Issue link: https://iconnect007.uberflip.com/i/1511130
NOVEMBER 2023 I DESIGN007 MAGAZINE 27 But that's not all. Figure 4 shows a simulation where we have placed two thermal vias under the right-hand pad. e thermal vias are 20 mils in diameter and plated to almost 3 ounces. e thermal vias are relatively large so that we can see the results in the figure. Even so, we have to zoom in optically and narrow the temperature range in order to visualize those results. As seen in Figure 4, the pad is hotter in the center than it is around the edges. is is because the edges of a pad or trace cool more efficiently than does the midpoint. Heat can conduct away from the midpoint primarily only in the vertical (Z-axis) direction. But heat around the edges can conduct away both verti- cally and horizontally. All normal pads, and all traces are cooler at the edges than they are at their midpoint. But more importantly, the thermal vias only have a small impact, and they have that only at a point 7 . 1. Each thermal via only changes the temperature by less than two degrees. 2. And they do so at only a very small area around the via. at is why almost every article that dis- cusses thermal vias recommends using a lot of them. But even if you covered the entire pad with thermal vias, the temperature still would not change by much; the gain has already been achieved with the underlying plane. e only thing thermal vias do is use up a lot of routing channels. Summary By learning more about how PCB design decisions impact thermal temperatures around the board, we discover there are several areas where we can make significant impacts in design efficiencies. DESIGN007 References 1. IPC-2152, Standard for Determining Current Carrying Capacity in Printed Circuit Board Design, IPC, August 2009, page 26. 2. PCB Design Guide to Via and Trace Currents and Temperatures, by David Brooks and Johannes Adam, Artech House, 2021. 3. We talk about temperature sensitivities in Chap- ter 7 of our book referenced in No. 2. 4. We used a simulation program called ther- mal risk management (TRM), which was originally conceived and designed to analyze temperatures across a circuit board, taking into consideration the complete trace layout with optional Joule heating, as well as various components and their own contri- butions to heat generation. 5. See Section 8.4 of our book referenced in No. 2. 6. The maximum "pixel" resolution and density in a simulation is determined by the smallest dimension in the X-Y plane. Via simulations require about one to two orders of magnitude greater resolution than "standard" trace simulations, placing a significantly greater load on the computer CPU and memory. For this reason, modeled board areas for via simulations are typically quite a bit smaller than would be the case for regular trace simulations. This results in a slight upwards bias in model temperatures from what might be otherwise expected. Relative tem- peratures within the model, however, exhibit much smaller upward biases. 7. Temperature is a "point" concept. That is, it varies from point-to-point around a board or trace. So, too, are resistance, resistivity, thickness, and sometimes even thermal conductivity coefficients. We discuss this in numerous places in our book, and especially in Chapter 13, "Do Traces Heat Uniformly?" Douglas Brooks, PhD, is a veteran signal integrity instructor and the founder of UltraCAD Design in Issaquah, Washington. Dr. Johannes Adam is a thermodynamics physicist and founder of ADAM Research.