Remora - ethernet NVEM / EC300 / EC500 cnc board

Has anybody got these working on a lathe with index and threading working correctly?

i think, encoder with index-pin will not work

I had a test working with the spi variant of remora.
I used the 3d printer main board that has support for hardware encoder input.
I coupled it with an axis to test rigid taping on my mill and it worked.

fintech wrote:

Has anybody got these working on a lathe with index and threading working correctly?

Hi Fintech, what's your proposed setup? Encoder + Index input?

Current firmware does not have an encoder module because of the static config. I'm still working on getting the C++ version of the code working on the RT1052, which would allow for dynamic configuration. Like for the RP2040 which I've just got working.

In the meantime, it would be possible to create a special static config firmware version with software encoder on a couple of pins.

I was thinking something simple like one HAL sensor like I have it setup in mach3. One pulse per revolution. I know this not a good solutions in case the RPM drops under load. I've seen some setups with Dual HAL sensors.

Should be possible then with the current firmware.

forum.linuxcnc.org/38-general-linuxcnc-q...the-spindle-feedback

scotta wrote:

I just ordered one of these kits (thanks to everyone who contribute to the GoFundMe).

www.aliexpress.com/item/1005005021450761...eLogUid=uaetAnuca0fD
 

 
I bought this one : www.aliexpress.com/item/1005005137425987.html
I see both has got an Arduino-style socket match CNC-shield board.
The CNC-shield has low quality and I not recommend it, but it may helpfull for tests, and maybe it is a good idea to assign pins as at GRBL.

scotta wrote:

Should be possible then with the current firmware.

forum.linuxcnc.org/38-general-linuxcnc-q...the-spindle-feedback

But how to do it when you have A B but not Z ? It can be done with other linuxcnc encoder components, but the way the pins are broken out in pru_encoder, I dont think it can happen. I think it needs access to A or B count respectively, and not total count

tuxcnc wrote:

I bought this one : www.aliexpress.com/item/1005005137425987.html
I see both has got an Arduino-style socket match CNC-shield board.
The CNC-shield has low quality and I not recommend it, but it may helpfull for tests, and maybe it is a good idea to assign pins as at GRBL.

The cnc shield is well worth the price ($1usd). Im doing a similar project using STM32 nucleo boards, and its not that bad. Its not as nice as newer designs but its perfectly fine. Maybe we need to make a nicer cnc shield to keep with the times.

Those ch32 dev boards look pretty similar to STM32 nucleo boards, they are probably close to being pin compatible-ish. The ethernet included is nice also. It even has the break away debugger like nucleo boards have.

Some good news. I think I've finally solved the Ethernet issues with my C++ attempts. The issue was related to the caching of the Ethernet buffers. When I converted the C code in MCUXpresso, the following additional linker script settings did not flow over to the C++ Linker settings. Definitely a trap and something I hope others can learn from my mistake.

 

I've started a new repo for the development of the C++ version for the RT1052, which all going well will have the all the same dynamic configuration capabilities as the original STM32 version for the NVEM / EC300 / EC500.

github.com/scottalford75/Remora-RT1052-cpp