I engraved and cut a thing and it mostly worked…except for one place where it did a pointless move to X678 (x max) with the laser still turned on. This was running from the program over USB (not using a gcode file) but I exported a gcode file and cannot find any moves that far to the X+ direction.
It happened at the top of a ‘slot’ shape. The design has dozens of the exact same profile that didn’t have issues.
Screenshot, picture of issue, and gcode file below:
Is that photo supposed to show the unexpected move that you’re talking about? If so, it’s not clear to me where that is.
Try running the gcode that you generated again from LightBurn with the laser off. Does it repeat the motion? If so, there may be a quirk in your firmware.
I don’t see anything in the g-code like you’re describing but it’s also not clear to me where X678 is on your laser.
Is there a reason why you chose Marlin for a custom CO2 build?
There is a slot shape to the left of “4MM”. After completing this slot, machine moved all the way to the right of the image (where the paint is burned off of the black frame) and then back to start the 4. The white line engraved into the frame ends at X678.
I should have tried that again but the job is gone now.
I have two close friends with different CO2 lasers. All three of us had different Ruida controllers. All three controllers got failed X axis because the transistor they use burned out. Wasn’t about to give them more money if they refuse to use quality parts. GRBL only supports 6 motors so that wasn’t an option.
Okay, so it’s the faint white horizontal lines originating and ending near 4MM mark that is the issue? Interesting that the intensity also seems to be lower.
Can you check Preview simulation and see if there is any traversal move that would account for this? Or is this impossible when you say the job is gone now?
Do you ever have other occurrences where traversal moves are being burned? How long have you been using this setup and is it otherwise reliable?
That seems unusual, were these all DIY systems with the same basic design?
I’m not aware of GRBL having a motor limit. Different variants have different axes limits. How many motors and axes do you have?
I didn’t think to the save the job; just started cutting it up to try to save the wood. I did a ‘FRAME’ before starting and this right move was well outside of the frame that lightburn traced. I tried recreating the job from the same source and coordinates but don’t see any far-right moves in the preview.
This is a relatively new setup; the first time I’ve tried a job this big actually.
All 3 of our machines were factory made. Three different companies but very similar designs other than size and power. Mine was heavily modified; the others were stock.
I have 4 axis and 7 motors. X, Y, A, and 4 motors (with 4 endstops) for Z. There were some other big features it was missing too but that was the killer.
It’s not true ABL; there is no probe. But it does level the bed and reset the level every time I home. Originally had a long belt that went to all 4 corners, it would skip no matter how tight it was. I had to manually turn each corner to get it level again. That got annoying fast so I ran the machine as if the Z was welded in place for like a year before I designed the 4 motor leveling setup.
I have a setup like this. I don’t think I’ve ever had it skip where it wasn’t associated with crashing the bed.
I’m reconsidering the issue you and your friends had with failed X axis… I’m not certain why the controller side would have failed but I know that some machines come from the factory with the driver current settings set too high. The controller should be isolated from that, however, so not sure why that would cause it to fail.
That might be. I am looking into ways to get gcode files onto the board without needing to pull the microSD every time. Might have found one. It’s extra steps but it’s worth a try.
