Missing Features in LCNC for FFF/FDM Printers

tommylight wrote:

I could think of some more drawbacks of Duet and Klipper and Marlin, but right now i am fighting the extrusion rate and that is something you can not change while printing, or at all as it requires stopping>slicing>start over>throw the partially printed part
Pretty sure that could be done in LinuxCNC during printing.

Why not use an external offset? I've been toying with the idea for pressure advance but we could monitor the A axis command and add / subtract a %. 

scotta wrote:

Pretty sure that could be done in LinuxCNC during printing.
Why not use an external offset? I've been toying with the idea for pressure advance but we could monitor the A axis command and add / subtract a %. 

That is exactly what i was thinking, on both accounts, but pressure advance would also require a delay between extruder and XY motion, not much important for direct drive but required for bowden as soon as you pas about 20 to 30mm/s depending on the length on the tube.
So sending the commanded position for X and Y through an adjustable delay should be easy enough.
Or am i overthinking this ?
P.S.
@Scott
I was just checking what else i need to print to finish your printer, started waaaaay back and bumped into some rail size issues ! Building a small one at 160x160x??? .

Hello,

I actually manage the extrusion feedrate through a pressure control.
I linked the extrusion feedrate with the xyz nozzle speed, and I removed the A coordinate from the gcode.
Actually we are considering only the prints in spiralize (aka no retraction ) with this method, if not the post processor for the gcode should be more complicated.
The worst barrier in my opinion about lcnc and 3d printing is the trajectory planner that doesn't have the look ahead option for more than 3 axis.
The pressure control act as a workaround for it, reducing the number of axis from 4 to 3, but absolutely we need to develop a better tp for this specific aim.
I'm waiting with anxiety the Grotius job. I saw that he was working on a standalone version of tp that maybe could be the best solution for this.

I think Andy mentioned somewhere using A axis as Z and using the Z axis for extrusion, should work perfectly but i do not have time to test it.

This swap between A and Z axis should work if you are not printing with the spiralized method ( as we do because of very large nozzles) where the Z value is increasing constantly or in some fancy prints where you slice the object along curved surfaces instead of planar surfaces.
In general we need to use a different "tp" to solve this problem at the root, also because interruption or not smoothie extrusion behavior acts as very bad prints quality and strongly increased printing time.

The extrusion feed override is really useful, especially with different materials suppliers and differents material behaviours.

Roiki wrote:

There pretty much no point in using lcnc when duet and klipper exist.

I have a Voron 2.4, but I also have a large CNC router. I would love to be able to throw an extruder on this machine and make some large 3d printed parts. Yes, it would be slow, but I know it's reliable, and it's paid off. And the ability to print on a 5-axis machine with LinuxCNC opens up some incredible print quality and topology optimization opportunities.