Homebrew Accent Table Build

Depending on whether you really want to build this or buy it, I bet Bruce would sell you a sisbot sans table for your custom furniture. Or get a Mini and transplant the mechanism into that table.

I am sure that the table will be awesome to watch, and I'll enjoy making new tracks for myself and the community, but my main purpose currently is the physical build and optimization. I am also looking forward to the programming aspect of this project. Some of the ways that the physical interaction between Theta and Rho affect each other are going to be fun to dive into. It is pretty mind bending thinking about the differences between this machine and my printer and laser cutter.

I have a bunch of parts modeled in CAD, and I am starting the buildup of the Raspberry Pi with stepper motor and lighting control. My son is encouraging me to make the UI web based, so that visitors to my office can play with (limited) settings and see which tracks are coming up.

Brains!

Now serving!

I aim to use these two servos to control dual sided ball parks. One per side of the table for a total of 4 ball locations. If I can get the balls out of the play area, I should be able to have more room for cleanup tracks that go outside Rho 1.

Tomorrow is testing day for the servos and optical end stop limit switches.

Decisions, decisions...

I have received and tested out all the electronic parts of my build. Stepper motors are stepping, 64 LEDs are making fantastic patterns in two loops, optical limit switches are itching to be triggered, and the two servos are ready to receive the parked balls. My power supply and step down converter are able to power the accessories noted above, along with the Raspberry Pi, and motor hat without breaking a sweat.

Now, on to the prototyping stage!

In order to make real progress, I had to make the hard decisions about the physical limits of my table:

1. The Rho is going to be from 0 to 1 with no overrun. I don't need to watch out for the ball parks, because they won't be in the drawing area.
2. The radius for the drawing area is 171mm. This fits tightly inside the legs of the table I bought. There is a bit of extra room outside the drawing area, but inside the table. That's where I am putting the parking garages.
3. The radius for Rho is 158mm from center to center of each magnet. The extra 13mm is the radius of the magnets I am using + the amount the servo hangs inside + a tiny bit of breathing room.
4. Drawing area depth is 32mm. Top 12mm for LEDs, bottom 20 is parking garage doors.

Time to get to work. I love it when it is laser time!

Parking hits the spot!

I have decided that this will be a truly mixed materials project.

After a while thinking of a good way to make the acrylic sides of the drawing area with nice even bends, I thought it would be a good time to explore different materials. Aluminum angle would be easy to snip and bend every 5° or so, but it would be a pain to isolate the LEDs from it. I found some PVC edge trim at the local hardware store that is 1-1/8" tall. A bit shorter than I wanted, but it will work out. I took my parking "valet" with me, and lo and behold, it fits just right inside an 1-1/2" PVC coupling. PVC welds to PVC, so I am on the right track! (pun intended)

Testing a very crude model with a less than precise "robot" shows that things should work out fine. Each of two garages will house a large and small ball.

All your base are belong to us.

How do you laser cut the sand bed base when it is larger than your laser bed?

Measure twice, cut twice.

Tomorrow is the day when I get detailed and start putting together the final product. I may have a lit sand bed with hand powered magnets in my office by the end of the week.

These lights are loud!

I finished my sand bed with lights in the traditional place, and also edge lighting the acrylic bed. Not sure how loud your tables are, but this (maybe wrong type) sand, ball, and thin, drum like bed are really loud when I drag the magnet below.

Anyway, I am really happy with the way the lights work from below. Obviously, the video doesn't do it justice, and it would definitely look a lot better if I was drawing with a computer instead of by hand.

Any tips on the kind of sand in use on your tables would be appreciated. Mine isn't white, but it seems fairly fine. Maybe finer wouldn't be so loud.

Here are some missing steps in my build.

Top photo is riveting the sides on the base after I carefully cut out the places for the garages.

Middle is after I bonded extra 3/16" acrylic to the LED cutouts. I cut apart each LED and soldered them back together with a ~1/4" gap.

Bottom is the nearly completed sand bed. I have since finished up the garages.

On to the mechanics of CNC art!

A carriage approaches!

Mechanical build is coming along nicely. I have a laser cut turntable with 480 teeth, slide mechanism is bonded precisely using alignment holes. I have the optical limit for theta using a short piece of black zip tie. Now the hard part:

I have to get this precision machine to mate to the table that is anything but precise. The band isn't round, or even oval. It is kind of egg shaped? The legs aren't at 90°, and there are only 4 screws holding down the top, but 5 holes.

Time to abandon the futuristic machinery in favor of old fashioned paper templates.

Hi Maker Bob, I've spent the last year or so lazily working on a sand table build and have gathered my research during that time over at https://markroland.github.io/sand-table-build/. The Resources links at the bottom may be of particular interest to you.

I've also invested significant time in build a web-based pattern generator as you mentioned, which is all open-sourced over at https://github.com/markroland/sand-table-pattern-maker.

Best of luck with your project.