PIErography

Design Concepts:

For the base of the machine, we laser cut a pegboard sheet to act as the work surface. We chose pegboard because it already has a consistent hole pattern, which makes it easy to mount accessories and gives a built-in grid for clamping anywhere on the board. To make the board actually usable for fastening, we installed wood T-nuts on the underside. That gave us strong threaded mounting points without needing to access the bottom during every setup.

To hold down workpieces, we laser cut clamping bars that span across the T-nut locations. Each bar extends just onto the top face of the wood, so it presses down near the edge of the part instead of covering the design area. We added long slots in the bars rather than fixed holes, which made the clamp positions adjustable. That way we could rotate the bars, slide them closer or farther from the workpiece, and clamp in different orientations depending on the part geometry.

A big design goal was keeping the clamping system low-profile. Since our tool moves close to the surface, tall clamps would risk collisions or reduce the usable area. By using thin laser-cut bars and keeping the hardware compact, we reduced the chance of the Z-axis or pen mount hitting the clamps during a burn. This also made it easier to burn pieces that were close to the edges of the board.

Another benefit of this setup was flexibility for weird shapes. Flat rectangles are easy to clamp, but irregular pieces can shift or vibrate during long burns if they’re only held in one spot. With the pegboard + T-nut grid, we could add multiple clamps at once and place them wherever we needed, which helped prevent rotation, sliding, or lifting. If a workpiece had a slight warp, we could clamp at multiple corners to force it flat against the board, which improved consistency in burn depth.

In practice, this was one of the parts that made the machine feel “usable” instead of just “functional.” It made setup faster, it let us swap parts without redoing the entire layout, and it gave repeatable positioning because we could reference the hole grid each time. If we had more time, we would likely add reference features (like a printed/etched coordinate system or a physical fence) so that aligning a piece to a known origin was even quicker and more repeatable.

Design Concepts:

For our x-axis, we started out with the basic idea of having an idler pulley on one side and a stepper motor driving the pulley on the other side. We wanted to attach the pen holder to the belt so that it would move along the x-axis as the motor moved the belt. To make sure that it would be secure, we added the mounts for the rail.

We were originally planning on having each side supported by the same linear slide, but we found that each rail introduced a lot of friction into our y-axis movements, so we changed to a slider that mounted onto the extrusion. This made it so that we had to change the base of the mount so that it would be able to fit on the new slider.

We extended the slots for the motor to allow it to slide back and forth. This would allow us to tension the belt. We also added a wall on the side with a screw hole in it which would allow space for a press plate that could be screwed in or loosened to apply more or less pressure to the side of the motor, for tensioning.

When we tested the early versions of the x-axis, we had the motor and the idler pulley much lower down and they were at the edge of the mount. To make the spacing work better with the z-axis, we had to move the pulleys higher and further in. As a result, we found that we needed much more support in a lot of key areas when testing our design.

The tension from the belt was causing the idler pulley support to bend inwards and we were worried that the axle it was on would snap. We also noticed that when we would push on one side, the other side would lag behind slightly, resulting in a noticeable gap when the side we were pushing on would reach the end of the y-axis.

We figured that this was a result of the shallow supports for the rail as well as the pliability of the 3D prints. We also noticed that moving the motor in made supporting the face it was mounted to much more difficult since there was a screw hole right next to the motor’s side. Since the other side of the motor hung out over the edge of our machine, the best side to support the face on to prevent bending was the one that was over the screw hole.

After our initial testing, we made a bunch of adjustments to improve the stability. First, we changed the depth of the rail supports from 5 mm to 10 mm and added support holes that we would be able to run a metal axle through to provide more stiffness without losing more workspace to the rail being pushed further into the mounts.

We also added a second side for the idler pulley mount and changed it so that the axle would no longer be 3D printed and would instead be a metal axle that we would put in after the print. Lastly, we added an I-beam under the motor so that it would no longer be hanging off the back of the mount, and we added a roof to the motor mount which would connect it to a wall that was perpendicular to the mount, allowing for better support.

As a part of some final adjustments after more testing, we added a slot on top of the rail mount for the limit switch to go in and a slot on the side of the mount to allow for wire management. We also added holes into the top of the motor mounts so that we would be able to see if the rails were fully pressed into the mounts.

Finally, we extended the rail mount holder further back. After the initial testing, we also added the belt mount on the side of the motor mount, which would allow it to be mounted to the y-axis belt so that it could be moved back and forth.

After doing more testing, we found that the 3D print tolerancing was causing issues where the distance between the two sides of the mount was slightly less than that of the rail length, so the rail would be pushing the two sides apart, causing issues with movement and leveling. We also extended the back part of the rail mount holders back to allow for better support for the rail holder. This also allowed us to support the pulley walls on the idler pulley side, so they would be less likely to bend.

Design Concepts:

The Y-axis mounts were designed to clamp onto the aluminum extrusions while also allowing belt tensioning. We added slots into the design so we could adjust tension by shifting parts slightly rather than being locked into a single position. That ended up being important because belt tension made a huge difference in how smoothly the axis moved and how much backlash we got.

To attach the mounts to the extrusion, we added press-fit style protrusions that line up with the extrusion profile. The goal was to make it easier to assemble and keep everything square without needing a bunch of extra hardware. We also learned pretty quickly that if the Y-axis isn’t level, the whole machine feels “off” and the motion gets inconsistent, so the base of the 3D print had to be thick enough to keep the axis aligned and stiff. Since that thickness adds weight and extra 3D print time we added light-weighting cutouts to remove material where it wasn’t doing much structurally, so the part stayed strong but didn’t feel unnecessarily bulky.

For the limit switch placement, we mounted it on the motor side because it was closer to the motor box, which reduced wiring complexity and kept the switch wiring shorter. That also made it easier to route everything through the extrusion slots instead of running wires across open space.

Looking back, one improvement would be adding more fillets—especially in the direction the belt pulls. The belt applies a lot of force in a pretty specific direction, and some of our sharper corners / thin features were definitely stress concentrators. More fillets would make those areas less likely to crack or snap over time, especially since 3D printed parts are more sensitive to stress concentrations.

Design Concepts:

The Z-axis assembly moves the wood burner up and down and is the part that is connected to and moves on the x axis.

Design needs

First, we determined the design needs for the Z-axis mount. It had to attach to the linear slide on the x axis, clamp the ends of the pulley, hold the wood burning pen, and be able to move the pen up and down. We wanted the Z-axis mount to be small and lightweight so the other axes had less mass to move. When testing the woodburner, we found that the depth of the wood burning tip into the wood had a large impact on the line produced. We knew we needed precise control of the pen, and it had to be sturdy and easily adjustable. We determined a lead screw would be best for this.

Fundamental design

The fundamental design for the Z-axis mount is a woodburning pen clamped into a slider that rides on support rails and is moved by a lead screw. A motor is attached to the lead screw with a coupler. A mounting piece holds the rails and the motor in place. The back of the mount has holes to attach to the X-axis linear slide and a clamp for the belt.

Manufacturing

We considered using a mill to make the mount and slider so that the holes could be precisely positioned, reducing friction and instability. This would have been very time consuming and difficult to iterate on, so we looked at 3D printing instead. The rods and 3D prints would have some give to them and we designed the slide to be tall enough to not be able to rotate and jam on the rods, so it is not crucial to have exact precision. We decided to go with 3D printing the mount and slide.

Sizing

We debated a lot about the travel length of the Z-axis. We wanted a enough movement to allow for different thicknesses of wood to be used in the machine, but not make the assembly too tall and unstable. The shortest lead screw available was 100mm, which was close to the length of the axis we wanted, so we designed the mount around it.

Rod mounts

We were able to get free metal rods that were about 95mm and thought it would be easy to use them. The first problem we ran into was the coupler and motor shaft adding length at the top of the lead screw such that our rods couldn’t span the whole distance. Our first design had a large block that offset the motor and shaft collar to counter this.

We were not able to find bolts that exactly matched this size so we couldn’t attach the motor. It also would have been very difficult to get the bolts in with this configuration. We tried another design where the motor mount was thinner and there were parts extending out for the rods.

Ultimately, we decided this was not necessary and got longer rods that we could cut to length to match the motor mount.

Hole tolerancing

Another problem we ran into was tolerancing the holes for the rails. The mount needed to have a tight press fit to keep the rod in place and the slider needed to be just loose enough to slide easily without wobbling. We tried printing the pieces then drilling out the holes to exact size, but this melted them and made the holes worse. We printed a test piece with different holes to see which would be the best fit for each.

We somehow read test piece upside and backwards, and made the holes way too big on the next print. With a little more brainpower, we got a perfect fit.

If we had more time we would add something to secure the rods in place other than a press fit. The press fit was good at first, but wore out over time, allowing the slide to move the rod and fall out of place.

Slide and wood burner clamp

Our teaching team raised a concern about the tip of the wood burner getting caught when it hit different grains in the wood and recommended adding some compliance to it. We thought about using springs or elastics, but figured it would be difficult to make it stable enough. We decided to mount the pen in a sponge. Unfortunately, the two sponges we tried shrunk and hardened when they were dry. The sponge had to be clamped extra tightly, but it still came loose over time. If we had more time we would have tried different kinds of foam. The sponge also may not have been needed at all, because there was enough play in the rest of the machine.

3D printing

When designing the parts, we kept in mind how they would be 3D printed. We decided to print the mount on its side so the filament layers were strongest across the height of the piece and supported the top and bottom. We printed the slide flat so that the holes were as smooth and consistent as possible. The areas that needed precision could not be printed with supports.

Belt clamp

On the back of the Z-axis mount there is a small channel that the two ends of the belt rest in. Two other pieces have grooves in them that match the teeth of the belt. They are bolted to the mount, clamping the belt in place.

Limit Switch

We added a limit switch at the bottom of the axis to make zeroing to the lowest point easier and stop the tip of the wood burning pen from hitting something and breaking. At the end of the project, we learned that it would have been better to mount it to the top since the pen should not have to go all the way down to zero. While testing, there were a few times where the slide was raised too high and hit shaft coupler, binding it and getting stuck. We had to use pliers to get it loose. A limit switch on the top would help prevent this. Ironically, a previous design with the rod standoffs would have had a better hard stop that also prevented binding.

Design Concepts:

We made the electronics house to keep all of the electronics in one organized place: the TB6600 drivers, the proto board, and the Arduino. Early on, our wiring was kind of chaotic, and troubleshooting was taking forever because we couldn’t easily track what was connected where. Having a dedicated enclosure made it way easier to wire everything cleanly, label things, and quickly access components when something wasn’t working.

Inside, we laid it out so the motor drivers each have their own “shelf” row, because each driver needed similar wiring and it helped us keep the wiring grouped by axis. Then we mounted the Arduino and proto board up top like a service section, since those were the things we were constantly touching while debugging GRBL, changing pin connections, or checking ground routing.

We laser cut the motor house out of MDF because it was fast, easy to iterate, and stiff enough for an enclosure. MDF also worked well because we could cut precise slots/holes for wires and mounting points without needing complicated manufacturing.

Looking back, even though having exposed wires was nice for troubleshooting, the final design would be better if the box was bigger and more enclosed. A door or cover would make the machine look way cleaner and also protect the wiring from getting snagged or bumped. Ideally it would still be easy to open when we need to debug, but closed during actual operation so everything is protected and hidden.