Why touchgfx doesn't support stm32f1series mcu's? And will it be supported in the future?

[AksultanK]

Hello! I asked this question at recent webinar, but my internet connection was horrible, and I couldn't understand the answer unfortunately.

I mostly develop on stm32f1 series mcu's and it would be amazing to have this tool. So, why touchgfx doesn't support them? And will it be supported in the future?

I didn't try the touchgfx software myself, because I don't have compatible hardware yet. So maybe I'm wrong and you can make it work with the f1 series. Please can someone cclarify this?

[Romain DIELEMAN]

Hi,

I believe it is not supported because the F1 serie have an ARM cortex -M3 core. This used to be supported by TouchGFX but has been removed since there were no ST dev kit with an F1 and a display kit using Chrom-Art accelerator to be tested on (and I dont think there are still any yet).

It could be something supported in the future if there is a push from marketing, and I don't think it would be in the near future unfortunately.. You are not the first one to ask, so I have made the teams aware of this.

Could you also add this demand in the Idea Zone of this forum ? This will help us keep track of those demands. Thanks

/Romain

[Stefano Bettega]

I was just wondering the same thing, we are planning to add a SPI display like on the STM32G071 Nucleo board, to our existing STM32F103 project. As the Nucleo board has more or less the same resources of our board, it could be a easy task to do so.

Stefano

[Stefano Bettega]

After some digging I found a couple of useful information:

In x-cube-touchgfx.pdf document (installed into TouchGfx directory) it is stated (bold is mine):

Minimum CPU load and footprints

– Taking full benefit of STM32 graphic hardware accelerators

– Optimized library footprints requiring from 20 Kbytes of Flash memory and 10 Kbytes of RAM

– Partial frame buffering minimizing graphic buffer sizes and enabling graphic UI support from lowest-cost STM32

options

– Running with RTOS or on bare metal, supporting the STM32 microcontrollers based on the Arm® Cortex®-M0+, M3, M4, M33 and M7 processors

And if you check out Cortex-M Wikipedia page here, you can see that M3 core has all of the features of M0+ (or better). So my opinion is that you can link against the cortex_m0 library in TouchGfx core, and have a working version. BTW this won't be supported in CubeMX and you'll have to configure the library by hand (but at this moment our project does not use CubeMX, so this is not a real big problem).

@Romain DIELEMAN​ can you have some more information from the involved teams? We are about to take a decision on how to proceed with the design.

Thanks and regards,

Stefano

[Romain DIELEMAN]

Hi Stefano,

I asked around again to get more info and this is not in our plans to support F1 series. The doc you found refers to what I wrote before as well where it used to be supported years ago. Theoretically it should be do-able by yourself since the libraries are close like you wrote but I do not know where to guide you.

/Romain