Laser Cutting an Enclosure to Compare to 3D Printing and Milling

There’s always a next step, a new opportunity to learn. For me, that’s the best part of being a maker.  I’d wanted to come up to speed on 3 modern fabrication technologies, a filament 3D printer, a resin SLA 3D printer and a CNC router. At the Columbus Idea Foundry, a local makerspace, I had access to these devices. That led to my article, published on Make: in January. I wanted to understand the strengths and weaknesses of each of these technologies and as a beginner with them all, it was the perfect chance to compare.

My comparison project was to create an enclosure for a “Camera Axe,” a camera controller for high-speed photography developed by another maker, Maurice Ribble. I’d built the board from a kit. While it worked perfectly without an enclosure, it just seemed wrong to use it “naked.” But this was a challenge, the board was not designed with an enclosure in mid. ICs and capacitors stood taller on the board than the switches used to control it and the LCD display. Not an easy task.

My colleagues at the Idea Foundry were delighted with the article but they wondered why I’d stopped at 3 fabrication technologies. What about our Trotec Speedy 400 Laser cutter/engraver, they wondered. And so round 4 began. I’d never used a laser cutter either.

(Re)design

In comparison to additive approaches like 3d Printers and subtractive techniques like CNC routers, working with Laser Cutters is a different game altogether. One might call this fabrication model “assemblative.” Laser cutters can’t create cavities so creating an enclosure requires cutting an assortment of parts and assembling them. But as I looked at laser cut enclosure designs on the web, I was bothered by their appearance. Some looked like sandwiches gone wrong, with the height of the enclosure provided by many stacked layers of acrylic. Others seemed more like3d jigsaw puzzles with the sides and top made of interlocking parts. I wanted to produce an enclosure with a more professional look. As before I also wanted to create the enclosure using only one tool, in this case the laser cutter, and limiting myself to fasteners as add-ons. I chose acrylic sheet as my material, both because it is relatively easy to cut and because I had a source of inexpensive scrap at a local distributor.

The Plane’s the Pain

Having worked in the world of true 3d, rethinking the design as a set of planar surfaces seemed very restrictive. As mentioned above, this enclosure needed to accommodate a range of bumps and recesses.  What would I be giving up? In some ways, the step forward seemed simple. Just use Fusion 360, my 3d CAD software of choice, eliminate any chamfers on horizontal surfaces and then cut the enclosure into planes. I’d then just cut out each plane on the laser cutter. A top, a bottom and 4 sides.

But thickness is infinitely variable in a 3d print. Acrylic sheet comes in defined thicknesses, 1/16”, 1/8”, etc. My original design had recesses in the top to make the buttons accessible and recesses in the top’s underside to accommodate the parts that projected above the height of the tall buttons. Since the laser cutter can’t create recesses in a surface, each of those levels would require its own piece of acrylic.  But then too, the variance in height wasn’t necessarily 1/16” or 1/8”.  This was going to require a good deal of experimentation, both in the CAD tool and with actual prototype parts.
After a couple of iterations my design came to look like this:

From Fusion 360 to Corel

Taking a part from CAD design to production with a 3d printer or CNC router is a two-step process. For example, with the Lulzbot TAZ printer, the design is exported from CAD and imported into Cura where it is prepped and then sent to the printer. Going from a CAD design to the final product uses with the Trotec laser system is a 3-step process.

1. Export the drawing from the CAD tool
2. Import the design into a vector graphics program where specific colors are assigned to define the areas to be engraved or cut
3. Print the file from the vector graphics program which triggers the Job Control software to load. In that tool, you then position the laser, position the part, define the laser’s speed and power for each area of the design and then kick off the job.

In that setting of speed and laser power, laser cutting shares another common feature with CNC routing. In the world of CNC, “feeds and speeds” are a bit of an art form. What material is being routed? How fast should the bit be spinning? How many inches/second should the router move? How deep should each pass be?  Settings for the laser cutter are similar. What material are we cutting or engraving? How fast should the laser move across the material? What percentage of full power should be used? And at what frequency should the laser pulses hit the material? As with the CNC router control software, the Trotec Job Control software incorporates settings for a wide range of materials. Still experimentation is needed.

The Trotec laser uses a vector graphics editor as a design tool. I had a copy of Corel Draw so that was my tool of choice. But how to get a drawing out of Fusion 360 and into Corel. Corel does accept CAD drawings in DXF format (only the pro version, not the Home version). But for the life of me, I couldn’t find a way to export a DXF out of Fusion 360. After far too many wasted hours I learned that you have to turn off “Capture Design History” to export a DXF file from a sketch. This makes no sense to me and is a significant headache. Once you turn off history to do the export, you can turn it back on but you will have lost all previous design history. Ugly!

Learning the Limitations

The middle section of my design (between the two top layers and the bottom) stood about 1 inch tall. My initial thought was to cut it out of one block of 1 inch acrylic. That seemed like a good idea but I ran into two problems. First, the walls on the sides need to be quite thin to handle allow inserting 3.5mm cable ends into the board’s jacks. I was concerned that those thin walls would deform under the heat of the laser. So, I split that level of the design into 4 components, two endcaps and two side walls.

I still wanted to create the endcaps out of 1 inch acrylic. I ordered a 12-inch square of 1 inch acrylic online, but once I tried to cut it, I encountered the second problem. The Trotec Speedy 400 is a powerful laser cutter/engraver, with a 120W laser. But that was no match to 1 inch acrylic. I was able to cut through the piece but only slowly and the result was sad. The holes started small but grew in diameter as the laser cut deeper. The sides of the cut were badly deformed. Ultimately I split each endcap into two pieces horizontally and cut them from ½” acrylic. I still ended up with a sandwich but not a 10-layer stack.

Surprisingly, while the half inch endcap sections held their shape, they did lose a bit of height, perhaps 1/16 inch. In the end, I used some shorter standoffs to mount the board and that rescued me from the height problem.

Registration

If you look closely at my drawing, you’ll see that the rear endcap has an opening for a power cable. To create this, I’d need to rotate one of my endcaps on its side and cut a U-shaped opening. That brought its own challenge. Once I’d pulled the part from the laser, how was I going to accurately position it for this side cut. After several failed attempts, I stumbled onto the solution in one of those late-night sleepless moments. I created a quick sketch of the endcap side with the U-shape opening:

I then placed a piece of ¼” inch acrylic in the upper left corner of the Trotec, a location I could replicate.  Next I positioned the laser to start the cut and set a “Marker” using the Trotec Job Control software. This gave me a laser position I could also replicate. This focus on replicating the location of the laser and the part is critical because each time you open the Trotec, the laser goes back to “home.”

With this system, I first etched my design onto the acrylic sheet in the Trotec. After the engraving was complete I placed my acrylic over the part to check that the hole for the cable was properly located. I then returned the etched sheet to the laser cutter and this time cut out the perimeter of the shape. Once the perimeter was cut, I removed the blank and replaced it with my actual endcap. Finally, I adjusted the job control settings to ignore the perimeter but cut the hole.  The result was a perfectly positioned hole through the side of the endcap.

Positioning Screws

Since I was using screws to assemble the enclosure I needed to cut accurate holes. That turned out to be effortless. But I was bothered by the notion that my screw heads (or the nuts holding the screws) would be projecting out of the bottom of the case. I didn’t want it to scratch a work surface. That led to a bit more experimentation. I was able to adjust the settings to engrave very deeply into the 1/8” acrylic for the base. But engraving half-way through the acrylic I could create counterbore holes that nicely held the 2-56 nuts. Problem solved!

The Joy of Labeling

If there was one limitation I found irritating in my experience creating this enclosure with 3d printers and CNC routers, it was the lack of a good way to provide labels for the switches and ports on the circuit board. I’d printed the filament version of the case in blue and creating decal labels with a blue background and white lettering turned into a genuine headache. I’d had to use color match the background of the label with the plastic, leaving the letters in white. It worked but the result was not very professional. Learning from that experience, I created my resin printed enclosure in white. That way I could use traditional decals with black lettering. But even here the result was less than professional looking. The decals were difficult to align precisely on the case. As for the CNC routed version in cherry, I just gave up. Nothing I could think of would allow me to create good looking labels on that surface.
By contrast, creating labels with the laser engraver is a piece of cake. Corel (or your vector graphics editor of choice) is adept at working with and positioning text. I edited the design for my “lower lid” to add the labels and printed it on black acrylic.

While the masking paper on the acrylic was still in place as a mask, I then filled the letters with white paint. Finally, I cut the upper lid in clear acrylic to protect the lettering.

So here it is: The Camera Axe enclosure in laser cut acrylic.

Conclusions

If the filament and resin/SLA printers from my earlier experiments are 3D and a CNC router is 2.5D, a laser cutter/engraver is decidedly 2D. In one way that’s an advantage. The parts coming out of the laser have a finished appearance. I certainly didn’t miss the time I spent sanding the plastic and wood versions of the enclosure. But more than with the other technologies, I found myself battling with the planar restrictions of sheet materials.

Surprisingly, the other downside to the laser cutter is in the choice of materials. While one can engrave many materials including glass, stone and some metals, cutting materials is much more limited. Wood and sheet plastic are the most common materials for laser cutting. Others like paper, cloth, leather, etc. can easily be cut but these soft materials aren’t appropriate for producing hard objects. And even with a powerful laser like the Trotec Speedy 400, as I discovered it may not be possible to cut deeper than one half to three quarters of an inch. Of course on the web I’ve seen laser cutters slice through three quarter inch steel. But those run to the hundreds of thousands of dollars.

On the other hand, the laser wins hands down for speed. Even though I was using a mix of 1/16”, 1/18” and ½” acrylic, the actual time from when I kicked off the job to when the part was complete was measured in minutes, sometimes seconds. The 3d printers and routers took hours to do the same thing.  Another way to think about this:  Once the design was finalized, the 3d printers needed between 12 and 18 largely unattended hours to produce the two halves of the enclosure.  The CNC router took 6-7 hours to do the same thing. By contrast, I could produce the 7 parts of my laser enclosure in less than an hour.  Sanding and assembly for the filament printer took about 3 hours, mostly sanding. For the SLA printer, it was a 1 hour task. Assembling the multiple parts of the acrylic case takes about 15 minutes.  Laser Cutting/Engraving wins that battle easily.

Despite the limitations mentioned above, I found laser cutting to be very satisfying experience. The machine felt exceptionally well designed with software well developed for ease of use, even for casual users.  That is certainly not a description I’d use for 3d printers. They all feel far less polished and rough around the edges.   Of the technologies I’ve tried (so far), laser cutting is the first I’d recommend to someone lacking the patience and determination of a determined maker.

I ended the last article with a table listing the characteristics of the 3D printers and the ShopBot CNC router. Here’s another column for that table: