Custom Board and application based on nRF Connect SDK

[masterLee]

Hi,

recently I started migrating my company's firmware from nRF5 SDK to nRF Connect SDK. I have been following the devAcademy courses and so far
1. I have created my custom board files
2. Defined the .dts for the custom board and populated it with the required GPIO pins and serial protocols(SPI, I2C) that are used for the peripheral sensors
3. Created a simple application based on the "Hello world" example

Now I am at the point where I want to add my sensor libraries into the new application in order to start re-creating the app main code.
In lesson 7 (nRF Connect intermediate course), an introduction to device driver model is presented and to my understanding, this will take considerable time to make the changes in the way the library for each peripheral is used with the device driver model.

I wanted to ask you if it is ok to use my existing .c/h library files for my project in conjunction with the I2C and SPI driver APIs of nRF Connect SDK, instead of using the device driver model for all my sensors' libraries? Would this issue any errors or conflicts with zephyr in the long run?

The reason I don't want to use the device driver model is because I will need considerable amount of time from creating a new driver model for each sensor based on the current libraries.

What do you think?

Thank you!

[Einar Thorsrud]

Hi,

I do not see any fundamental problems with ignoring the Zephyr driver model. The advantage of that is that the interface is standardized and portable, but it is fine to implement drivers in your application in other ways, and in many cases that makes more sense. (You can refer to this unofficial sample to se a simple way to implement for instance SPI communication.)

[masterLee]

Good! I want to make some progress the classic way and then perhaps I can test the device driver model when I have set up the main structure.

Thank you for your time and help!