The failure rate of an electronic device doubles with every 10°C increase in chip junction temperature.
Thermal Management Tips and Frequently Asked Questions
1. What are the most important factors that effect cost effective thermal management?
The following components comprise an effective thermal management solution:
• A good heat sink or heat pipe to remove heat from the power component.
• A high thermally conductive interface material.
• As thin as possible thermal interface material
• Void free interface material.
• Void free interface between the power device, interface material and the heat
sink (or heat pipe)
The order of the contributing factors of these components for a fixed power or heat generation device are:
• Heat sink design and proper airflow is the first and most critical factor in any
thermal management design. The component should be optimized for
the physical space limitation. Airflow and cost consideration are also
main factors.
• The thermal interfacial resistance between the heat sink, interface material and the power
device is at least 2-3 times more important than the thermal conductivity
of the interface material when the thickness is less than 3 mils.
• The most important factor of thermal resistance is the elimination of voids along
the interfaces. Voids along the interfaces contribute to more than twice
the interfacial thermal resistance than voids inside the thermal interface
material. It is critical that there is adequate interface material and thickness to accommodate
any warpage of the surfaces.
• Only through adequate flow of the interface material will one be able to eliminate
the trapped air between the interface material, power device surface
and heat sink surface.
• Higher bulk thermal conductivity of the thermal interface material is only an indication that
having a lower thermal resistance is a possibility. This does not always directly
correlate the same in actual performance. Testing is the only way to confirm
the performance.
• The thermal interface resistance related to phonon scattering and thus not
completely predictable by the thermal conductivity of each part of the
components. The compatibility factor in heat transfer is a very complex phenomenon.
2. Is the initial performance of thermal resistance a good indicator for long-term reliability and life of the power device?
• Not really.
• Sometimes interfacial stress due to thermal expansion over thermal cycling may
change the initial thermal resistance performance.
• When interfacial mechanical stress induces voids over time along the interfaces,
the reliability and life of the device will be shortened with the increase in
thermal interface resistance and junction temperature of the power devices. If the interfacial
stresses are high due to using high bond strength and high modulus adhesives they tend
to have more problems.
• Sometimes thermal greases will be “pumped” out of the interfacial areas and thus
create voids. This will induce higher junction temperatures and shorten
the life of the power device. Accordingly thermal gels with the same initial performance are
always more reliable than grease. A flexible, highly molecular weight pad
that properly flows and eliminates interfacial trapped air will perform the
same as gel.
• Thermal interfacial materials that remain flexible during all phases of the operating
conditions from cold to hot are more reliable than materials that are hard
and solid during cold conditions and become more fluid during
hot conditions.
• In general, a stress-free interface between the thermal interface material, the power device
surface and the heat sink surface is critical for long-term reliability.
3. When one thermal interface material has a measured higher bulk thermal conductivity than another, will it also have a lower thermal resistance?
• Not always, please refer to question #1.
• Only a direct thermal resistance measurement can confirm this.
• Only if the thermal interface material will do the following will the thermal interfacial resistance
be reduced.
• The thermal interfacial material must be induced to flow to remove any trapped air along
the interfaces and within the interface materials.
• The thermal interface material must be flexible and compliant from the lowest temperature to the
highest temperature to ensure there is no interfacial stress that can
generate voids.
• Some thermal interfacial materials are more compatible in heat wave transfer (phonon and
electronic) than others in relation to the heat sink and power device materials.
4. Is heat spreading or heat removal more important?
• Heat spreading is more important since it is the first part of thermal management.
• Heat removal becomes critical when the power generation is much faster than heat
transfer to the air.
• If the physical space allows for a larger heat spreader it is a better solution than
faster air circulation.
5. Is it good to spread the thermal interface material-compound over the complete area of the power device?
• No. By using this approach, it tends to induce the user to think that by simply
placing the device over the area will form good thermal interface with
low thermal resistance.
• No matter what kind of shape, size or the interface compound being used, the key
is to ensure FLOW of the compound that will force out any trapped air when
two surfaces are being mated together.
• The best proven method is to dispense a “star” shape pattern of material. The
thickness or quantity of the heat transfer compound should be enough that
you can observe compound oozing out of the perimeter. Remove excess compound
whenever possible.
• This flow should be observed either in the placement of the heat sink over the
power device or in the case of a phase-change pad, during the first power-up
cycle of the device.
6. What if I have a large area power module, how can I compensate for the height variations over the area containing multiple components?
• AIT has a patented solution with the use of a compressible phase change thermal interface pad when
it is placed between the module and heat sink.
• The compressibility of the thermal interface pad material allows for the first critical part of accommodating the
difference in height tolerance so there are no air gaps existing anywhere on the
module’s components.
• The second property of the phase-change thermal interface pad is that it is engineered for most
applications at 50°C to ensure the elimination of trapped air between each
of the components and the heat sink surface.
• The flow aspect is critical to ensure the best thermal interface performance that cannot be achieved
by just being compressible.
Previous page: Insulated Metal Substrate Copper Clads & Pre-Preg
Next page: Flexible and Advanced Circuit Substrate Materials