5 Case 2: Polygon Sizing

The second case analyzed is a reduction in the polygon size for the same current sweep. From the above, the 4oz copper weight profile serves as the starting point for this experiment, as this is the full 100% of plane topology. Each polygon leading into the EVM is approximately 900mm² in terms of PCB area, equating to 1800mm² of two 2-layer, 4oz via-stitched planes. From this base calculation, ratios of 40%, 60%, and 80% of this area were calculated, and removed from the PCB while keeping the designed for shape of the PCB polygon pours. Figure 5-1, Figure 5-3, and Figure 5-5 show the topology post copper removal, while Figure 5-2, Figure 5-4, and Figure 5-6 show device performance for the respective ratio.

Figure 5-1 TMCS1123EVM, 40 Percent of EVM Plane

Figure 5-3 TMCS1123EVM, 60 Percent of EVM Plane

Figure 5-5 TMCS1123EVM, 80 Percent of EVM Plane

Figure 5-2 TMCS1123EVM Thermal Response Curves for 5 Devices, 40 Percent of EVM Plane

Figure 5-4 TMCS1123EVM Thermal Response Curves for 5 Devices, 60 Percent of EVM Plane

Figure 5-6 TMCS1123EVM Thermal Response Curves for 5 Devices, 80 Percent of EVM Plane

The responses were then once again overlaid to examine behavioral performance over the various percentages in Figure 5-7.

Figure 5-7 TMCS1123EVM Thermal Response Curves for all Copper Sizes

From the chart, draw the following observations:

• While there is marginal improvement due to the polygon sizing, current carrying capability remains roughly equal across all of the polygons
• The 60% polygons were capable of slightly higher current capacity, but this was concluded to be an artifact of the board to board distribution. The expectation is that over additional samples, the true mean of performance does not demonstrate this advantage.
• Again, board manufacturing tolerances and capabilities altered the spread of devices from board.

The final observation here is that while the current capacity is roughly similar in this distribution, thermal response can most likely reach equilibrium more quickly for the smaller polygons over the larger polygons, due to the additional copper and therefore additional time for the total thermal mass to reach an equilibrium point. This effect, however, was not examined during data gathering.