I have this shifting on the X/Y axis every once in a while, it’s completely random, sometimes it goes away if I split things into layers, sometimes it goes away when I change the direction from 90 to 0 or viceversa.
It’s almost always at 50% of a given layer’s content, the motor should change direction (bidirectional engraving, constant power, fill) but it just stops and burrs like it’s reached the end, then continues as usual, but obviously shifted, and it’s always the whole length of the layer’s contents.
It usually happens on the X axis, sometimes on the Y.
Also, it almost always happens on these business cards I engrave, on the larger layers, never on the small ones. And the more stuff there’s in a layer, the higher the chance. I’ve marked plenty of cardboard boxes which take up almost the whole table, and haven’t had a problem with any of em. The issue is only when the layer is on “fill”, never on “line”.
Restarting everything doesn’t help, I have the latest LightBurn version (updated it today, still had the issue). Machine is CNC 3018 PRO grbl, laser is 80W, but it did the same thing with my 15W head too.
I see you have a screw drive machine. I also am pretty sure it is a mechanical problem. Check the set screws (both X&Y) in the coupling connecting the stepper motor to the screw shaft.
The thing that I see is that the offset is consistently the width of the graphic, every time. My first thought was drag chain catching on the frame (if there is one)…but the consistency is too coincidental for me to not want to keep looking.
That makes me think there is a communication error rather than a mechanical error. I am wrong a lot, at least twice a day
Are you using a USB cable or Ethernet? Or are you watching the files on a thumb drive? I just woke up, and may not completely understand the setup he is using. A few more details may help us find your gremlin.
That was my first thought too because I think coupling issues are rare in screw machines. But I picked mechanical because he did not mention it randomly stopping. Yes, definitely need more info.
The complete offset is what I find odd. It looks like the Gcode skips a “stop and turn around” command on the axis?? Then the rest of the object completes.
Not a very technical explanation .
+1
We would do better detective work if we could get our fingers on a sample file where that error manifests.
Edit:
The term I was looking for is that the machine is loosing steps. Most of the time I have seen step loss, it is staggered rather than completely consistent.
A missing “stop and turn around” is what describes is most accurately. Or more like stopping in one spot instead of turning around, but just for one line.
I am using a USB 2.0 B cable, I have ordered a new one to try and see if the issue comes from there somehow.
All set screws are tight and loctited in. No drag chains.
The interesting thing is that when it stops instead of turning around, the motor makes the “krrrr” sound as if the axis has reached it’s end. Sorry for the dumb explanation.
Attaching a file which had this issue 6 of 10 times yesterday:
I am on mobile this morning, but have been thinking about how to troubleshoot.
I suggest you test on scrap cardboard or similar. Save your good metal for finished product. Adjust your power setting accordingly. Have a bit of water handy in case you need to dampen out some flames!
Try running these tests:
Change the scan angle from 00 “flat on X axis” to 45 degrees and the 90 degrees.
Did the error happen again? Show a photo please.
Scale your project down 50% or so. Show a photo please
Slow your speed by 50% (mind your power level again). Show your results
Move your piece to another place in the work area, as far to the opposite side as you can, then as far back as you can. Move From 0,0 to… max,0, then move max,max then to 0,max
I suspect your problem will still be present, ruling out a software error.
What driver and software version are you using?
Is you system well grounded for electric?
We love helping people find success to their laser mysteries.
Re-reading (oops), I see you already covered most of what I suggested.
My inclination is that your gantry head is moving fat too fast or too high of acceleration settings. The inertia is over loading your stepper motor.
That’s a perfect description of a stepping motor losing sync with the driver pulses.
Verify the small trimpot at the top center of each driver PCB is set to provide the proper current for the motors. You’ll need a voltmeter, because the data sheets give the current as a function of the voltage on a PCB test point.
If the motors continue to misbehave with the proper current, then you must reduce the maximum acceleration for the axes to reduce the torque required to get the axes up to speed. You can set those directly in the console with the GRBL $120 and $121 parameters or through the LightBurn Machine Settings.
Reduce them by a factor of ten, which should make the axes respond very slowly and may increase the overscan region beyond the machine boundaries. Increase the acceleration by a factor of two until the motor fails again, then find a compromise between those two values.
Thank you! Will try reducing the acceleration and do some tests. I don’t have a multimeter yet, just ordered one.
The machine used to have all the wrong settings by default, and the acceleration was set to very slow. I’ve used it plenty of time with the wrong settings, and it had the absolute same issue then.
Which speeds did you use? Be aware that such mechanics are only made for very slow movements, around 1000 mm/min. If you command it to be faster, you might get into stepping issues.
.