Nevertheless, to achieve good thermal performance the gap filler
thickness in such a setup needs to be reduced, as the thick gap filler increases the thermal resistance of the
path from chip junction to coldplate.
Board #18, using a gap pad as interface, shows a notably higher Rth_j_cp_avg value. Also, the variation of the
individual Rth results for the individual transistors is significantly higher than the variation on the board using
liquid gap filler.
In conclusion, the liquid gap filler material shows better thermal performance than the pre-cured gap pad
arrangements. This can be explained by the higher conformity of the liquid gap filler materials, which reduce
the contact resistance (and therefore improve the total Rth_j_cp_avg).

2021-09-30
applications
Assembly methods
4.4 Improve thermal performance by reducing mechanical tolerances

A general recommendation in the cooling assembly is to minimize the mechanical tolerances as much as
possible to optimize the thermal performance, while guaranteeing a minimum distance of the TIM to guarantee
the isolation strength. Infineon is contributing to this optimization with narrow production tolerances for the
HDSOP family.
4.4.1 Applying a permanent force with pre-cured gap pads

A permanent force applied on the power semiconductors will reduce the mechanical tolerances, because the
standoff tolerances of the package leads are reduced. Hence, the thickness of the TIM can be reduced and the
thermal performance can be improved. Figure 18 gives an example of this technique: the left picture shows a
distance measurement from the surface of the PCB to the body of the TSC power package. A standoff of 96.2
µm is visible if no external force is applied to the power package.
The right side of Figure 18 shows the same measurement with an external force of 100 N. Here, the package
standoff (and therefore the distances in the vertical direction) is reduced to ~15 µm. As a result, the TIM can be
thinned by ~80 µm and the thermal performance can be improved as well.
Note: No impact to the thermal-cycling test was observed during reliability testing. This can be
explained by the spring-like behavior of the package leads. Details about this investigation are
available in Chapter 3.3.
Figure 18 Reduction of package standoff tolerances by applying a vertical force
4.4.2 Reduce warpage of PCBs

We recommend reducing warpage as much as possible to optimize the thermal performance. This can be
achieved by using stiff PCB materials or mechanical reinforcements (e.g., with metal carrier structures below
the PCB). Also, the dimensions of the PCB correlate with the warpage: the larger the PCB, the more PCB
warpage is to be expected. Therefore, it could be beneficial to use many distributed mechanical contact points
to improve the conformance of the PCB to the heatsink.

2021-09-30
applications
Assembly methods
Figure 19 PCB warpage (exaggerated). Recommendation: use as many distributed mechanical contact
points as possible (more screws).
4.4.3 Reduce tilting of packages

To maximize the thermal performance, the PCB surface should be as planar as possible to avoid any tilting of
the power packages. Cleaning of the board can be one way to avoid unwanted tilting by particles of the power
packages during assembly.
Also, tilting effects (such as “tombstoning”) of SMD components caused by the reflow process should be
avoided [13].
Figure 20 Tilting of SMD components caused by a non-planar surface below the QDPAK.
Recommendation: clean PCB of any particles before assembly. Reduce tombstoning effects in
reflow process.

2021-09-30
applications
Assembly methods
4.5 Insulation considerations

Besides achieving good thermal performance, it is important to ensure electrical insulation between the HV
potentials and the coldplate. Partial discharge events in particular can erode the insulation and eventually lead
to a complete breakdown of it.
Table 5 shows the simulated risk of of partial discharge with a power device attached via an insulating gap filler
material to the coldplate. The parameters that were varied were the gap filler thickness, the applied voltage
and the void diameter within the single-layer insulation. The risk is indicated as “low”, “medium” and “high”.
For the simulation, an analytical method was applied, which is based on a multilayer dielectric model.
The results show that a single-layer insulation with 300 µm thickness is sufficient for 600 V applications, even
for void diameters up to 200 µm. For 1200 V applications the single-layer isolation must be applied more
thickly, as 500 µm or more, to avoid the risk of partial discharge.
Note: The two-layer isolation must be at least 50 µm thick to be be considered as double-layer

insulation. If one layer is thinner (e.g., the coating), the results for the single-layer insulation
(Table 5) must be used!
Independent of these results, the chosen insulation thickness in the application must undergo a holistic
assessment, and not one single judgement. During the manufacturing process it is crucial that the void rate of
the liquid gap filler is monitored so a minimum void level can be guaranteed.
Table 5 Risk of a partial discharge event for single-layer insulation

Maximal voltage (V) TIM thickness (µm) Void diameter (µm) Risk (likelihood) of a
partial discharge
event
600 100 10 Medium
600 100 50 High
600 300 10 Low
600 300 50 Low
600 300 200 Low
600 500 10 Low
600 500 50 Low
600 500 200 Low
600 1000 10 Low
600 1000 50 Low
600 1000 200 Low
800 100 10 High
800 100 50 High
800 300 10 Low
800 300 50 Medium
800 300 200 Low
800 500 10 Low
800 500 50 Low
800 500 200 Low

2021-09-30
applications
Assembly methods
partial discharge
event
800 1000 10 Low
800 1000 50 Low
800 1000 200 Low
1200 100 10 High
1200 100 50 High
1200 300 10 Medium
1200 300 50 High
1200 300 200 Medium
1200 500 10 Low
1200 500 50 Medium
1200 500 200 Low
1200 1000 10 Low
1200 1000 50 Low
1200 1000 200 Low
Table 6 Risk of a partial discharge event for double-layer insulation

partial discharge
event
600 100 10 Low
600 100 50 Low
600 300 10 Low
600 300 50 Low
600 300 200 Low
600 500 10 Low
600 500 50 Low
600 500 200 Low
600 1000 10 Low
600 1000 50 Low
600 1000 200 Low
800 100 10 Low
800 100 50 Medium
800 300 10 Low
800 300 50 Low
800 300 200 Low
800 500 10 Low
800 500 50 Low

2021-09-30
applications
Assembly methods
partial discharge
event
800 500 200 Low
800 1000 10 Low
800 1000 50 Low
800 1000 200 Low
1200 100 10 Medium
1200 100 50 High
1200 300 10 Low
1200 300 50 Medium
1200 300 200 Low
1200 500 10 Low
1200 500 50 Low
1200 500 200 Low
1200 1000 10 Low
1200 1000 50 Low
1200 1000 200 Low
Double insulation, consisting of a combination of foil and gap filler (like board #4) improves the general
insulation robustness (as shown in Table 6). A thickness of 100 µm is sufficient for system voltages up to 600 V
peak. A thickness of 300 µm fulfills insulation up to 800 Vpeak and a 500 µm thick TIM can handle peak voltage up
to 1200 Vpeak safely. The void rate is also not so critical if a double-layer isolated approach is chosen.
4.5.1 Maintain a minimum vertical distance

The mechanical design should consider distance holders to guarantee a minimum distance between the
coldplate and the cooling plate of the power package (which is on HV potential). This could be achieved by
using spacers.
Additionally, liquid gap filler materials are available which ensure a minimum vertical distance with glass balls
(which are part of the material mix).

2021-09-30
applications
Assembly methods
Figure 21 Problem: a minimum thickness of the compressible TIM is needed to safeguard the required
insulation
Figure 22 Recommendation: use spacers to guarantee a minimum thickness of the compressible TIM
4.5.2 Reinforced insulation – use of additional insulation foil

The partial discharge simulation presented in Chapter 4.5 revealed that a single-layer insulation can be used
for HV applications only if the void rate and size can be controlled. The use of a double layer of insulation
reduces the risk of partial discharge significantly, as shown in Table 6. Therefore, an additional insulation foil
that is applied on the coldplate can be used to achieve more robust insulation. This approach is shown
schematically in Figure 23. With this approach the gap filler material can be applied as thinly as necessary
without the need for a minimal gap filler height to fulfill the insulation requirements.

2021-09-30
applications
Assembly methods
Figure 23 Recommendation: use additional isolation foil for a more robust insulation
4.5.3 Creepage and clearance optimization

Figure 24 shows that both creepage and clearance violations can be observed if the gap filler material is only
covering the top side of the power package. The distance from the package leads to the coldplate also needs to
be considered. To improve the creepage distance from the package leads to the heatsink, we recommend
applying the liquid gap filler or the insulation foil beyond the extent of the package leads. The insulation foil
could also be applied to the complete coldplate area, as Figure 25 shows. In this way the creepage and
clearance requirements from the package leads to the coldplate can be fulfilled.
heatsink / coldplate
compressible TIM
Semiconductor die
QDPAK
Figure 24 Location of clearance and creepage violation

2021-09-30
applications
Assembly methods
heatsink / coldplate heatsink / coldplate
incompressible isolation foil
compressible TIM compressible TIM

Semiconductor die Semiconductor die
QDPAK QDPAK
Figure 25 Recommendation: also apply gap filler over lead area or cover coldplate with isolation foil
Figure 26 Example: gap filler manually applied on exposed heat pad and package leads
Note: During curing the gap filler material may shrink. This can be ignored for spring-loaded coldplates,
but it is important that after application and curing of the liquid gap filler, the screws of the
connection between PCB and coldplate are retightened to ensure good contact between the
coldplate and the power device.

2021-09-30
applications
Conclusion
5 Conclusion
TSC power packages are a promising solution to improve thermal and electrical performance. These packages
also help to increase power density and reduce manufacturing effort.
One design target is to minimize all mechanical tolerances to achieve the thinnest possible gap filler thickness
for best thermal performance. This can be achieved by different methods, such as using a stiff PCB, pre-
cleaning of the board, and ensuring adequate mechanical connections between PCB and coldplate.
Investigations have shown that there is no negative influence on the long-term reliability when using HDSOP
packages even if a significant force is applied to the package. The impressed force is absorbed by the leads
because they act like springs, and the solder joint connections are not negatively influenced during TCoB tests.
Infineon recommends using a liquid gap filler with high thermal conductivity to compensate the tolerances
introduced by the mechanical assembly. The thermal contact resistance is lower for a gap filler compared with
a pre-cured gap pad, as shown in Chapter 4.2. Also, by using a liquid gap filler no force is needed to press the
package to the coldplate, which facilitates the assembly process.
Another important design goal is to ensure the required insulation. If single-layer insulation is used, a certain
thickness and a minimum void size needs to be guaranteed. In the case that a gap filler is used as the TIM, a
distance holder also needs to be used to guarantee a certain thickness of the compressible gap filler material to
ensure insulation. Infineon also recommends applying the gap filler over the whole package area including the
leads to improve the creepage from the package to the coldplate.
Greater insulation can be achieved by adding extra dielectric foil to the TIM (double-layer insulation). As shown
in Chapter 4.5, the risk of partial discharge can be reduced by introducing this foil. Especially for high safety
requirements or HV applications (with peak voltages reaching from 800 V to 1200 V), Infineon recommends
using double-layer insulation. However, this will reduce the thermal performance because it introduces
additional thermal resistance.
In summary, the TSC solution offers many benefits in application if the necessary mechanical challenges are
handled accordingly.
Infineon’s TSC package family provides the best solution to ensure successful top-side cooling in the
application.

2021-09-30
applications
References
6 References
[1] Infineon Application Note “600 V CoolMOS™ G7 and 650 V CoolSiC™ G6 come in a new top-side cooling
package – the DDPAK”.
[2] AEC – Q101: Failure Mechanism Based Stress Test Qualification for Discrete Semiconductors in Automotive
Applications. Rev. D1.
[3] Infineon Publication Additional Information “General Recommendations for Assembly of Infineon
Packages”, JEDEC Solid State Technology Association: IPC/JEDEC J-STD-020*. Moisture/Reflow Sensitivity
Classification for Nonhermetic Surface Mount Devices.
[4] JEDEC Solid State Technology Association: IPC/JEDEC J-STD-033*. Handling, Packing, Shipping and Use of
Moisture/Reflow Sensitive Surface Mount Devices.
[5] Infineon Guideline “Storage of Products Supplied by Infineon Technologies”. Revision 7.0.
[6] Infineon Application Note “Electrical safety and isolation in high voltage discrete component
applications and design hints”.
[7] Technical Paper “Using Forward Voltage to Measure Semiconductor Junction Temperature” by
Keithley Instruments, Inc.
[8] Technical Datasheet “BERGQUIST GAP PAD TGP HC5000” by Henkel.
[9] Technical Datasheet “GFL 3020 & GFL 3025 Gap Filler Liquid” by KERAFOL.
[10] Technical Datasheet “BERGQUIST GAP FILLER TGF 3600” by Henkel.
[11] Technical Datasheet “Gap Filler SDP-5040A/B” by Shin-Etsu.
[12] Technical Datasheet “2-PART SILICONE GAP FILLER TDG-U-SI-2C” by HALA.
[13] Technical Paper “Tombstone Troubleshooting” by EFD.
[14] Technical Information „Henkel TGF 7000“ by Henkel.

2021-09-30
applications
Revision history
Revision history
Document Date of release Description of changes
version
V 1.0 2021-09-30 Initial release

2021-09-30
Trademarks
All referenced product or service names and trademarks are the property of their respective owners.
IMPORTANT NOTICE
Edition 2021-09-30 The information contained in this application note is For further information on the product, technology,
given as a hint for the implementation of the product delivery terms and conditions and prices please
Published by only and shall in no event be regarded as a contact your nearest Infineon Technologies office
description or warranty of a certain functionality, (www.infineon.com).
Infineon Technologies AG condition or quality of the product. Before
81726 Munich, Germany implementation of the product, the recipient of this
application note must verify any function and other WARNINGS
technical information given herein in the real Due to technical requirements products may contain
© 2021 Infineon Technologies AG. application. Infineon Technologies hereby disclaims dangerous substances. For information on the types
any and all warranties and liabilities of any kind in question please contact your nearest Infineon
All Rights Reserved. (including without limitation warranties of non- Technologies office.
infringement of intellectual property rights of any
Do you have a question about this third party) with respect to any and all information
given in this application note. Except as otherwise explicitly approved by Infineon
document? Technologies in a written document signed by
authorized representatives of Infineon
Email: [email protected] The data contained in this document is exclusively Technologies, Infineon Technologies’ products may
intended for technically trained staff. It is the not be used in any applications where a failure of the
responsibility of customer’s technical departments product or any consequences of the use thereof can
Document reference to evaluate the suitability of the product for the reasonably be expected to result in personal injury.
AN_2101_PL52_2103_112902 intended application and the completeness of the
product information given in this document with
respect to such application.

Infineon-Package Top Side Cooled Package For High Voltage application-ApplicationNotes-v01 00-EN

Uploaded by

Copyright:

Available Formats

Infineon-Package Top Side Cooled Package For High Voltage application-ApplicationNotes-v01 00-EN

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Infineon-Package Top Side Cooled Package For High Voltage application-ApplicationNotes-v01 00-EN

Uploaded by

Copyright:

Available Formats

AN_2101_PL52_2103_112902

Innovative top-side cooled package solution for

About this document

Application Note 2 of 30 V 1.0

Figure 1 TSC package concept

Application Note 3 of 30 V 1.0

2 Advantages of top-side cooled devices

A graphical representation of these formulas is given in Figure 2.

Rth vs. thickness of TIM

Thickness of thermal interface material in mm

Application Note 4 of 30 V 1.0

2.2 Comparison to through-hole devices

2.3 Comparison to bottom-side cooled devices

Application Note 5 of 30 V 1.0

Application Note 6 of 30 V 1.0

Figure 5 Power density increase of TSC approach

Application Note 7 of 30 V 1.0

3 HDSOP package family

Table 1 HDSOP package family details

3.2 Positive standoff

Application Note 8 of 30 V 1.0

3.3 Reliability of HDSOP packages

Application Note 9 of 30 V 1.0

Application Note 10 of 30 V 1.0

4.1 Thermal interface materials

Table 2 Thermal interface materials

4.1.1 Gap pad

4.1.2 Liquid gap filler

Application Note 11 of 30 V 1.0

4.1.3 Phase-change material

4.2 Experimental results

4.2.1 Description of the test PCBs

Figure 8 QDPAK test board

Table 3 Test board configurations

Application Note 12 of 30 V 1.0

Board Interface Manufacturer Type Thermal Comment

Figure 10 Dispensing of liquid gap filler

Application Note 13 of 30 V 1.0

4.2.2 Thermal performance results

Application Note 14 of 30 V 1.0

Figure 13 Diode forward voltage vs. temperature on test board #40

Table 4 Rth [K/W] results – test boards

Application Note 15 of 30 V 1.0

4.2.3 Interface thickness

Application Note 16 of 30 V 1.0

Figure 15 3D optical surface measurement of the TIM of board #24

Figure 16 3D optical surface measurement of the TIM of board #40

Application Note 17 of 30 V 1.0

4.3 Interpretation of the measurement results

Application Note 18 of 30 V 1.0

4.4 Improve thermal performance by reducing mechanical tolerances

4.4.1 Applying a permanent force with pre-cured gap pads

Figure 18 Reduction of package standoff tolerances by applying a vertical force

4.4.2 Reduce warpage of PCBs

Application Note 19 of 30 V 1.0

4.4.3 Reduce tilting of packages

Application Note 20 of 30 V 1.0

4.5 Insulation considerations

Note: The two-layer isolation must be at least 50 µm thick to be be considered as double-layer