Analog voltage input on STM32Fxxx driver

[jas-mic]

Hello,

I would like to test the plasma module with a board based on the STM32F1xx (STM32F103RC).
This is a "High-density" chip that has I/Os for reading and writing analog signals.
Datasheet of the STM32F103(RC)
The pins are labeled "ADC123_IN" in the documentation on page 13, and there are 3 12-bit ADCs.

Is it possible to use these inputs to read the reference voltage of the plasma arc (coming from a voltage divider, plasma plugin mode 1) ?
I don't think this is currently implemented in this driver, either through the ADC pins or via I2C ?

Thank you for your work
Best regards

[terjeio]

It is possible but code has to be added to support both analog input and step injection. It can be ported over from the STM32F4xx driver, currently I have no plans to do so myself.

[jas-mic]

Thank you @terjeio. I will rather focus on an STM32F4xx since I don't have the skills to port the code.
I did notice the file that implements the analog input, either through GPIO or via I2C (using the MCP3221).

I understand that pin assignment is done in the x_map.h file, but I can’t find any functional examples to help me define the analog input. The web builder doesn’t seem to support this on this driver at the moment.

Please, could you let me know what steps to follow to enable the plasma plugin with a GPIO-based analog reading ?
Have a great day.

[terjeio]

The board map has to include definition for at least one analog input, like this (it is in the only board map that has one defined for now). Analog ports/pins are defined here, which are available depends on the MCU variant.

Are you going to use a galvanically insulated front-end for the input?

[jas-mic]

Thank you, @terjeio, for the information and the links to the files. I’m going to order a board with an STM32F4xx to conduct some tests, most likely a 401 or a 411.

Yes, galvanic isolation will definitely be necessary to prevent electromagnetic disturbances from affecting the STM32. The method for igniting the plasma arc is also very important, even if isolation is implemented.

I hope to retrieve the reference voltage from the plasma cutter without directly measuring the arc voltage. I’ll need to open it up and check if there is an internal voltage divider.
I also need to verify if the STM32’s ADC is linear over 0-3.3V or if it would be better to use the MCP3221 for 0-5V.

I’d like the simplest solution with the fewest additional components.
I’ll keep you updated
Thanks