The Smell of Molten Projects in the Morning

Ed Nisley's Blog: Shop notes, electronics, firmware, machinery, 3D printing, laser cuttery, and curiosities. Contents: 100% human thinking, 0% AI slop.

Tag: Laser Cutter

Layered Paper: Mariner’s Compass Block

The Mariner’s Compass pattern on page 133 of Jinny Beyer’s The Quilter’s Album of Blocks and Borders:

Mariners Compass – Beyer 133

Becomes a laser-cut layered paper design:

Layered Paper – Mariners Compass – Beyer 133

A face-on view with different colors:

Layered Paper – Mariners Compass – Beyer 133

This seemed like an appropriate use for the stack (well, several stacks) of colored paper I’ve accumulated over the years.

The illustration in the book is apparently a photograph of a quilt block Beyer put together, so I had to reverse-engineer the Platonic Ideal Block from the image:

Mariners Compass – minimal shapes – LB layout

Fortunately, after a bit of fiddling around, I could take advantage of the radial symmetry to duplicate most of the fundamental shapes, so producing the layout really wasn’t all that difficult:

Mariners Compass – LB layout

The blue tooling lines (upper left) run along the centers of what would be seams in a fabric block, with 2 mm circles defining the endpoints for ease in snapping the lines.

This being the first block I attempted, I did it all wrong. LightBurn can form the convex hull over a group of shapes, so I just selected pairs of circles, created the hull, and iterated for the minimal shapes required to generate the whole design. That produces the basic layout, but what you really want is the collection of shapes between those hulls that define the actual cutouts, which appears in the lower left image.

Don’t do it that way, as explained tomorrow with a different block.

With all the shapes in hand, you duplicate them for all the paper layers you need, removing the shapes corresponding to the color of each sheet. Sheets lower in the stack have fewer cutouts, with the pattern in the lower right being second from the bottom.

The four holes in the corners fit over rivnuts in a fixture aligning the sheets in a tidy stack:

Layered Paper – alignment fixture

Yes, that’s a blooper sheet.

All in all, it’s easier than I expected to get nice results.

2023-10-16
Layered Acrylic Smashed Glass Junk Collector
What else would you call it?

Layered Acrylic Desk Junk Collector – overview

It’s a test piece with adhesive sheets between acrylic layers:

Layered Acrylic Desk Junk Collector – edge detail

From top to bottom:
- Acrylic 4.3 mm – sidewall to corral the junk
- Acrylic 1.5 mm – top plate
- Acrylic 2.4 mm – two layers
- Acrylic mirror 3 mm
- Cork 2 mm – PSA backing
The pair of 2.4 mm layers add up to just an itsy more than the 4.8 mm thickness of the shattered glass atop the mirror. Unlike previous epoxy sealed coasters, the glass sits on a sheet of 3M LSE adhesive film to keep the pieces together, with the top 1.5 mm acrylic layer containing any slivers. Because there’s no epoxy involved, the project is finished with no muss, no fuss, no curing time, and no drippy edges.

The geometry comes from a scan of the glass piece:

Desk Clutter plate – smashed glass – Quick Mask

That’s the GIMP Quick Mask result of manually drawing around the perimeter with the center of a 25 pixel diameter pencil, thus creating a 12 pixel gap to ensure the glass will fit inside the cut shape: at 300 dpi, the 12 pixel gap is about 40 mil = 1 mm wide. Slightly less would work as well, although I’ve discovered some of the glass cuboids have non-vertical walls sticking out to the side above the scanner’s depth of field.

Scribbling over the interior with a bigger pencil clears it out and a few fill operations produce a binary mask perfectly suited for LightBurn’s Image Trace tool:

Desk Clutter plate – smashed glass – binary mask

Trace that outline into vectors, throw away the mask, and use the outline for a conformal cut.

The rings of acrylic and adhesive are 3 mm wide, generated from the outline by offsetting it 3 mm outward:

Desk Clutter plate – LB perimeters

The tooling circle around the perimeter simplifies drag-and-drop alignment, because the geometric center of the perimeter shape isn’t quite in the middle of where you’d (well, I’d) want to align it. Grouping the outline with the circle keeps the center snap point where it should be.

Narrow rings of adhesive sheet turned out to be even more unmanageable than I expected. Perhaps a better way:
- Cut the ring with tabs holding it to the center area
- Stick the ring + center to vinyl transfer tape
- Peel the protective paper off the adhesive ring
- Stick the acrylic ring atop the adhesive ring
- Sever the tabs to release the adhesive ring
- Peel the transfer tape off the ring
The “adhesive tape sheets for craft” are paper-based, rather than a plastic film, and are neither transparent nor durable. I used it mostly to get an idea of how well it sticks to acrylic, as it’s primarily intended for paper crafts.

The 3M LSE backing layer is plastic and the sheet becomes nearly transparent as the glass squishes down, although you wouldn’t want it on a mirror where you cared about the optical quality of the reflection. Underneath a chunk of smashed glass, it’s just fine.

All in all, it turned out well.

Next: how long does that craft adhesive last in abnormal conditions?
2023-10-13
Blockface Grips From Warped Acrylic
This is definitely not recommended practice:

Laser-cutting warped acrylic

It’s the sheet of 3/8 inch = 9 mm vintage acrylic that Came With The House™ and what you see is the relatively unwarped result of several months of pressure. The wood sheet in front puts the left edge mostly flat on the platform, with the steel disk weighting the rear end down; the front end sticks out through the laser’s pass-through door.

The general idea was to cut a rectangle that would be flat enough for the shapes I actually needed:

Blockface grip cutting

They’re grips for a set of Blockface printing shapes. The alert reader will note the rectangle sits atop three spikes, not the usual four; it was flat enough for the purpose, not actually flat.

The OMTech laser claims to be 60 W, but has an over-amped 50 W tube that’s barely adequate for the task. The cuts required two passes:
- 5 mm/s @ 90% cuts almost all the way through
- 10 mm/s @ 90% clears the cut enough to have some shapes drop out and the others be easily pushed through
Cranking the tube current that high isn’t recommended practice, either, but sometimes ya gotta do what ya gotta do.

All’s well that ends well, though:

Blockface grips in matrix

I also cut the Blockface shapes from double-faced adhesive foam and stamp pad rubber:

Laser cutting rubber stamps

To my utter astonishment, the rubber cuts did not smell like a tire fire and the neighbors did not rise up in arms.

I cut several sets of grips / foam / rubber and sent them off for final assembly. The recipient seems delighted with the results and has been hand-printing up a storm.
2023-10-12
Wood Inlay Earring

Trying a sequence of offset and raster filling produced a credible 0.5 mm deep inlay pocket in a chunk of acrylic:

Earring samples – veneer spade recess

The odd pattern inside the pocket comes from the offset fill:

Veneer spade pocket – LightBurn preview

Combining the two fill patterns produces a smoother bottom in the pocket, but it’s a good thing nobody will ever see it from the back side, because the offset fill concentrations chew through the entire 3mm sheet.

A spade-shaped adhesive sheet bonded the veneer into the pocket, after which I sanded the surface flat with 200 grit sandpaper and hit it with some polyurethane sealer. Next time I’ll knock the finished surface flat with 400 grit paper.

The gray acrylic disk has a rebate around its rear rim for the “gold” ring, with the adhesive disk gluing the black disk to the front:

Earring samples – veneer spade parts

Which looks like this when it’s stuck together:

Earring samples – veneer spade side view

I don’t like the raggedy surface of the rebate above the ring, although it’s not too awful in person.

Having gotten reasonable results with the acrylic pocket, I cut an acrylic spade and stuck it in place with another bit of adhesive sheet, leaving it standing proud of the surface:

Earring samples – veneer spade acrylic

The red earring used up one of the hundred slugs left over from the capsule plate; I’m ready for a surge in demand.

The adhesive sheet is workable in disk form, but the spade shape is entirely too fiddly. A dab of acrylic adhesive would likely work just as well and be easier all around.

Stipulated: they’re cute, but not that cute.

2023-10-11
Laser Cutter: Moving Mirror 3

With Mirror 1 and Mirror 2 aligned, the next step is positioning the laser head to put the beamline at the center of both the aperture and Mirror 3 inside:

OMTech CO2 Mirror 3 mount – realigned – Z screws

Raising the laser tube by 5 mm put the head’s Z axis screws in the middle of their slots. This had the additional benefit of letting me rotate the head slightly around the X axis to make it perpendicular with the bed, thus fixing its mysterious from-the-factory misalignment.

Centering the beamline horizontally required a few iterations of Mirror 2’s position along the Y axis, but eventually produced this result:

Beam Alignment – Mirror 3 detail – 2023-09-16

Those are five manual pulses with the head at the corners and center of the platform. I put the 3M target on the mirror rotated 90° from the proper orientation with the stretched scale aligned vertically and parallel to the slightly oval beam mark.

The F target shows the beam position inside the head just above the focus lens:

Beam Alignment – Focus detail – 2023-09-16

The little target in the middle gets centered on the nozzle by feel and shows the beam position within a 2 mm circle. The initial position was off against the side of the nozzle, but slight twiddling of the Mirror 3 screws centered it.

I centered the lower F target at the beam position using the red dot aiming pointer, then pulsed the laser to put a pinhole almost exactly at the graticule center. The larger scorch shows the beam size with the platform lowered 10 mm from the focus level. The Z axis leadscrews are not particularly precise and the platform moves by about a millimeter in X and Y as they rotate, so that’s about as good as it gets.

After all that, the laser behaves at least as well as it ever did and I feel better about having the beamline actually travel along the center of the optical path.

Now, back to cutting out interesting shapes …

2023-10-06
Laser Cutter: Moving Mirror 2

With Mirror 1 moved 10 mm to the right (as seen from the front of the cabinet), Mirror 2 must move 5 mm to fix the problem that started this whole thing and put the beam in the center of the mirror:

Beam Alignment – 2M detail – 2023-09-16

This puts the bracket holding Mirror 2 closer to the center of its x axis adjustment range:

OMTech CO2 Mirror 2 mount – realigned – X screws

Remember, the flange is fixed to the gantry and the bolts move with the mirror mounting bracket.

Raising the laser tube by 5 mm requires Mirror 2 to go upward by a bit to put the beam at the center:

OMTech CO2 Mirror 2 mount – realigned – YZ screws

The least awful way to make simultaneous X and Z axis adjustments seems to be by feel. Tighten the screws just enough to prevent the bracket from moving easily, then slide it while aligning the top edge with respect to the flange on the gantry.

When it feels about right, stick a target to the aperture, fire a pulse, check the results, and iterate until it is actually right.

The two screws holding the mirror mount to the bracket sit in slots allowing some adjustment in the Y axis, as well as a slight amount of rotation. AFAICT, the mount was rotated enough that the test pulses passed through the center of the aperture, but hit the mirror off-center as shown in the top picture. I aligned the aperture parallel with gantry, which should put the holder at 45° to the beamline, and hoped for the best.

With the pulse mostly centered, twiddle Mirror 1’s alignment to make the beam parallel to the Y axis, which eventually produced these results, with each target getting a pulse at each end of the Y axis travel:

Beam Alignment – Mirror 2 detail – 2023-09-16

Not perfect, but much better than where it started.

2023-10-05
Laser Cutter: Moving Mirror 1

Raising the laser tube 5 mm required nothing more than cutting and inserting 5 mm spacers and finding slightly longer M4 screws:

OMTech CO2 laser tube – 5 mm Z supports

I used an ancient adjustable inside caliper to put the tube the same distance from and aligned parallel to the partition.

Sliding the tube an inch to the left provided enough space to drill & tap two new holes for the Mirror 1 mount to move the beamline 10 mm along the X axis:

OMTech CO2 Mirror 1 mount – redrilled screw holes

I briefly considered crunching rivnuts in there, but the mirror mount expects to sit flat on the floor with no room for rivnuts. So it goes.

Although Mirror 1’s mount has some vertical adjustment, the central stem was already close to its maximum extension, so I cut a 5 mm plywood pad to raise the base:

Laser Mirror 1 – baseplate scan

Despite what the lighting suggests, it’s concave. The image was clean and contrasty enough to just trace into vectors with LightBurn, then Fire The Laser to cut the spacer:

OMTech CO2 Mirror 1 mount – 5 mm Z shim

If you’re wondering how that worked with the tube jacked up, Mirror 1 sitting on the scanner, and the beamline in disarray, there’s considerable benefit in doing things out of the obvious narrative sequence.

Reassemble the mirror, square the entrance aperture to the partition, fire a couple of test shots to center the mirror on the beamline:

Beam Alignment – Mirror 1 detail – 2023-09-16

And that part’s done.

2023-10-04