This article details the development of a bill of materials (BOM) tool, Circuit AI, which integrates with OpenAI’s GPT-4 model to streamline component selection and optimization in electronic design. Circuit AI leverages GPT-4’s Function Calling, Code Interpreter, and fine-tuning capabilities to simplify BOM tasks, such as switching loss optimization and capacitor RMS estimation. By automating tedious searches for components, availability, and datasheets, engineers can focus on critical design work and make informed decisions on cost-effective, reliable parts (Figure 1). Tested with a Texas Instruments BOM, Circuit AI demonstrated its efficiency in optimizing reference designs, identifying alternatives, and enhancing workflows from selection to production. A key application of Circuit AI is in power electronics, where it leverages GPT-4’s finetuning capabilities to estimate capacitor RMS currents (Icap-rms), a crucial factor in preventing overheating and potential failure.
For more on the article, click here.
