Thermal Design 101: Why it’s Necessary and How to Get Started

All electronics produce heat during normal function. But without an efficient way to cool down, that heat can cause dielectric materials in the components to break down, resulting in faulty performance and, eventually, product failure.

14 September 2021

Good thermal design is only becoming more critical as electronics skew smaller and components denser, and high-power designs generate heat faster. Engineers must incorporate ways to dissipate the generated heat, while managing heat transfer between components, their interconnects, and printed circuit boards. This blog will discuss the importance of effective thermal management and some of the methods for implementing it.

The Basics of Heat Transfer

Thermal design is based on the basic theory of heat transfer and fluid mechanics: where there's a temperature difference, heat will transfer from the higher temperature zone to the lower temperature zone.

Heat transfer occurs in the following ways:

  • Heat conduction: The transfer of heat between substances that are in direct contact with each other, travelling through a solid material. As a substance is heated, molecules gain energy and vibrate; when a warmer molecule “bumps” into a colder one, it transfers its energy, and so on down the line.

  • Radiation heat transfer: Heat traveling in the form of visible and non-visible light; low-wavelength, non-visible infrared radiation can carry heat directly from warm objects to cooler objects.

  • Natural convection transfer: Thermal energy that naturally flows from hot to cold places, ie: hot air naturally rises, resulting in a continuous circulation pattern.

Each of these three methods of transfer can be used in cooling electronics, depending on the product and a variety of factors. In thermal management, one size does not fit all.

Engineers need to consider multiple pieces of the puzzle to determine the best thermal solution, including:

  • heat output of components

  • area of heat dissipation

  • direction of highest heat dissipation of each component

  • the heat absorbing medium

  • the surrounding environment

  • which applications the product is being used for.

Since thermal engineers are often not brought into the process until later stages, it’s up to designers to incorporate thermal management best practices early on. Doing so can prevent catastrophic product failures, maximize mean time between failures (MTBFs), and result in a highly reliable product.

Performing thermal evaluation in all phases of the design cycle will also ensure that any major issues are caught early, rather than requiring massive teardowns or setbacks later on. To get started, the first step is to establish specific thermal operating requirements to design toward. For example, you might set a target a 100°C junction temperature. In a successful thermal design, the junction temperatures of all critical devices in a system will be kept below their worst-case temperatures. In a successful thermal design, the junction temperatures of all critical devices in a system will be kept below their worst-case temperatures.

One Size Does Not Fit All

As mentioned above, there’s no generic, silver bullet thermal solution that will work for all designs. Product designs differ in all kinds of ways. Differences in factors like the manufacturing process, the materials used in packaging, and the device’s physical environment, may all have a significant impact on the required thermal design. For example, newer technologies and methods for creating additively-manufactured devices has brought about novel “3D printed” heat removal schemes that can be optimized for flow and conduction considerations.

On the other hand, high-performance devices that utilize components with greater power dissipation like workstations, gaming PCs or tower servers require more aggressive cooling technology including heat sinks, liquid-air hybrid systems, cold plates or pumps to circulate liquid coolant.

Methods and Products for Cooling

There are several common methods for cooling electronics:

  • Forced-air cooling: Fans are used to “force” air to flow through electronic devices or components, transferring heat from heat source to heat sink through a ventilator or ram air device.

  • Direct fluid cooling: Components are directly soaked in a fluid coolant.

  • Indirect fluid cooling: Components do not make direct contact with fluid coolant; cooling is instead carried out through heat exchangers or cold plates.

  • Liquid-air system: A high-performance alternative to heat sinks, liquid-air systems are generally comprised of a cold plate, a liquid-to-air heat exchanger, a pump to circulate single-phase liquid coolant through the cold plate and the heat exchanger, and/or a fan or blower that forces ambient air through the air-side of the heat exchanger.

It’s important to note that successful integration of commercially viable liquid-to-air cooling systems may be largely dependent on by system-level factors such as manufacturability. That is to say, the “thermally best system designs” and the “best manufacturable designs” may not be the same thing. Thermally optimal designs may dissipate more heat than practical designs, but their complexity and/or specialization creates a manufacturability gap that designers should consider.

Air Management Best Practices

A device’s heat will be carried by airflow to other surrounding devices, which will impact their temperatures. As a result, effectively managing the airflow to carry heat away from the device plays another key role in thermal management.

Below are several key air management guidelines:

  • Temperature-sensitive components should be placed in the area with the lowest temperature, such as the bottom of a product. Place as far from heat-generating components as possible.

  • Place components that run hot from high amounts of current towards the center of the design (power parts, microcontrollers, etc.) This allows the heat to diffuse out through the board, whereas if placed closer to the edges, heat would accumulate.

  • Space hot components away from each other.

  • Using more copper on the board will also help to dissipate heat. This can be done by increasing the copper that the board is built with, and the width of the traces.

  • Increasing board thickness as suggested above is also helpful for devices that run with high currents.

Putting it to the Test

Once the initial product design is done, it is extremely important to test under real-world conditions. This way, engineers can alter the design before it goes to market and ensure the final result provides the best possible performance and reliability. At the board level, heat transfer effectiveness will depend on materials and their thermal conductivity. At the system level, there are even more factors to consider. Will the device be used indoors or outdoors? Is it going to be outside in a rainy location? What will be there in terms of air filters, vents and openings? Be sure to consider all possible real-world aspects of the application and test for them as closely as possible.

Managing Heat During Manufacturing

That’s right – thermal management isn’t only for preventing problems during normal operation.

The manufacturing process itself exposes devices and components to potentially dangerous heat. For example, PCB fabrication requires both heat and pressure to laminate the layers together. In addition, the soldering process will heat the board up during assembly. Luckily, proper heat management will protect components and boards from defects caused by heat such as warpage, broken traces and component degradation. The manufacturer must follow best practices during board handling, inspection and testing. One important measure is to ensure no unnecessary stresses are applied.

For example, during PCB handling, it’s important for the board to be sufficiently cooled down to room temperature before exerting mechanical or vibration pressures. Otherwise, it may break and need to be replaced.

Conclusion

There is a lot to consider when it comes to thermal management design, but it’s extremely important to get it right.

Be sure to address the following questions as you finalize your design:

  • Will the layout support the expected current and power requirements?

  • Do you expect any future component changes that might alter the airflow and should be taken into account now?

  • Are there any sensitive components that could be impacted by being too close to high-powered and hot components?

  • Does the product comply with safety and quality standards – including industry, third-party or internal, company-specific standards?

A knowledgeable Electronics Manufacturing Service (EMS) provider can ensure that your design conforms to good manufacturing principles for the selected process and meets all required standards. PCI has more than 30 years of EMS manufacturing experience across industries as well as in Printed Circuit Board Assembly (PCBA) and box build.

Our expertise in Design for Manufacturing and Assembly (DFMA) methodologies ensures the smoothest and most cost-efficient production cycle possible with the fastest time-to-market. Contact us today to learn more about PCI’s full suite of capabilities.