Introduction: Maze Table That MOVES!

About: Wood working and electrical engineering....oh yeah and building stuff

Welcome! To my project that has been stressing me out for

the past 12 weeks. I’ll explain to you my journey building the Maze table mainly because my friends and family are sick of me talking about it. Im currently enrolled in a CNC furniture design class at Purdue that is focused on the sustainable design of wood products. Wood working is passion of mine and I thought this class wood be the perfect excuse to get into a shop that has every tool I could imagine plus a few extras. This helped me understand which ones will go into my future shop and find out where my current skill level is at when given the keys to the kingdom.

Majoring in Electrical Engineering Tech I wanted to incorporate some of my knowledge on that side into the semester project. Through some back and forth with the class we came up with a table that would have a maze embedded and be controlled with a joystick. I absolutely loved this idea because it was something that I’ve never seen before and it had some potential to become a product that could be sold. So, if your reading this and would like a Maze table of your own reach out, I would be happy to discuss it.

At each point in this project where a choice had to be made in-between making it simpler or more complex, I decided to make to more complex. Everything from designing the electronics to how the top was milled out wasn’t the simple route. All this was done to create a seemingly simple design for a simple idea, making the maze tilt and pan, a bit ironic. A lot was tackled in this project, but it was all done to learn, sort of learning how to swim but starting in the deep end. I hope you enjoy this project and don’t forget to vote for it!

Step 1: Design

To start this, build off I first had to come up with a design

on paper then transfer that into CAD. This was done to not only get a scale of each of the parts but also check and make sure that there was proper spacing for the maze to pan and tilt. Fusion 360 was used for this process since it was the CAD I had the most experience in and it was free for students! This was the first major project that I used a CAD with, and the use of joints and multiple components was completely new to me. I’ll attach the file in this step, I attempted to model it parametrically Unforutally that only partly worked.

The entire build was designed around a 24” circle of glass that can be found on Amazon HERE because it was the best value to size ratio I found. Since this was a fixed diameter the rest of the table was built into proportion around it. Another size considered was that of rough lumber, the legs were made of 6/4 boards and the top made of 4/4. This was done to minimize the amount of glue ups and surfacing I would have to do.

The top it is made of 16 22.5deg segments to complete the full circle. This was done because I could lay them out on a 7” board and mill them out with a CNC to get perfect parts as well as reduce the waste material to a negligible amount.

One goal for this table was to have it wirelessly controlled with a joystick. I ended up getting the code to work but the board malfunctioned due to improper storage and a wired solution was created instead.

In summary the design was based around standard measurements and efficient layout techniques to minimize waste material. This was done with a simplistic focus where the end user would have a low barrier to entry of understanding how the table worked.

Step 2: Purchasing Materials

Table materials

Even before going to the lumber supplier I planned out exactly what I needed and that’s what I bought. This was a mistake because I made a ton of mistakes milling the lumber and ended up having to go back to buy a few more BF.

The lumber purchased was locally supplied and urbanely harvested.

-4BF of Hard maple

-17BF of Walnut

Wood cost: $75

1 x Glass

1 x Ball Joint


1 x 9V battery clip

2 x Nema 17 motors

1 x Arduino nano

1 x power supply

1 x Joystick

2 x stepper motor drivers

2 x 5V regulators

NRF24L01 - For wireless com


Step 3: Top

As previously stated, the top was made from 16 segments,

each one was attach using a biscuit jointer. Referencing the top surface meant that any variation in the thickness of the piece would be showed up on the bottom. The top was glued up over 5 days with 12 glue ups and was glued up one at a time into two halves. This was then put through a large sander so the squeeze out could be removed. Gaps were filled in with an epoxy and sawdust mixture to match the color of the wood.

After the top was glued and thicknesses to its final dimensions a ¼” round over bit was used to break the edges and give a more finished look around the edge. Around this same time the glass arrived and to my surprise it didn’t fit! Turns out there is a slight bevel along the edge which increased the overall diameter. This was fixed with a rabbiting bit on the router table. I don’t believe I would make a circle like this again frankly, making a box with 4 miters is hard enough now try doing 16 of them. This step ate up a lot of time and a simpler solution would have been to make a large panel then get the circle from that.

The wood grain does look nice though!


16 segments milled out

Glued up one at a time


Round over


Check the calendar and realize that 2 weeks have passed

Try to use it as a hula-hoop

Step 4: Maze

This step was mostly spent on a computer generating the tool

path. An online maze generator was utilized to output an SVG file that was then uploaded into Incscape. Here the SVG lines were increased in size to make the walls of the maze thicker. Next the new SVG file was imported into Fusion, scaled and output as a .step file that MasterCAM could use. For the pocket to be generated the lines had to be cleaned up, trimmed and connected. This part alone took around 12hrs since each of the vertices on the maze had to be trimmed and connected. (This is what the coffee was for) This was done because a simple bitmask of the file in Incscape led to noticeably uneven lines that would show up in the final piece.

Before cutting the hard maple, a test was run on some scrap plywood to test out the overall look and size. It was here were I decided to add round overs to all the corners on the inside of the maze. My reasoning beside this is that it was less prone to chip out and it looked nicer. This added another 2 hours to the CAM process.

After all this hard work it was time to cut the maple piece annnnd it went excellent! I even skipped a class to stay for the final few passes, woops. I placed it leaning up against a wall and left for the day. When I came back a few days later it had completely cupped, as seen above, this was due to the heat being turned on in the building and the maze being next to a heater. I placed the maze in an area with a higher moisture content and a significant amount of weight on top. After a few more days the cup was mostly gone!

Step 5: Legs

In the pictures above you can visually see how the mortise

and tenon joint was carved along with how the 90deg cuts were made on the X frame.

The legs for the table were 30” long with a taper along two faces that started 8” from the top and going down to the bottom. The taper was cut on two of the faces using the table saw, this gave the slender look while maintaining some of the rigidity of the table.

To connect the legs to the X frame a mortise and tenon joint was used. The X frame was joined using a half lap and a 90deg notch was made using a band saw on the four corners. On the underside of separated by 90deg two notches were mortised out to hold and hide the stepper motors.

I wanted the ½ lap joining to be removeable since flat packing this piece of furniture was a goal of mine. Therefore, the joint isn’t glued together but rather two pocket hole screws connect the frame on the underside.

Step 6: Electronics

This aspect of the project has given me the most grief and

turned out lack luster. I used NEMA 17 stepper motors which don’t have the torque to run the table quickly and makes the game hard to play. These motors also heat up to a point that I’m concerned that the table might catch fire, which would not be good. I also tried to make the joystick wireless but ultimately improper storage lead to the breakage of that board.

Ill still post my code for the multiple attempts because at one point in time they did work, and maybe you have a project that it is better suited for.


For the transmitter a joystick, Arduino nano, voltage regulator, RF-24, and 9V battery were used. The Receiver used an Arduino nano, voltage regulator, Rf-24, and 2 A4899 Stepper motor drivers. This board did drive the stepper motors wirelessly, but I stored it next to a large breaker box and I believe the EMF induced something on the RF24 chip and fried it. Running out of time and patience the scope of the project was changed to a wired connection.


The wired connection was simpler to make and easier to code for. This code is also linked in this step with text explaining each line. The basic principal of it is that the stepper motors move 0-180deg based on the position of the joystick. This will mean that whatever the angle of the joystick is at the maze will also reach that position.

Future upgrades

I plan to swap out the motors completely to some heavy-duty servo motors and to use some RC car linkages. If this ends up working, I will update the Instructable. So stay tuned for that!

Step 7: Electronics Part 2

Furniture Contest

Participated in the
Furniture Contest