AD8403 Ready

[Ale11Re]

changed SPI_MODE1 to SPI_MODE0
changed Reset, Shutdown logic
changed demo example

[RobTillaart]

@Ale11Re
Thanks for the issues and the PR, I will look at them today.

[RobTillaart]

verified the documentation and merged your code.
Thanks again, I will update the version to 0.3.0 (going to add RP2040 in build support too)

[RobTillaart]

@Ale11Re
Could you verify if the software SPI does works as intended (I assume you have some hardware nearby)
Thanks for your time.

[Ale11Re]

@RobTillaart Yes, I confirm it's working with software SPI too.
I have a question now: does it work with AD520X after my changes?

[RobTillaart]

Thanks for verifying.
I have to check if I have a spare to make a test setup.
Otherwise I need to order one.

Just to know, what board do you use?

[RobTillaart]

No chip available, ordered - https://nl.rs-online.com/web/p/power-motor-robotics-development-tools/2475645
(not cheap but looks easy to test)

[Ale11Re]

I'm using an arduino mega2560. I could also verify ESP32 VSPI but not HSPI as I only have 1 model which has HSPI already in use.
I only have one AD8403 lying around.
I used an oscilloscope to verify sawtooth and sinusoidal wave and as you mentioned, software SPI is about 15 times slower than hardware SPI.
Just to clarify: this is my first contribution in github, I believe I understood how this platform works so if I made a mistake, please let me know.

[RobTillaart]

Your contribution is super 👍 really appreciated.
I have written a lot of libraries in past decade and are not able to test all possible configurations.
Some are even only tested by others

I am interested in the ESP32 VSPI and the software SPI. The latter is very fast on ESP32, in some other tests it was roughly on par with hardware SPI. The SW SPI pulses are slightly slower thus longer, but SW SPI has not the transaction overhead. In fact this results in nicer square waves with the SW SPI. It would be nice if confirmed here too.

[Ale11Re]

So, as I'm writing, I'm testing HSPI and VSPI on ESP32 and they work as intended.
Software SPI works too but it's a bit slower:
1 tooth of the Sawtooth wave takes about 4ms on HSPI and VSPI while on Software SPI 6.5 ms.
Just for reference, it's about 5 ms long on the hardware SPI of AtMega2560 and 60ms on software SPI.

[RobTillaart]

Thanks for these numbers, they look good imho.