…then you have written it so that it could be misunderstood when you compared the problem with this picture. (or I interpreted it differently)
But it doesn’t matter, as long as there is a solution in the end for your problem.
I hope it will succeed.
This should drain my will too!!!
Or you are doing something wrong or you need to change tube supplier.
Unfortunately I can’t help you in neither way.
With a new tube, yes.
The only real mode test we have is looking for uniform scorches on the targets. The picture you showed has an irregular toasty rim and a light center, which is why I think the tube is kaput.
You’ll want consistently better scorches on better target material, but in the absence of those, I think the tube is kaput.
Cutting through a sheet requires a nicely cylindrical focus zone. The beam from the tube is not nicely symmetric, so it cannot focus into a symmetric cylinder.
QED.
Those are different results than in the low-res picture you showed before:
They are also different results than the original description & image.
The mechanics are (now?) not quite right, in a different way. Seeing as how you’ve replaced nearly everything, it’s a matter of fine tuning rather than something grossly wrong.
If the machine does not position the laser head correctly, fix that. Verify the fix using the same test patterns as previously failed; when it shows no further problems, you’re done.
Those are not different results at all, I did ask if you wanted a closer shot.
Didn’t seem necessary, given what I could make out.
Rule of thumb: better pictures are always better.
What are the chances this may be controller related?
I’ve tried changing the Acc factor, G0 Acc factor, speed factor. Swapped out mirrors again, checked theyre tight again. Slackened every belt off, tightened back up. I’m truly now stumped, changing the backlash settings doesnt either seem to make a difference in the wobble.
IMO: zero.
All of the changes you describe have no effect on mechanical issues.
Explain what we’re seeing here:
Is the laser head carrier running on what looks like a linear guide barely visible inside the beam or on the set of ball bearings crunching along the crusty beam?
This is good, because if that crusty beam was involved I would be forced to crawl out of your monitor and strangle you with the X axis belt …
It’s likely down to the intrinsic flex in the gantry.
This is a layered-paper design:
The long sides are parallel to the Y axis and the RepRap calculator suggests the laser head gets up to full speed about a third of the way along the side:
I do not know how the speed varies around the 1.5 mm radiused corners, but you can see the damped wobble along the short side immediately after the corner. It’s consistent on all the rectangles, so it’s inherent to the machinery.
This is on a 700×500 mm machine, considerably smaller than yours, with reasonable attention paid to mechanical stiffness.
The first wobble is about 0.3 mm peak-to-peak, so the deviation from the nominal line is maybe 0.15 mm. The second wobble is about half that size, the third is barely visible, and the trio covers maybe 15 mm.
Similar results on some beam targets at a jaunty angle suggest the wobbulations involve the entire gantry:
I’d expect somewhat worse results from your machine, due to the larger masses and longer beams involved. Running slower will certainly help, but there’s a lower limit to the practical speed.
That your machine does not return to the same position when commanded to the same coordinate remains a mechanical problem not amenable to software or firmware correction.
So the weird thing is this is a new issue not something that has always been occuring, it ran fine as a production machine for months so I know it can do it perfectly as is. Regarding speed changes it doesn’t really change much, even down at 30mm/s when im doing full depth cuts it still induces these wobble discrepancies.
Absent any proof to the contrary, the laser tube is defunct and will not cut properly.
It’s perfectly OK for you to maintain that’s not the problem, but we’ve already been around this loop twice and I have nothing further to add.
Hmm thanks. Very nice of you.
Is that the best quality you can achieve cutting paper?
Probably.
The machine gets less wobbly with lower acceleration, but the entire gantry / laser head seems underbuilt for real precision: too much off-center mass dangling from bent sheet steel.
The image was a test piece to see if I had the colors right, so I wasn’t worried about geometric perfection. It’s adapted from the Subpixel Zoo; other patterns will require much better fixturing.
I’m definitely not playing in your league!
I was helping someone here a few years ago that makes some nice vector engraved wooden medallions. He mentioned that he had a 20 year old GCC Laser that had more precise vector etching than the OMTech he purchased. But, the GCC was too old to justify fixing, when he could simply buy the OMTech
I’m struggling with what is the test for laser accuracy. @RalphU could you post the concentric diamond and circles file? I’d like to see what I get.
I am not sure what files you are referring to.
Perhaps the one from this post:
Thanks, interesting test pattern. My machine has a 400 mm/min speed and 1000 mm/min acceleration limit on X-Y. Not sure where this came from. X motion looks really clean but Y is a complete mess. its like the axis has a rubber band in it so anytime I come from some other Y location and changes direction it takes a while for everything to settle down, putting a wave in the line as it travels in X. you can see which way the machine moves by just replaying the pattern as a slower speed in Lightburn.
I’m not seeing stepping in circles, just a lot of wavy lines if I’m moving Y and X at the same time or trying to stop Y then move X. pretty wild and the controller solution is to just slow Y acceleration way down. Not sure what can be done to stiffen Y. note - I’m clearly not loosing steps here. this axis has a harmonic frequency that is excited by fast engraving.
So I don’t do a lot of fast engraving and I don’t really see this much in normal slow cutting. if this bothers me, I could slow down Y acceleration but on short distances that’s the same as just slowing down the speed. When you do a fill engraving the Y is changing small amounts as the X sweeps, so again I don’t think this wave would hurt.
back to this thread. In my machine for all the rubber band looking Y axis movement, I’m not seeing in stepping in the circles. Stepping to me means that it starts and stops the circle in a different place. I have seen some waviness before and this test pattern really helps show that.