Remora - ethernet

Hi satsuma, this branch is still in development and totally untested on the LPC1768 on the SKR1.3.

Also the LinuxCNC Remora Ethernet component has developed faster and the packet data structure is slightly different.

In configuration.h the following line would need to be changed to 68
#define SPI_BUFF_SIZE 64

Again, this branch has only been tested on an SKRv2.

Hi scotta, thank you for your reply.
There is always an error and it is very difficult to find a solution.
The last problem: joint 1 following error.
It starts but gives an error message.
I'm testing an Octopus(F446) board now, and need this because estop don't work.
/remora-eth.hal/

#net user-enable-out <= iocontrol.0.user-enable-out => remora.enable
#net user-request-enable <= iocontrol.0.user-request-enable => remora.reset
#net remora-status <= remora.status => iocontrol.0.emc-enable-in
net estop-loop iocontrol.0.user-enable-out => iocontrol.0.emc-enable-in

That is the intended behaviour and is telling you that the communication to the board is not working.

Did you change the line in the configuration.h file and recompile?

Line 640 in main.cpp also would need to be corrected from else if ((recBytes == 64) && (rxBuffer.header == PRU_WRITE))

to

else if ((recBytes == SPI_BUF_SIZE) && (rxBuffer.header == PRU_WRITE))

There could be other issues as well and I note that the code on github is by far not the latest as it is using a polling method for the W5500.

Thank you for your answer. Recompile does not work, mbed studio say lots of errors there. I used the available bin file. The logic analyzer does not see SPI communication. I have too little knowledge to solve it, so I give up and look for another alternative.

hey satsuna, are you using the octopus firmware in the ehternet repo? I never tried the ethernet ones yet. Ill give it a try tonight and see whats going on. what pins are connected to your w5500 => octopus ?

Hi cakeslob, I just followed the instructions:
remora-docs.readthedocs.io/en/latest/hardware/octopus.html

I think the wire is too long for SPI. There are several rules that must be followed for high speed, because SPI not will work well. If this works, a new board is required to function properly. Maybe a LAN8720 or similar would be better. and would use a very fast SPI line for the motor controls and each motor have an MCU that manage the motor and the encoder. SimpleFoc might be a good option and would avoid the step/dir/en logic.