Experimental study demonstrates reduced thermal resistance and mechanical stress in chip-on-heatsink SiC modules, indicating superior packaging reliability for power converters.
Power module packaging remains one of the constraints preventing silicon carbide (SiC) devices from realizing high power density and optimal reliability in converters. This article proposes a non-thermal interface material (TIM) air-cooled power module architecture, i.e., chip-on-heatsink, to enhance both thermal performance and structural reliability. The chip-on-heatsink packaging bonds conductive copper traces and an aluminum nitride (AlN) ceramic heatsink together without requiring TIM. This non-TIM packaging design streamlines the manufacturing process by eliminating certain steps, such as attaching the bottom copper layer on substrates and assembling heatsinks, since there are few layer stacks between chips and the heatsink. Two types of 1200 V, 36 A power modules are manufactured and experimentally compared. One utilizes the non-TIM structure integrated with <tex-math notation="LaTeX">50× 38× 24 </tex-math> mm AlN ceramic heatsink, while the other follows the standard conventional packaging equipped with the same size 6063 Al Alloy heatsink. Tested under various air cooling and power loss conditions, the non-TIM power module consistently exhibits approximately a 2% reduction in junction-to-ambient thermal resistance compared to the traditional module, indicating the enhanced thermal performance of the non-TIM packaging. Furthermore, continuous performance testing confirms the suitability of the non-TIM power module packaging for operation at 650 V dc-link with 2 kW, making it a feasible choice for power converter applications. Moreover, an electro-thermal-mechanical finite element analysis (FEA) model and the digital image correlation (DIC) test are employed to evaluate the in-plane deformation. Results reveal that the maximum stress of the MOSFETs for the non-TIM packaging is significantly reduced by up to 40.2% along the defined path compared to the conventional structure, demonstrating the potential for better reliability with the proposed packaging.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2024) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: