SPRADB7 September   2023 AM2431 , AM2432 , AM2434 , AM2631 , AM2631-Q1 , AM2632 , AM2632-Q1 , AM2634 , AM2634-Q1 , AM263P4 , AM263P4-Q1 , AM2732 , AM2732-Q1

 

  1.   1
  2.   Abstract
  3.   Trademarks
  4. 1Introduction
    1. 1.1 How to Use This Application Note
    2. 1.2 Glossary
  5. 2Thermal Resistance Overview
    1. 2.1 Junction Vs. Ambient Temperature
    2. 2.2 Package Defined Thermal Resistance Characteristics
    3. 2.3 Board Defined Thermal Resistances
  6. 3Board Design Choices that Affect Thermal Performance
    1. 3.1 Thermal Vias
    2. 3.2 Board Size
    3. 3.3 Air Flow, Heat Sinking, and Enclosures
    4. 3.4 Copper Thickness
    5. 3.5 Relative Position of Heat Emitters
    6. 3.6 Layer Count
    7. 3.7 Breaks in Thermal Pathing
  7. 4Thermal Design Best Practices Review
  8. 5AM263x EVM Thermal Comparison with Data
    1. 5.1 Test Setup and Materials
    2. 5.2 Measurement Logging Software
    3. 5.3 AM263x EVM Comparison
    4. 5.4 Measurement Results
      1. 5.4.1 Lid Temperature Readings
      2. 5.4.2 Power Readings over Temperature
      3. 5.4.3 Calculated Thermal Resistance Values
      4. 5.4.4 Recorded Junction and Ambient Temperatures
      5. 5.4.5 Calculated Junction Temperature at Ambient Temperature Extremes
  9. 6Using the Thermal Model
  10. 7References

Calculated Junction Temperature at Ambient Temperature Extremes

The measurements that were recorded for this procedure, only account for an ambient temperature range of approximately -30℃ to 130℃ due to the limitations of the testing environment as described in Section 5.1. The trendlines of Figure 5-7 can be used to extrapolate out the data to approximately calculate the junction temperature in a given configuration of the board, R5 core utilization, and ambient temperature.

Table 5-6 shows the trendlines for junction temperature for different configurations as they relate to ambient temperature.

Table 5-6 Trendlines for Approximate Temperature Readings
Configuration Junction Temperature Trendline Equation
CC only logging script with no Kapton tape on SoC y = 1.0758(x) + 6.3812
CC only logging script y = 1.0716(x) + 8.4798
CC logging script + Load Core 1 y = 1.0644(x) + 8.5352
CC logging script + Load Cores 1,2,3 y = 1.0669(x) + 8.7881
LP only logging script with no Kapton tape on SoC y = 1.0916(x) + 9.293
LP only logging script y = 1.0951(x) + 10.589
LP logging script + Load Core 1 y = 1.09(x) + 11.55
LP logging script + Load Cores 1,2,3 y = 1.1(x) + 11.682

Junction temperature can be approximate for ambient temperature extremes such as 125℃ by plugging in 125 into any of the Table 5-6 equations. For example, an AM263x LaunchPad at full utilization can be represented by the equation y = 1.1(x) + 11.682. Therefore, an LP-AM263x at full R5 core utilization with an ambient temperature of 125 ℃ has a junction temperature of approximately 149.18 ℃.

AM263x junction temperature approximations for ambient temperature extremes have been calculated using the trendline equations and are shown in Table 5-7.

Table 5-7 Approximate Junction Temperatures at Ambient Temperature Extremes
Measurement Approximate Junction Temperature at -40℃ Ambient Approximate Junction Temperature at 125℃ Ambient
CC only logging script with no Kapton tape on SoC -36.65 ℃ 140.86 ℃
CC only logging script -34.38 ℃ 142.43 ℃
CC logging script + Load Core 1 -34.04 ℃ 141.59 ℃
CC logging script + Load Cores 1,2,3 -33.89 ℃ 142.15 ℃
LP only logging script with no Kapton tape on SoC -34.37 ℃ 145.74 ℃
LP only logging script -33.21 ℃ 147.48 ℃
LP logging script + Load Core 1 -32.05 ℃ 147.80 ℃
LP logging script + Load Cores 1,2,3 -32.32 ℃ 149.18 ℃

As predicted in the EVM Comparison chapter, The AM263x Control Card has a better thermal performance due to the increased number of ground layers. For a complete list of best practices and rules of thumb when designing a PCB for the best thermal performance, see the Section 4.