waynegramlich / kicube32

A helper program for taking STM32Cube output and generating KiCAD processor schematic symbols.

kicube32

kicube32 is a program that takes output from the STM cube and generates an a .csv (Comma Separated Value) for input the the kipart program.

Installation:

For now, please clone this repository to get the code.

    cd SOMEWHERE
    git clone https://github.com/waynegramlich/kicube32.git
make

The binary is bin/kicube32, unless you are using virtual enviroments in which case you should be able to just run the program using kicube32.

In addition, this program currently useds a program called kipart to create the KiCAD schematic symbols. This is installed via:

    easy_install kipart
    pip install kipart

The cube data file ends with a .ioc suffix.

Usage

The first step is to start STM32CubeMX software:

    ./STM32CubeMX

This is some Java code with the usual Java annoyances. It is kind of slow and sluggish running on Linux.

• Select a chip or a development board.

• If you select a development board, a number of the pins will be orange to indicate that the development board is interested in thost pins. Try to avoid them if you can.

• Bind pins to peripherals (e.g. UART's, Timers, etc.)

• It is recommended that GPIO pins be given a User Label [Right Mouse Button] to give them a label. Label names that start with an underscore ('_') will show up in the schematic symbol with an overbar over the label name and an inversion circle on the schematic pin.

• When you are ready to generate the KiCAD symbol do the following:

  • Use the [File=>Save Project] or [File=>Save Project as ...] button to save the STM32CubeMX output into a .ioc file.

  • At the same time use the [v Pinout=>Export pinout without Alt. Functions] (or Control-U) to write out a .csv file that lists all the pins.

• It required that the generated .ioc and .csv files be located in the same directory and have the same file name (e.g. f767zi.ioc and f767zi.csv).

• Now run the kicube32 program:

    kicube32 IOCFILE.ioc STM32CUBE.csv KIPART.csv
  

  where:

  • IOCFILE.ioc is the .ioc file output by STM32CubeMZ

  • STMECUBE32.csv is the .csv file that you manually generated using STM32CubeMX.

  • KIPART.csv is the to use.

  • LIB.lib is the name of the KiCAD schematic symbol library file to store the generated KiCAD schematic symbol into.

  kicube32.py will insert/replace a new schematic symbol into KICADLIBDIR/LIB.lib with a name of BASENAME (converted to upper case.)

• Now restart KiCAD and bring up the schematic capture editor.

  • It will likely complain that it noticed that you changed the .lib file behind its back. It will ask you if you want to "rescue" the symbol that changed. Just click on [Cancel].

• Now you can use the schematic capture editor to insert the symbol into the schematic. The generated symbol has multiple units from A to whatever. In general, the program attempts to asign port A to unit A, port B ot unit B, etc. The last two units are miscellaneous pins (e.g. NC's, BOOT, RESET,) and power pins (VSS, VDD, AVDD, VBAT, etc.)

• Each time you update the symbol via kicad.py, you need to go into the schematic capture editor, and force the schematic capature to [Revert to Libary Defaults] using the 'E' key.

Comments

1. This code is not properly commented internally yet.

2. There is a good chance that kipart will be removed over time.

3. There are plans to support Nucleo-64 board swapping whereby a single PCB can be designed to support multiple Nucleo-64's. The Nucleo-64 boards tend to have different pin bindings and peripheral pin mappings, and this will help alleviate that pain.

4. There are plans to support mulitple board that stack on top of a Nucleo board (call them daughter boards) and make sure that the pins do not conflict.

5. It may be possible to eliminate the .csv file output and directly read all the necessary information from the .ioc file.