I'm coming over from PIC32s, and I'm trying to wrap my mind around the differences, and what I can and can't do. I've been successfully tinkering with a Nucleo-L052R8 board, and now I'm looking at making my own stand-alone prototype PCBs. I initially want to use the STM32F030 (20 pin TSSOP package).

I've read a number of posts, web pages, and the Nucleo user manual that say to the effect of "it's easy to program external devices from the SWD header on the Nucleo board", but I haven't seen any examples of this being done. With PICs, In Circuit, Serial Programming is as simple as directly hooking the 5 lines (power, ground, clock, data, reset) directly from the chip to the programmer. Is it that simple with programming from the Nucleo to an external STM32?

Also, the mbed examples are all for the specific STM32s on the Nucleo boards. While using mbed, are you limited to those specific microcontrollers, or can you write code for other variants?

Joe

Hi Joe Take a look at the "platforms" section. You'll see that MBED is across Cortex platforms - not just ST. I have successfully ported it back to ARM7 as well (based on advice from this forum).

You need some method to download code. You don't have to have the USB interface to which you upload the .bin file - you can use JTAG via an SWD or JTAG (5 wire) interface. Most of the implementations have some sort of bootloader that register as a filesystem interface. You drag and drop your compiled code to the filesystem and it uploads to the Cortex processor via a secondary processor on-board that uses SWD to upload.

I hope that helps - post back if you need more explanation. Bill

I don't know if the goofy mass storage download capability will work with other targets, but you should be able to use the Nucleo as an SWD interface for arbitrary STM32 chips, either using ST's software on windows, or https://github.com/texane/stlink on Linux or (possibly) OSX.