Cut end points issues. PLEASE READ

Ok folks. I know this topic has been covered hundreds of times but here goes.

When cutting circles, squares etc, the end points have a step.
IE, they don’t meet up using LB.

I have a 6090 Opus Laser 100w Ruida controller. The X/Y axis are at 90 degrees, belt tensions across the 3 belts “appear” to be of equal tension.
So annoying is this issue, i have decided to buy another laser with servo steppers at a huge cost!

So, I drove to Wales and Gloucester yesterday (from Southampton 10/07/23) to visit with Thunder laser and Trotec.

Our host at Thunder was fantastic, Wayne, the owner couldn’t have of been more helpful and accommodating.
Armed with a LB file on a memory stick, I aske Wayne if he could cut the flower shape on his Thunder laser 24 and 35 machines. I was hoping that this would prove, once and for all that our machine was the problem.
But…er…no! :rage:
The same issue was replicated on both the Thunder lasers, even tried on Wayne’s computers turning on and off “hide backlash.” Results were the same. Misaligned end points…!

So, maybe, I don’t need to plough 15k into a new machine I thought!
Wayne said that he uses RD Works and would try cutting the LB file though RD Works.
RD Works cut the circles perfectly. Endpoints married up with out any step at the end of the cut.
Amazed as much as I was, Wayne then tried again on LB, and guess what? Misaligned end points. Both of us were as perplexed as each other. Clearly, there is an inherent problem with LB software somewhere? I don’t buy it, that the “hide Backlash” option needs turning on, this is masking a LB issue, is it not?
So, we then went to Trotec, met with Jamie, again thumbs up to him, great guy. Tried out the R400 machine, (beautiful in every aspect, but to big for us), we tried our shape out on the machine, end points perfect, via the Ruby software that Trotec use.
I since contacted Aeon in Croydon to arrange a site visit to look at the Mira 9 laser, the sales chap has been amazing, asked me to send the LB file to them so they can look at the LB parameters and hopefully come out with an answer as to why the end points are misaligned using LB and not when using RD Works.
I have tried now, using RD Works to cut the shapes out on my machine, and the end points meet without any step on the circle cut out or the final shape cut out.
Tried LB and same results, end points misaligned…!
I am using LB Ver 1.3.01. I am loathed to spend any funding to “see” if newer versions eliminate the problem and I would ask for LB team to contact me with a proposal for ironing this problem out, not just for me, but for everyone who has a similar issue, because it is a serious issue, yet to be properly addressed by the LB team.
Sorry LB team, but this is so frustrating and costly, it needs sorting out once and for all, if every other suggestion has been exhausted, belts, X/Y alignment, laser beam alignment, bed level etc , then the issue must be at the door of the software.
I have proven, beyond doubt, its not my machine settings, nor Thunder Laser nor Trotec machines, its a software driven problem being somewhat “masked” by Hide backlash.
BTW, i love LB, hate RD Works, but what option do I/we have but to use RD.
I would encourage all other users to try RD Works, if they have the same issue and see if it replicates in RD and report back.
Really look forward to having a serious conversation with the dev engineers at LB and would welcome all valuable input.
As mentioned. all my machine mechanical parameters are set up as per all the YT channels
Belt tension…check
X/Y axis 90 degrees to each other…check
Laser alignment…check
Bed level with gantry…check
Free running linear bearings and rails…check
Stepper motor belt tension…check.

Thanks for reading my rant, hopefully it can be sorted out!

Seems very odd that this hasn’t surfaced before, many of us cut circles and I think it would have been noticed…


First suggestion is to get a current version of Lightburn… I think currently, it 1.4.0

Check results again…

If it is the same results, I’d suggest contact with Lightburn… if this is a bug, they need to know about it…

:smile_cat:

I’d like to better understand the issue. I’ve cut literally thousands of circles on multiple lasers and do not have the issue you’re describing but I may be misunderstanding the issue.

Can you provide the following:

  1. Photo of cut circle showing the artifact you’re describing
  2. Full screenshot of LightBurn with design loaded and ready to burn. Can you have Cut and Laser windows visible during the screenshot
  3. Screenshot of Edit->Device Settings
  4. Upload .lbrn file with your design
  5. Upload .rd file generated from your design
2 Likes

Is something preventing this?

Took a look at your design. I don’t see anything inherent in the design that would be a problem.

Ran it without fuss with only changes to the speed and power.


Also ran a concentric circle test.

These were without “Hide backlash” enabled and run on 1.4.01 beta although my results are no different than on previous releases.

Unless the RD file is somehow being generated incorrectly not sure what would be going on. I wonder if somehow the RD file is being generated as if “Hide backlash” were enabled. This could possibly explain why RDWorks looks correct while LightBurn does not. Reviewing the RD file would confirm this.

If the RD file is being generated correctly then this type of issue is almost always related to mechanical backlash. However, your results in RDWorks would make that seem less plausible.

Do you have a backlash setting configured at the controller? Under Miscellaneous in Edit->Machine Settings in LightBurn.

I didn’t realize .RD was a forbidden format. Try adding .txt to the end of the file.

I just created a new LB file and seem to have recreated this problem, seems to be triggered by Hide Backlash optimization, AND its only a problem on the last shape to cut. I created a square using the tool and a circle using the tool; the circle was cut last and it clearly did not terminate the cut on the start point, as viewed by the Preview window. Then I created a 3rd shape, another square using the tool, and it was cut last and left its end point short of the start also, according to the Preview window.
To be clear, I did not actually cut the shapes on my GRBL device, rather I zoomed to max in Preview. I’ve never known Preview to show anything different than what is generated as gcode.

I don’t see anything odd with the RD file.

Can you check if you have backlash enabled in the controller?

Can you upload the file where this is occurring?

This sounds like a different issue than what @Paulco01 is getting, however. I don’t think what he is getting is seen in Preview, only the final output but he can confirm.

Cant see backlash in the controller.

Glad you understood the RD file… :wink:

Only seen in the final output not in the preview window.

Are you using “Backlash reapy optimize” in RDWorks?

yes, I agree, I believe what I’ve stumbled into is a different situation, although I’m not ready to call it a problem or bug. As I was rebuilding the test case to upload, the Preview shot did not exhibit the problem!, but I had changed layer settings. I’ve also discovered that Preview seems to show a red crosshair when the cut does not terminate on the start point, that convenient, and I suspect all is working as designed for now. I’ll open a new thread with test case uploaded to pursue this once I get my head around it. So far its related to feedrate and Hide Backlash as triggers, and I’ve excluded Constant Pwr and Scanning Offset Adjustment. Odd, but I’ve learned to trust the power of Oz @LightBurn :wink:

1 Like

I tried RD in standard format, really found it clunky. Went back to LB…lol

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.