Theta JA (θ_JA) Test Conditions for Assessing IC Package Thermal Design

In the previous article, we learned about the concept of junction-to-ambient thermal resistance, often known as θ_JA or theta JA, and how it affects the thermal performance of an IC package.

Here, we will discuss the standardized test conditions under which θ_JA should be measured. This discussion will reveal that it might not be a good idea to use this parameter that is measured under specific test conditions for evaluating the performance of an application-specific design.

 

Measuring θ_JA Under Standardized Test Conditions

The previous article covered the idea that the θ_JA value can be affected by factors such as board surface metal coverage, the number of power planes, and the presence of other "hot" devices on the test board. As a result, manufacturers must perform θ_JA measurements under standardized test conditions such as those described in JEDEC specifications. 

 

Image courtesy of Richtek

 

Defining Appropriate Test Conditions

We need to keep in mind that a datasheet θ_JA value reports the thermal performance of a single device on a standard test board. The JEDEC test coupons are relatively large, at least 76 mm x 114 mm, and place the test device in the center of the board.

Test boards extend the traces by at least 25 mm from the edge of the package body and use thick copper traces especially on the outer layers. For example, a four-layer board uses copper thickness of 2 oz on the outer layers, and 1 oz on the inner layers. These thick traces and large sizes are chosen to reduce the variation in thermal resistance measurement caused by variations in board fabrication process. 

With exposed-pad packages, the choice of thermal vias can significantly affect the overall thermal performance. JEDEC standards define thermal vias for the test boards to have consistent and scientifically sound thermal characterization data. A typical test board can have an array of 4, 9, or 16 thermal vias that are connected to the nearest ground plane. The board power and ground planes must also be unbroken except for via isolation clearance.

 

θ_JA Only Gives a First-Order Approximation of the Thermal Performance

JEDEC test conditions can be very different from that of our end design. For example, the copper thickness of many application boards is 0.5 oz instead of the 1 oz and 2 oz traces mandated by the JEDEC standards. Besides, the number of layers, the board size and orientation, and the number of adjacent components can change from one application to another.

Power dissipation, the moving air characteristics (velocity, direction, turbulence), and environment variations (for example, JEDEC natural convection tests are done in a 1 ft3 chamber) can also make our final product have a different thermal performance.   

Hence, the θ_JA value reported in the device datasheet might not provide us with a reasonably accurate estimation of the thermal performance of our end design. The figure below shows how board size can change the value of the θ_JA parameter.

 

Image courtesy of Texas Instruments

 

Hence, the datasheet θ_JA can only be used as a first-order approximation of the thermal performance in an application-specific environment.

 

Why Do Datasheets Report θ_JA?

The remaining question is, what is the point of reporting θ_JA if it is so dependent on the test conditions?

θ_JA provides us with a metric that can be used to compare products from different manufacturers (assuming that different companies are reporting θ_JA under the same test conditions). If a device offers a θ_JA of 40°C/W and another one reports θ_JA of 45°C/W, the one with lower thermal resistance will likely run 10% cooler.

θ_JA measurements under standardized test conditions are mainly developed to let us compare package performance from different suppliers without concern that our measurements might be affected by more favorable testing conditions.

 

Next: Junction-to-Case Thermal Resistance (θ_JC)

In the next article, we’ll take a look at another important thermal metric: the junction-to-case thermal resistance θ_JC. We’ll see that θ_JC can be used to evaluate the thermal performance of a design that attaches the package to a heatsink.

To see a complete list of my articles, please visit this page.