This file has two layers. The 1st cuts the inner shapes in the leaves and the 2nd cuts the branch shape out. The path that cuts out the branch completes, but many of the inner paths are not being completed. The image has the worst example but many others also have one small segment not cut.
Work Speed 250mm
Power 40%
Its hard to say, in general there are a few meanings about.
1st. Is your Model out of focus?- Because of a warped bed/carriage/platform?
2nd. Different Material surface could cause not cutted lines.
3rd. Maybe its a known Marlin bug.
Share your gcode to see if the lines are closed and not open like your picture.
Here is something to read (the solution is in the last post):
In my opinion your work speed is too fast, the power is too low, and it is not allowing sufficient energy transfer to cut all the through at the start.
Power and speed separately, as itās somewhat confusing to explain how they interact.
When you are cutting through thereās no reason to limit power to 40% like you have, put that at 100%, or at least sufficiently high.
The arrow location is the start of the segment, not the end. This is a fairly common failure mode when the laser parameters donāt match what the material needs to begin combusting.
This laser is not a CO2 laser, it does not ablate material away. It burns, and needs time and energy to do so, while also being in very good focus.
It could be just slightly out of focus, or moving just slightly too fast, but otherwise is mostly working. So dropping from 250mm/min down to 200 or 150 should be enough.
More sophisticated software, like Lightburn, allows you to specify a dwell at the start of each segment. It allows the laser to stabilize the cutting at the start, then you can move quickly.
But if you turn on the laser and immediately start moving away (like Luban) you need to have a slightly slower travel speed.
Likely doing one of the above would fix your issue. You could also likely increase power to 100% and possibly even speed up from 250mm/min to 300 or more, some experimentation would be required on a test piece to optimize the speeds.
Now I know what ablate means and how CO2 lasers on youtube donāt create char marks like our diode lasers.
Kalmdown,
If you are using the 1600mw laser like I do, on cardstock, I would be using 400mm/min and 100% power.
Yours looks like just plain thin paper, which might work with even faster speed.
Did you run a test grid on it? That will let you know what is a good speed/power ratio.
I always try to go 100% and as fast as possible if the results are good.
Lastly, if you paper was not perfectly flat, and areas of it are bubbled up, you will get incomplete cut lines as well.
This helped a lot. Printed it and was able to pick best speed for 100% power.
To cut standard copier paper that was 1000mm/min.
@rustyssjones
I was trying to do one of those in Luban, but the values where not saving
(problem reported, probably solved next version)
Can I suggest you to make a new post sharing this here and at Thingiverse?
It would help many people.
PS: my next test board is a slightly angled base to find the best height/multipass tolerance,
(to account for small changes in paper and Tetra Pack milk containers)
but only will produce it after being sure the values are saved.
It was very frustrating to discover that 50 different definitions were not saved last time,
but was very happy to find your post.
Thanks
Nelson,
Yeah, I spent a whole evening in Luban configuring 200 separate squares, only to find that it did not save any of my changes.
Lesson Learned. Iāve been using Lightburn ever since. Even so happy I paid their $40 after getting 3 months free.
Iām not sure why you are angling the baseā¦
Are you trying to create something like my third post here?
I can send the .nc file if you want. I didnāt post it because it is in 11 different files. Due to Lightburnās max 29 layers.
Need a channel to post these tests on so others donāt need to do them. Would be good to post a pic from the sideā¦ know what 6mm is but stillā¦
@rustyssjones, the ramp is really to find how many passes would override the effects of curved materials.
I was trying glass like Tip for laser cutting complex designs // use glass but only had picture frame glass nearbyā¦
So the ramp was a test for evaluating the effect of focusing errors.
In tetrapack milk packages there is often a 5mm focus variation
(it all started with the Snapmaker contest, will post results after meeting some zombie deadlines)
About Luban: reported the error and got an answer that next version it will be working. Probably one of those things no one noticed it stopped working.
About your model:
already have your .nc, you shared it alreadyāTHANKSā, and I have another suggestion:
mix the squares and the curves (a corner of the square with an arc) and it will address curves also
(theoretically the speed is the same, but in practice no)
and the cuts in the overcooked areas will be more solid (they burn away sometimes).
And please publish it in Thingiverse, many people will appreciate your effort,
and move on to other improvements.
Huh, yeah, youāre right, a half circle, half square shape makes a lot of sense.
On a side note, I did a test previously, testing if like the y and x-axis the z-axis can change mid burn/scan. And yes it can.
I used two z depths (starting and finishing) that ran the same time the x-axis changed.
This was just a simple note pad edit of the .nc file.
I canāt find the file now, but it basically etched a ruler marking at 1mm increments then went back and started drawing a straight line above the ruler left to right, rising in the z-axis (in a smooth slope) at the same time.
So say it was 14mm long, with a mark at every 1mm.
I set the z depth at -4mm (otherwise I hit my material) to +10mm. Fortunately the rate at which the x-axis motor and the z-axis motor travel are the same, so they finish their movements at the same time.
This will give you a ā¦ lets say āvirtualā rampā¦ without having to physically slope your material.
Yeah, when Iāve got time. 1000 projects going on.
You want the 6mm pic stacked on top of the snapmaker boxās cardboard for depth/height reference?
Great solution for a virtual ramp, would like to see that when you have time.
Also it would be great if the test pattern would have that feature,
there are many complaints about the factory height of the focus and that would help a lot the starting community: it would be hero stuff for starters.
For now there are too many computers in my life to enjoy ānote pad edit of the .nc file.ā
Waiting for the promised revision of Luban, the one that saves data to go back to it.
PS: sorry for the late reply, just now realized it has been 15 days since I did something other then papers in the computer!
