AI-Enabled Investigation of the Effect of PCB Thickness on Conversion Losses and Efficiency of a DC–DC Buck Converter Under Heatsink-Assisted Operation

Abstract

This paper presents an Artificial Intelligence–assisted simulation based on python algorithm to investigate the effects of FR-4 PCB thickness on conversion losses and conversion efficiency of a DC–DC buck converter operating with a heatsink. An XL4015-based buck converter delivering 5 V at 5 A from a 6–36 V input was used as case study and modeled over five PCB thicknesses (1.60 mm, 2.00 mm, 2.50 mm, 3.00 mm, 4.00 mm) while maintaining identical electrical components, copper weight, layout geometry, and switching frequency. The results obtained show that conversion loss decreased from 0.6016 to 0.3344 W at 6.00 V as the PCB thickness increased from 1.60 mm to 4.00 mm, and similarly, 0.5900 to 0.3299 W at 10.00 V, 0.5944 to 0.3316 W at 16.00 V, 0.6410 to 0.3498 W at 24.00 V, 0.7069 to 0.3756 at 30.00 V, 0.7998 to 0.4117 at 36.00 V, indicating reduced energy losses due to increase in thermal pathway for heat dissipation as the PCB thickness increases, and the corresponding conversion efficiency increased from 97.65 to 98.68 % at 6.00 V, 97.694 to 98.698 % at 10.00 V, 97.678 to 98.691 % at 16.00 V, 97.5 to 98.62 % at 24.00 V, 97.25 to 98.52 % at 30.00 V, 96.9 to 98.38 % at 36.00 V, demonstrating improved conversion efficiency as the PCB thickness increases. The findings established PCB thickness as a decisive board-level efficiency parameter and provide quantitative guidance for power electronics thermal management in converter design.