Optimize thermal design using actual components’ stress

Thermal analysis is an important part of any electronic system design.
Electronic devices that consume high power require heat removal mechanisms such as PCB substrate, heat sinks or fans.
Thermal analysis is usually performed by mechanical engineers or 3rd party companies that specialize in simulating heat and air flow. This analysis requires information about the components placing on the PCB, actual power dissipation and PCB materials and geometry.
Usually, the thermal analyst conducts the calculation using the absolute maximum power rating from the components’ datasheets. The actual power is often significantly lower than the maximum rating. This leads to overdesign of the heat removal mechanisms.

BQR offers software and professional service for component stress analysis by circuit stress simulation.

BQR’s software provides several methods for calculating and documenting the components’ actual stress (power, current and voltage) values:

  1. Semi-automatic: stresses can be easily assigned to components using the fiXtress E-CAD plug-in on the schematic
  2. Stress simulation: Components stress is calculated by a unique automated circuit stress simulator (fiXtress)

Benefits:

  1. During stress analysis, design errors and rating issues are detected before layout
  2. Exact stresses allow for optimal thermal design, saving space and cost
  3. MTBF calculation using exact stresses provides better (higher) MTBF values

Example 1:

Tj of 137.1oC was calculated for IC U2a based on the absolute maximum power rating (6.12W). Based on this result the mechanical engineer was planning to add a fan to the design. Furthermore, the IC failure rate was calculated to be 6.129 failures per million hours.
By using the actual power dissipation result from the fiXtress simulation for the calculation, (see IC U2 in fig. 1), Tj was found to be 95oC, and the natural cooling was sufficient. Additionally, the failure rate decreased to 0.804 failures per million hours.
A difference of 42.1oC was found in Tj, and the failure rate (FR) decreased by a factor of 7.6.
To summarize: The use of actual power dissipation reduces the unnecessary use of costly cooling elements, and results in a much higher MTBF.

Figure 1: Tj and failure rate for different power of an IC (screenshot from BQR fiXtress)

Example 2:

MTBF of an actual board was calculated for both actual stress (Parts Stress Method) and 50% stress (parts Count Method).
MTBF calculated with actual stress was found to be higher by a factor of about 1.5.

Screenshot of MTBF software
Figure 2: Comparison of board MTBF calculated using “50% stress” vs. “Actual Stress”