Got a nasty one here. FL Cutout-5in Zap (a.k.a. “Zap”) is the original file, and FL Cutout-5in (a.k.a. “New”) is a rebuild where I only used 1 state outline.
I originally thought my SF-A9 had a screw loose. In the “Zap” output, A, B, and C were the same size, but C, D, and E were all different in size and rotation. So I switched to my S30 machine with burn corrections.
The S30 produced the same results with the “Zap” file. I created the “New” file thinking “Zap”might be corrupt. In that one A, B, D, E, and F were all good, but C was the wrong size. So I reinstalled LB and rebooted Win, still with no happy.
I tripple checked (3X) the sizes of the outlines in the grid, and found they were all the same.
I ran some tests using a line instead of a cut, and that made no difference.
I can’t exactly explain it, and it is probably a good one for @LightBurn to look at, but I think it has something to do with the “Xform” for those shapes.
I ungrouped the states of FL, and broke them apart, and then auto joined them. This “zeros” out the Xform(s) so they now look like:
No joy. In fact, the ones that were about right were wrong (A, B, and C, but F matched these 3 this time). D and E were actually distorted where the upper leg was bent down somewhat.
Thanks for trying and I appreciate the help! This file has gone thru 40+ Baltic ply without error. It was originally a screen capture of FL, then vectorized, then extra line removes leaving a single path. I then set the size so it would fit within a 6” square (1/4 of a sheet. I then did a CRL-D and made 5 copies. Finally I scooted them around so could get 6 to a sheet. And finally, I started cranking them out (I have a good customer).
Summary: We are chasing the tail, so I am going to create a new file from scratch.
I am not at my computer now, but I noticed that a few lines were large radius. Could they be a problem? I know LB breaks them up into line segments that approximate the arc. I was working on a new FL state shape in CAD, but am not home for about 2 hours.
Update: I switched to a laptop that has not had recent Windows updates. I was able to complete my customer’s order.
Because I am using the same Lightburn EXE and “…Zap.lbrn2” files, it HAS to be a Windows issue. Win tried to do an update while I was burning, but I blocked thaaaat in a hurry.
I will work around this, but if Lightburn Development (@LightBurn ) will pay round trip shipping for my laptop, I would be happy to let them research it. I have neither the software tools nor the patience to drill into this mystery.
If there is any file or information I can send to Lightburn (@Rick ) to help figure this out, let me know. The offending laptop has the latest Win downloads and is an ACER Win10 upgraded to Win11.
Not sure that will help because the error appears to be in the GCode output. But I will try it.
Now wondering if it might be the laser. Thinking MillMage could be affected, I thought of a test. Create the GCode for the program. Then use the good PC to feed it to my other laser. If it runs ok, it is the laser. If not, could still be the PC Windows SNAFU.
UPDATE:
The SVG used to create an entirely new file made no difference on the SF-A9 machine.
The GCode from the faulty PC output (to SF-A9) also produced non-matching outlines when run on the S30 machine.
So it is definitely an issue in the ACER laptop with Lightburn.
I am open to try something else if anyone has any ideas. I have exhaused my realm of experience here.
Can I send you the GCode file from MY PC? Your test produced ok output streamed and .GC file. From the same computer. Try my output from the wonky PC, it only has 2 FL’s on opposite sides. They will not match.
If you save the GCode, then import it back into LightBurn, you should get the same shapes. You should also be able to drag those shapes around and rotate them to see if the GCode tells them to be the same size or not.
If the generated GCode is telling them to be different sizes, then you might have rotary mode enabled, or something in your custom GCode profile that’s messing things up.
If the generated GCode does have them all the same size, then it’s likely an axis issue on the physical machine - If one axis is slightly off that could explain it - as in, if either your X or Y axis is outputting slightly scaled. A,B, & C are the same orientation, I’d expect them to match each other. D, E, & F are all oriented differently, so they wouldn’t match each other, or the first three.
