Att. Oz.
“User Origin” / Current pos. change made no difference.
Yes, my theory is also that it has to do with accelleration.
Can I slow that down in Lightburn ?
Go to ‘Edit’→’Machine Settings’.
Machine Settings
Opens an editor that allows reading and writing firmware settings from supported controllers.
Hey Kim,
thank you for sharing especially the video of your job, it was very helpful for me. You already loose your position while the laser is working on the circles so what I would suggest next is checking your Machine Settings / Cut Parameters and decrease the Idle and Max acceleration values (write them down before you do and do a Save of your settings before!).
The Dialog within LightBurn should look something like this:
So here is how I would proceed if I had that issue:
I don’t think that you have to always do the whole job to see if it improves, I would just use the original settings, start a job and maybe run the first line of circles and hit the stop button. This will move your laser back to it’s previous origin and you could do a pulse again and compare it with the first one. You will soon be able to see how many circles are necessary to get you noticeable off.
Check that back after decreasing the Idle / Max acceleration values and see if your situation improves. At first I would probably decrease those values to a very low value just to be sure that they have an impact. And don’t forget to write them to the laser after you change the values and read them back just to check that all went well.
Good luck and cheers,
Brian
First picture is the original settings.
Second is what I changed it to.
It made no difference. Still offset in origin…
Hmm… that values are not very high. Have you noticed any difference while watching the job? Was it noticable slower than before? And can you please try to narrow down how much your the job has to be done to get you out of your origin?
Please try to only do a “Frame” of your job and see if it will return exactly to the same origin as before by doing a short pulse, frame and pulse combination.
Cheers,
Brian
Try lowering your ‘Min Acceleration’ value as well - it’s currently set to 100, but lowering that could help. What kind of machine do you have, and more specifically, what size is it?
Att Oz.
Machine is red/black china, 80W CO2, Ruida 6442G controller, bed size 90 x 140cm
Ok, tried with these settings:
Att. Brian.
Tried to do the frame and cross, and then the same with 3 rows of circles, then the offset becomes visible.
If I only do the square and cross, there’s no offset.
Made with the same cut settings as my answer to Oz.
Your lines and circles are in general and compared with the not too high acceleration very shaky. Can you please check your belt tension and report how loose or tight they are? As well as if the lens is sitting tight within the tube and the tube itself has not much of a play (try to push/pull the laser head tube while the machine is on and see if it is able to move it a little).
And please try what Oz has mentioned as well, reducing the min. acceleration.
Despite not fully understanding why the RDWorks output is that much better I would bet that you have some serious mechanical issues that also relate to the drifting origin you are experiencing.
Maybe it could also help if you could post a picture of your full machine bed (including the mechanical driving parts in the back and on the sides and the stepper motors). Maybe I can spot something that might be an issue. I had a cheap blue and white Chinese laser and had all those issues you are facing right now (and some more) and decided to build a new one from scratch by myself and I have learned quite a lot on that road.
Cheers,
Brian
You have something fairly wrong in your mechanicals somewhere:
A straight line should not look like a drunken lightning bolt. 
As Brian says, I would look to make sure all your mirrors and lenses are tight, belts are tensioned properly, and also check that none of your bearings & set screws are loose, missing, grinding, etc.
I totally agree about my lines look like sh.t, but I would like to take one step at a time.
I believe, that all mirrors are firmly in place and secured and belts are pretty tight, about like a bass guitar…
It doesn’t explain the difference between the Lightburn and the RDWorks offset.
You asked for pics of the machine…
Actually it could - They plan the cutting in a different order, and the order of the cut can make a huge difference in the amount of induced resonance or machine slippage. If your machine is wobbling that much when running simple straight lines, your issues are almost certainly all mechanical, and it’s just the different cut order that’s causing the LightBurn output to make the machine slip more.
Ok, Oz. I might make sense…
So I should make the testfile, so the cutting order is exactly the same in both Lightburn and RDWorks ?
BTW: How tight is tight enough for the belts ?
That would be a much more accurate test, yes.
Regarding the belts, it’s hard to say with words - they shouldn’t be stretched, but they shouldn’t have any slack in them either. If they’re much too tight they can wear too fast, and cause vibration, but it would be from the pinion teeth impacting the belt teeth, and would look much different than what you’re seeing. That looks like a loose mirror mount, lens, or a bad bearing block on the laser head.
Thanks, Oz.
I’ll lok into that tomorrow, almost bedtime here.
Will check mirrors and lens are secure, bearings for free run, and I think I might loosen my belts a tad.
Then make a file that runs equal in Lightburn and RDworks…
Regards, Kim
UPDATE:
Checked mirrors and lens, all secure and in good shape.
Have checked the rails, and they are all running smooth.
To check the rails, I had to remove the belts, and when I mounted them again, I didn’t made them as tight as before, and that seems to have improved on my lines.
See before and after picture.
It didn’t solve the offset problem, but see my answer to Daboo
I had a similar issue when I received my new laser with a ruida controller. It drifted when it moved to the left after moving to the right. I turned off “pwm rising edge valid” in the controller settings under the x axis settings and that solved my issue.
1 Like
BINGO… Thanks David.
I had to turn off both X and Y “pwm rising edge valid”, but then it solved the problem.
I attach a picture, where I first ran the Lightburn file, and then the RDWorks file on top of it, with the same origin.
And… TADAA, no offset
2 Likes
It’s very curious that this didn’t happen with the RDWorks version but nonetheless, congratulations to a significantly better output and in the end with a quite small change. 
Now you could tinker with your acceleration settings again (change them back or increase them step by step and see how far you can go with still a clean output).
Cheers,
Brian
It depends on where the direction changes happen - for the output to shift, you have to change direction in X at the same time as continuing a move in Y, or vice versa, and if you perform the opposite moves, it’ll “undo” the lost step.
I haven’t seen an offset due to the ‘rising edge’ setting on vectors before - usually it’s so small that it’s nearly imperceptible, but this test has enough direction changes in it that it’s obviously showing it as well.
The usual giveaway for the “rising edge” issue is when an engraving leans to one side, like this:
Thanks for jumping in here Brian - I’ll have to remember this one myself. 
1 Like
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.