Unable to Calibrate X-Y Axis

LightBurn Software Questions
1

I’m not sure if this is software or hardware.

I usually work small dimensional projects. Recently, I downloaded a larger project (120 X 345 X 226mm) from Etsy. the designer was very precise in his joinery and it was impossible to get the continuous slots and tabs to line up. The actual “box” parts ended up about 2mm oversized. I called Omtech tech support who walked me through Lightburn calibrate axis in machine settings. I thought we had it with a 25mm cutout. I still had problems and tried recalibrating a 75 X 100mm rectangle. First a little under, recalibrate then a little larger, recalibrate a little smaller and so on.

What kind of precision should I expect? Over these distances, I’m about 0.5 to 0.7mm off. What else could be going on?

Thanks!

Dale

2

The reason you can’t use a cutout is the measurement doesn’t include the vaporized kerf: the hole will be larger and the cut out shape will be smaller than the nominal size. Adjustments that make any given size accurate will make all other sizes inaccurate, because they do not take the kerf into account.

Instead, engrave a square with a fine line on something cheap & flat, then measure center-to-center between the lines. This assumes you can make accurate measurements on, say, cardboard, which can be challenging.

Instead, I use the red-dot pointer on a good metal scale:

When you’re done, the laser should be pretty much dead on across the entire platform. When you care about the actual dimensions of the parts, you must apply a kerf offset to lay the edge of the cut at the nominal outline position.

All that assumes the machine is in good mechanical condition with no appreciable backlash. If that’s not the case, then fix the machinery before trying to calibrate it.

3

Makes sense! I have a Starrett digital caliper from my furniture building past.
Actually, 0.5mm differential on a 100mm test works out to about 1.5mm on a 350mm project piece; that makes a huge difference!

I assume you are referring to the “Offset Tool”. So the offset would be inside the cut line? Also, I need to know the size of the laser beam, correct? Do I then delete the original project line and cut the offset line?

Thanks!

Dale

Thanks guys!

4

Ed,

Now you have me thinking. Offset in or out? Here’s a screenshot of a portion of the project.

Dale

Screenshot 2024-04-01 084958
Screenshot 2024-04-01 0849581920×1080 418 KB

5

A little error grows when it’s proportional to the size!

That will work, but you can also apply an offset to the entire layer.

Negative values are “inward” and positive values are “outward” relative to the nominal perimeter.

The gotcha: layer offsets apply to closed shapes and can have peculiar interactions with complex patterns. You may not want to apply an offset to everything in the layer, which means you must move those shapes to a different layer.

With that in mind, I apply offsets to specific shapes needing a precise fit, so I can verify what’s about to happen.

The kerf width depends critically on the material, speed, power, and perhaps the moon phase.

A handy pattern / tool to measure the actual kerf in whatever you’re cutting will come in handy:

Think thrice, measure twice, cut once!

6

Than you Ed but I’m trying to puzzle out whether it’s inside our outside in this specific case.

7

For the black lines, all the fingers around the perimeter must have the nominal size in order to fit together. That requires an outward offset: the beam runs outside the nominal perimeter, cutting exactly on the line.

An inward offset would put the beam on the inside of the perimeter, so the material surrounding the part has the proper dimensions. Probably not what you want.

The red stuff in the middle is up for grabs. If the rectangular outlines in the upper piece are sockets for more tabs, then they also need outward offsets to put their borders at the nominal size.

If you think of the laser spot as a (tiny) round cutting tool, you won’t go too far wrong.

LightBurn counts boundary crossings on a per-layer basis to automagically apply layer offsets, so all the perimeters move in the same way relative to the part’s interior. This may be what you want, but it’s worthwhile to run a few tests even when you think you know what’s about to happen.

Some projects aren’t that ahem clearcut, in which case you must puzzle it out on your own. :grin:

8

Thanks Ed,

I made the kerf scale and top scale reds 5, bottom scale is 7. So, 5.7/20= .285mm. Sound right?

9

It’s in the right ballpark for a CO₂ laser cutting a wood-like substance: around 0.2 mm.

May be a bit large, but that’s exactly why measurements beat opinions: trust, but verify.

Remember that the kerf offset is half that value to shift the beam half the kerf width (in the proper direction!).

10

I’m back. Please see attached screen shots. While I have not changed kerf offset settings, I’m concerned that increasing the offset may make the tabs the correct size, won’t the slots also change?

Dale

Fish
Fish1920×1080 396 KB

Support Base
Support Base1920×1080 357 KB

11

Everything is deeply intertwingled!

Caveat: I don’t know how all those parts are supposed to fit together, nor whether the red crosses represent holes, so pay attention to catch my incorrect assumptions.

I assume all the lines in those layouts are on the nominal dimensions, so that an ideal zero-width Xacto-knife cut will fit everything together.

If you apply the correct outward (positive) kerf offset to the (black?) layer with the fish outlines in the first screenshot, then their posts come out exactly at the nominal width. The notches (slots?) in the black round bases (?) also cut to the nominal size, so the tabs should fit into the slots. In both cases, the kerf offset moves the beam outward so the inner edge of the cut = outer boundary of the shape lies exactly along the nominal perimeter.

If you apply the same outward offset to the (blue-ish? green-ish?) layer with the notches (slots?) in the second screenshot, they also come out at the nominal size to match the nominal size of the red crosses.

However, if the crosses in the red layer in the first screenshot represent holes inside the black bases, then applying an outward kerf offset to the red layer will make the holes too large, because the beam will travel around the outside of the cross: LightBurn does not know those are holes.

If you apply an inward (negative) offset to the red layer, then the beam travels around the inside of the cross so the surrounding material has the correct size after the cross drops out.

Instead, you can put the crosses on the black layer, in which case LightBurn knows they are holes by counting perimeter crossings from outside the layout into the closed shape:

  • Odd number of crossings = outside perimeter = object
  • Even number of crossings = inside perimeter = hole

So, yes, applying a kerf to a layer means everything changes: it’s your responsibility to know which way they’ll change. You should definitely practice on cheap material until you’re sure you understand what’s going on and, even then, expect a few surprises.

12

Preview DOES SHOW the kerf-adjusted output. If you want to see the relative directions of all the offsets, you can temporarily apply an exaggerated kerf offset value to make them visible. Kerf is typically a fraction of a millimeter, so it would be easy to simply add a digit to the left of the decimal.

13

Thanks Ed!