Is there any plan to upgrade or move to a more powerful 32-bit ARM based controller platform? The current 8-bit RAMPS board is very capable, but bogs down when trying to parse and run through gcode that defines a lot of curves and circular based movement quickly.
I have seen 32-bit based boards and looked into Marlin compiled for ARM, but I am unsure of compatibility with the existing printer platforms.
This is very interesting! The Archim looks to be the same form factor as the RAMBo 1.3L and 1.4. Does that mean that it can be a direct replacement to the board in the TAZ6? If so, are there any plans to support a controller upgrade with Marlin 2.0 firmware for the Taz6?? I’m sure a great many people would an upgrade path!
I haven’t yet, other priorities have been in the way, but several people in the facebook group have and report great success. All the new features, much quieter steppers. LCD implementations can be a little tricky.
@piercet, you probably shouldn’t be controlling an LCD with the same micro as the one controlling a stepper. Precise motor controlling should be a sub-system that reports to the primary controller that handles display and outside COM.
Any updates or progress here? My Yaxis shat the bed, looking to upgrade my Taz6 to 32bit thinking of going with an SKR board, but im out of my depth with working out how to set it up.
It is basically a plug and play conversion if you have all the right parts. IT-Works 3D put a kit together for $350 that includes everything you need. I think others have has success integrating other 32-bit boards, such as the one you mentioned and the Duet, but that will require a lot more work.
For FW, I am currently using Klipper, however drunken-octopus-marlin is also a tried and tested option with the Archim2. Links and more info at the end of the writeup.
yes I have, actually. The motion of the machine is much smoother which results in cleaner curves (which resulted in more dimensional accuracy) and more details. I’m pretty happy with the setup!
That’s super interesting. I secretly was hoping it was a waste of time, but I know even with the compressed messaging in Klipper that the 2560 is running up against hardware IO constraints. Now I have a rational reason to upgrade boards, damnit.