Thermal Management of Power Supplies
Thermal management is critical for power supply design. The biggest threat to reliability and the long service life is heat, and thermal management strategies ensure that power supply components run smoothly for as long as possible.
To tackle thermal management effectively, you need to understand some essential concepts, like the bathtub curve, thermal resistance, airflow, and heat transfer methods. Let's look at what these concepts are and how they relate to power supply performance.
Why Is Thermal Management Important?
With power supplies, heat is not your friend. It leads to increased failure rates and shorter service life, leading to higher costs for both manufacturers and customers. When heat can't dissipate, the thermal stressors build up and cause deterioration. The device may fail sooner or more often, and performance might suffer.
Poor thermal management could result in complete failure of the device, but it could also cause effects like dimming or reduced power delivery and other interruptions. Managing heat can be complex and depends on various factors, like the operating environment, enclosure design and heat transfer methods applied.
What Is the "Bathtub Curve" and How Does Temperature Affect It?
The bathtub curve refers to the shape you'll see if you graph out the failure rate of a product over time. It starts high, dips down into a steady, low rate, and then climbs again at the end of the product's life, giving us a line that looks like a bathtub.
The first section of the curve where the failure rate is high is called the early failure period. These failures typically come from defects in manufacturing or installation issues. After this initial stage, the failure rate drops sharply and plateaus for most of the product's life. This stage is the normal service life phase, characterized by a constant low failure rate. This phase might be measured with metrics like the mean time between failure (MTBF) in hours, mean time to failure (MTTF), or failures in time (FIT).
MTBF is a rate calculated across a large number of products. If you're operating 1,000 units continuously for one year, they run for 8.77 million unit operating hours. Say you have an MTBF of 250,000 hours. You won't get 250,000 hours out of every unit. Instead, the MTBF rating applied to the 8.77 million operating hours tells you to expect about 35 failures among those 1,000 units within the year.
High temperature can significantly affect your MTBF and the bathtub curve. A common rule of thumb is that an increase of 10 degrees Celsius doubles your failure rate. That brings your failure rate in the above situation to 70 a year.
At the end of the bathtub, the curve is the end-of-life wear period, where the MTBF spikes quickly as components like aluminum electrolytic capacitors and fans start to wear out. Higher temperatures can significantly reduce the time it takes to reach this period.
Understanding Heat Flow and Thermal Resistance
Thermal resistance works a lot like electrical resistance, and we can use that comparison to better understand and visualize it. Thermal resistance quantifies how difficult it is for heat to be conducted between two points, similar to how electrical resistance assesses current flow. It can help us calculate values like thermal impedance or the sum of thermal resistance and all contact resistances.
We can even replace the parameters of Ohm's law with heat values. For example, we calculate electric resistance by dividing the voltage difference by the current. Similarly, we can calculate thermal resistance by dividing the temperature difference by heat flow.
These values match up well with an electric model. Two heat sources might connect through a heat-conducting path with a component such as an insulator. A corresponding electrical model could have two voltage sources connected along a circuit with a resistor.
Types of Heat Transfer Methods
Heat transfer typically occurs through one of three methods. These cooling strategies can help keep systems within certain temperatures and prevent the need for derating. Derating refers to operating a device at a reduced capacity to avoid high temperatures that negatively affect reliability and equipment life. Cooling methods can help remove heat and maintain appropriate temperatures without derating.
1. Convection Cooling
Convection cooling involves transferring heat into the surrounding air. It usually comes in one of two varieties:
- Natural convection cooling: Natural cooling relies on the fact that hot air rises. It lets air flow on its own to carry heat away. This method is free and doesn't create any noise, but it isn't always enough to cool hotter components.
- Forced convection: A system with forced convection uses a powered system, such as a fan, to remove heat. It adds noise but removes more heat. Forced convection systems must consider the complex interactions of air pressure and airflow restrictions due to enclosure design and electrical components.
2. Conduction Cooling
Conduction cooling is typically used for higher-powered modules and those requiring environmental sealing. A conduction system transfers heat to a thermally conductive surface, providing a path for heat flow. These thermally conductive surfaces might include:
- Thermal pads: These thermally conductive materials facilitate heat transfer between circuit board components and fill air gaps.
- Cold plates: Cold plates are flat pieces of metal with flow paths cut through the sides for a conductive liquid to move through. They absorb heat and help it dissipate into the fluid.
Some particularly high-power systems will use liquid coolants that run through the system to draw heat away. Liquid-cooled solutions are effective and work much better in smaller systems than convection methods.
For example, we use conduction cooling in many of our environmentally sealed power modules (ESPM). These power supplies are filled with oil and transfer heat out of the unit through long fins that move the heat into the surrounding air or water. These units can withstand tough environments, whether they are dirty, hot, or even submerged.
3. Radiation Cooling
Another cooling method that's less common is radiation cooling. All objects hotter than absolute zero radiate heat, and radiation cooling methods improve the speed of this movement. They aren't viable for most applications but can provide some additional cooling.
How to Measure Thermal Temperature and Airflow
Using the right approach to thermal management relies on accurate measurements. Some of the instruments used to measure temperature and airflow include:
- Thermocouples: A thermocouple measures the difference in temperature between two wires. It offers the best accuracy but is time-consuming and requires a separate data acquisition unit.
- Surface probes: Surface probes have a flat surface for measurements. They're easy to use and offer good accuracy. They can only take one measurement at a time and generally don't offer data logging options.
- Thermography: Thermography systems use imaging to detect and visualize heat distribution. They're quick and allow you to see everything at once but can make it difficult to isolate small parts. Highly accurate systems can be expensive.
- Anemometers: Hot wire anemometers allow you to measure airflow within electronic equipment. They assess how much heat a wire loses while in an air stream. They're affordable and good for measuring localized airflow on power supplies, heat sinks, and other components.
Learn More From the Experts at Astrodyne TDI
Understanding thermal management can be complex, with many different variables and considerations. Astrodyne TDI has extensive expertise in thermal management and can help you find or create the ideal power supply for your projects.
To learn more about thermal management, register to watch our webinar and hear from the pros!