Remora - Rpi Software Stepping Using External Microcontroller via SPI

Ya, I used some cheap chinese jumpers with dupont cloned connector ends and maybe 7in long wires. Crappy ends and my rPi would get moved lots as I messed with things. Scotta did fine with short jumper wires too while his rPi and controller board were mounted to his 3DP so quite stable.

After making the ribbon cable adapters I could go to 70cm cable and even tried 130cm cable which worked for a while but I did get one SPI error so that was too long and maybe 100cm would be reliable. Having the connections all in one 2x5 socket and 2x5 header pin keeps all connections stable and in place.

tjtr33 wrote:

Doogie,
thanks,
I read I can assign Px.x to any free pin,
and should create the pin in config.txt.

Under what conditions is Reset active?

thanks tomp

Hi tomp, 

The Reset pin is available on the LinuxCNC side and I use a pyvcp button that allows me to manually reset the board from within LinuxCNC. This is handy when the SPI comms get interrupted and the buffer is not clocked out completely. I really should build in a SPI timeout functionality that switches the board into master mode and clocks out the tx buffer and switches back into slave mode. But, in normal operation this isn't a huge issue.

BR
Scott

For those wanting a simpler SPI connection situation, the prototype boards arrived from China today and I made up some adapters and tested the MKS and rPi ends while also ohming out the SKR adapter and all is A-ok.You can order your boards from dirtypcb.com or ask and I'll get you the KiCAD and GERBER files. 

Remora MKS and SKR ribbon cable Adapters PCBs

This uses standard 2x5(10 pin) header sockets which fit in the EXP connectors on the SKR board or on the MKS board the outside row of the 2x5(10p) socket slips onto J7 header pins. Any straight or 90deg 2x5(10pin) headers will fit in the other end and interface with the GLCD ribbon cables normally used on the EXP1 and EXP2 connectors.

OK, I found one goof and that was the rPi 40pin connector layout relative to the board. I made the mistake of doing the layout based on the very nice ARGON One 40pin layout and it turns out it's  rotated 180 degrees from the rPi itself.  So on the ARGON One the board/adapter sticks out the rear but on the rPi the board/adapter extends inward over the CPU instead of out the side. It works fine and if you use straight 2x5 header pins the ribbon cable and connector will extend upward instead of across the rPi.

I'll repeat my testing with the very long 700mm GLCD ribbon cable now that the grounding and signal traces are solid on both ends.

It is not easy to register on this website, for me.
In the past period of time I tried the initial use of REMORA (although there are still some configuring problems), I must say that this is a very creative project.
Thank you for your work, especially, thank you Scotta for answering my questions in the email.

So i have done some tinkering and made a new spi connector.
I had some thin coax cable from when i used to make FPV antennas and tried them out.

I have the system grounded on both ends of the coax and 2 dedicated ground wires.
The wires were just the right size to still fit into a dupont connector with the shielding pulled back.
So far no issues at all.

Kinda wondering what the best way to stress test them to make sure no phantom problems arise.

So I've been trying to create a HAL file to use the SKR1.4 turbo for a CNC mill. I have the basic motion working... I've spent the last few weeks doing a udemy class on HAL (hands-on-stm32-basic-peripherals-with-hal).. And really have not gotten very far... Basically, I would like to use the bed heat output as spindle control, The endstops input as endstops, the fan outputs as coolant and the probe input as probe. Since this board is readily available and the layout is the same it would be great to have a file that maps all of this and with 2209 support all in one premade file, for those who want mill control? Basically, have all the I/O pre-mapped and working just like if it was running in a printer but just plugging in what you need and go.

This is what needs to be done if anyone could lend a hand
heatbed output mapped to M3 with or without PWM

endstops inputs mapped and working

Some of the I/O (like fan output for M8) mapped

Probe input working as Probe input

2209 auto endstop function working

If we can make this plug and play with your connectors, an RPI IMG, and pre-made mapped HAL there is no reason this would not become the go-to for the little 3018 like mills. thank in advance for any help or direction.