Thermal Management - how to get the most out of your PCB

My experience as an electrical engineer has taught me that designing and building an electronic device that works on your lab workbench and one that can reliably work in the “real world” for the whole duration of its specified lifetime, while being exposed to elements (water, moisture, dirt, heat, cold, etc.) are two vastly different things.

Reliability becomes very important when a device is being used in applications where failure can cause significant material damage, endanger human lives and regular service (repair) is not possible or expensive.

The goal of this blog is to explain effects of ambient temperature on electronic devices by addressing following questions:

• What are the negative effects of working outside of specified temperature range?
• What makes a good PCB thermal design?
• How to validate that your electronic device is up to specification and will work in the “real world” scenarios?

1. Negative effects of operating outside off recommended temperature ranges

Recommended operating temperature is the temperature range of the surrounding environment at which an electronic component is safe to operate.

Outside of this range component will exhibit derating of its electrical characteristics, reduced lifetime followed by imminent failure. Failure is most often caused by thermal runaway which occurs in situations where an increase in temperature changes the conditions in a way that causes a further increase in temperature.

Maximum junction temperature (TJ) on the other hand, defines the maximum internal operating temperature of the electronic component.

Equation 1: Relation between junction temperature (TJ), ambient temperature (Ta), dissipated power (PD), and thermal resistance between junction and ambient of the component (Rja).

Therefore, optimal operating conditions are achieved by mitigating following factors:

• Ambient temperature (Ta): Is reduced by maintaining sufficient airflow that prevents heat buildup in your electronic device assembly (use of fans and vents or natural convection).
• Dissipated power (PD): Components should be selected so that they work in a area of highest efficiency (minimize dissipated heat) as specified in manufacturer documentation.
• Thermal resistance (Rja): Most cost-effective way to reduce thermal resistance is to maximize the copper surface of an electronic component’s thermal pads. This helps generated heat to dissipate into the surrounding environment more efficiently.

Since thermal resistance (Rja) is dependent on PCB size, copper surface area and thermal connection between the PCB and component, the question that emerges is; On what PCB was the parameter from the datasheet measured? In most cases, it is either defined by JESD51-7 standard or evaluation board for that electronic component offered by the component manufacturer. Most of the time the size of copper surfaces on those boards is larger than what we can allocate on our PCBs. It is important to be wary of this when planning your thermal design.

2. Good PCB thermal design

There are many ways to improve PCB thermal design. Planning your PCB layout in advance and maximizing copper surfaces that can be used as a heatsink is the optimal solution from a cost to performance perspective. Don't get me wrong, there are designs where the use of specialized components such as thermal bridges or aluminum core PCBs is needed, but that is a topic for a future blog.

When planning your thermal design, good practices are:

• Maximizing copper surfaces to which electrical component (power) pads are connected. This includes pads and an exposed central pad if present. Utilize top, bottom, and internal copper layers if available.
• Adding vias (via stitching) to the copper surface increases surface area, which in turn reduces thermal resistance. The further away from the heat source the via is, the smaller the benefit (thermal spreading resistance).
• When vias are placed on solder surfaces, using a hole diameter of less than 0.3mm reduces solder wicking which can cause solder voids (air pockets) that have a negative effect on thermal resistance.
• Vias can be filled with thermally conductive materials (epoxy) to reduce thermal resistance. This process is called “plugged vias” and is used when placing vias in the pad of BGA components.

In addition to above-mentioned practices, standards like IPC-7351 and IPC-7525 also provide a lot of information on how to improve your component land patterns with good thermal design in mind (among other things).

3. Design verification testing

Last and most important step of successful thermal management is verification testing. Verification is performed in specialized climate chambers that can emulate real world temperature and humidity conditions.

Image 1: Climate chamber test setup. Includes data logging equipment so that the test device can be observed during the entire testing duration

The goal of those tests is to ensure that the device will operate within its specification when deployed in the target environment for a minimum duration of its specified lifetime. It is important that during verification the device operates in a regime that is representative of its real-world application.

Depending on the requirements verification tests include:

• Basic testing: Simulation of normal operation at varying temperature, humidity, and duration.
• Thermal cycling: Rapid transitions between specified minimum and maximum operating temperatures.
• Accelerated aging: In addition to above mentioned tests also includes: Salt-Spray tests (corrosion simulation), UV Light tests (sunlight), ESD tests, vibration tests, etc.

Since climate chamber testing can be expensive and takes a long time to set up, it's good practice to do initial measurements using a thermal imaging camera. Data gained from such measurements is a good indication of possible failure points that need to be addressed before we commit to climate chamber testing.

When using thermal imaging cameras, it is important to take into account that some electronic components have surfaces with low emissivity (WLCSPs, ferrite inductors, modems with metal shields, etc.) and require the camera to be configured correctly (emissivity factor) for those measurements.

Emissivity is the efficiency with which an object emits infrared radiation, compared to a perfect emitter (blackbody, which has an emissivity value of 1).

Image 2: Pictures show emissive object properties. Due to the metal surface, the measured temperature of the modem appears lower (note the temperature scale). When thermographic paint is applied, the measured temperature is accurate

Alternatively to adjusting the camera emissivity factor, we can utilize thermal conductive coating that would make the component surface more emissive. There are specialized thermographic paints like “LabIR”, but since they are aerosol sprays, they have shipping limitations (high shipping cost). FLIR (large thermal imaging camera manufacturer) suggests accessible alternatives that give comparable results: masking tape, kapton tape, electrical tape, correction fluid, etc…