XY offset , part ruined

10 hours of printing and then boom it loses register to 1 mm in the +X,+Y direction , making the part unusable, leaving strings on the interior of the -X,+Y, and a lip on the exterior +X,-Y...

Anybody see this in anything they are doing?

Can a melt flaw on the part cause the machine to lose register by bumping the print nozzle and thus the head? This would have to cause the stepper to fail to go to its correct count... this would be a physical thing rather than a count error in the stepper Arduino, or can the Arduino have a count error itself?

 

Are you sure the plastic isn't curling and then blocking the extruder path, causing the belts to slip and become misaligned to where the printer thinks it is?

 

Yes Bob, I think that is what happened. I don't think the belts slip, I think the stepper gets physically bumped a step or two and that then creates a permanent error till you re-zero the X/Y axis. Had I been watching the print more closely, I could have re-zeroed and saved the part after cutting off the offending melt error.

Do you see many of these melt errors ? I seem to see way too many for this to be reliable for any significant print function. I ask why should you have a melt error after hours of near perfect printing?

 

The belt's aren't slipping per say. They're the only thing that can give slack when your stepper motors are pulling the extruder into a wall of plastic.

Happens alot to me when my heated bed wasn't working because plastic tends to curl more when the cooling is uneven.

It could also be the pathing. Your gcode might tell the extruder to move over an overhang section, which is curling upwards near the end of your print, and BAM, something has to give.

Try another slicer. I find that cura has some of the smartest pathing out there.