Nowadays, creating hundreds of 3D CAD models often leads to secondary views for mock-ups, analysis, and manufacturing. However, up to 48% of CAD models may fail after changes, sometimes rendering them unusable. The quality of a model can be defined by the absence of errors. This work aims to detect and quantify common errors during the CAD 3D design process based on six dimensions of CAD quality. This study analyzes the quality of 3D CAD models, focusing on fundamental dimensions of quality. Examples of 3D models constructed with SolidWorks were examined using a rubric designed for these six dimensions. The main goal was to identify common errors made by engineers, despite their training in 3D quality models. While this study has limitations, it identified frequent errors, providing a clear understanding of areas needing improvement to ensure higher-quality models. We propose extrapolating these results to models of similar difficulty constructed using the Feature-based Design method. This will allow us to identify and detail specific errors in 3D CAD models more precisely compared to previous works. This work is a first step towards future research proposing methodologies to improve compliance with established quality criteria. The ultimate goal is to prevent errors in 3D CAD models, enhancing their quality and reliability in industrial and design applications. We conclude with a summary of lessons learned and directions for future work.