Switching between Nucleo and custom board

[alanb]

I used a Nucleo F439ZI to develop initial software (with a custom daughterboard) and now have redesigned the whole lot onto a single board with the STM32 MCU.  I'd like to remain able to support both boards in my Cube project as the nucleo one is much easier to modify.

I used the pinout -> compatible MCU selector in Cube to find an MCU which digikey stocked and which was the cheapest pinout compatible which met my needs which found the F429ZET6.  I'd have imagined that thanks to the HAL I would simply need to switch "Build configuration" or similar to go between them but this doesn't seem to work it sems I need to completely retarget, code generate the project. Am I missing something obvious? This seems to be the design journey the software encourages me to take.

Did I make a mistake in not using the identical MCU (STM32F439ZIT6) to the one on the Nucleo Board?

[TDK · Accepted solution]

You're not missing anything. Cube isn't set up to switch MCUs in a developed project (with user code) seamlessly. Since you've already started with one chip, you have two options:

• Use the new chip, but don't tell CubeIDE about it and develop using the project made for the previous board on the new hardware. Since your two chips are near-identical, this will work just fine as long as you stick to the subset of features present in both.
• Duplicate the project or create a new project for the new chip. Copy over user code manually.

If it's a project that matters and will exist or need time/development in the future, I'd choose the second option. Especially if the user code is minimal or can be copied in 15 minutes or whatever.