3 SAMA7D65 SiP Measurements

The board used for the measurements is an evaluation board with the same characteristics and size as the SAMA7D65-Curiosity evaluation kit. Supported by a mainline Linux distribution as well as bare-metal software frameworks and RTOS, the kit allows easy use case implementation and performance measurements.

  • The power consumption is measured using the on-board Microchip Quad DC Power Monitor PAC1934.
  • The junction temperature is measured with the SAMA7D65 internal die temperature sensor.

Two use cases are considered in the following sections:

  • Use Case 1: Linux idle, the O/S prompts for commands.
  • Use Case 2: iPerf3 test on the Gigabit Ethernet interface, generating a high CPU load along with large memory transfers.
Note:
  1. The results presented hereafter strongly depend on the PCB on which the SAMA7D65 SiP device is mounted, and more specifically on the effective thermal resistance (RJ-A) of the chip soldered on the PCB. For the SAMA7D65 SiP board, the junction-to-air thermal resistance was first characterized with a set of measurements (see the method presented in the SAMA7G5 Series Power Consumption and Thermal Considerations (AN4797) application note, section In-Application RJ-A Measurement) and its value is estimated to 19°C/W (±10%).
  2. Use Case 1: Linux Idle and Use Case 2: iPerf Test show the measurements of two parts, called “Typical” and “Maximum,” respectively. “Typical” corresponds to a part having a power consumption identical to the average of the measured batch. “Maximum” is an extrapolation of the maximum production part (μ + 3 x σ).
  3. TJ measurement accuracy is ±5°C.
  4. To ease comparisons between curves, the x-axis always represents the device junction temperature, not the ambient temperature. A vertical line at TJ = 105°C indicates the maximum specified junction temperature for the mounted (industrial grade) SAMA7D65 SiP devices.