12/14/2025 8 PM - Made the sketch

I drew out the marble run path today.

I plan to have the run use 10mm steel marbles, so the elevator can be completely hidden. This is because I want to make this run similar to a "perpetual" motion machine, where the power source is hidden.

My idea is to use electromagnetic coils to move a steel ball up a ramp without any moving mechanisms, similar to how a coil gun works.

12/14/2025 9 PM - Made two elevator designs

The hardest part of this design will be the elevator, because it needs to be able to move the marble up the ramp without any moving parts. To do this, I chose two different designs.

The first design consists of coils hidden under a track, to pull the marble. This is easier to hide, and will require less work to make work.

The second design has coils completely surrounding the track, which should be able to move the marble better. However, this looks less clean.

The CAD for the first design has the coils bolted in with M3 bolts.

The M3 bolts through the center also help focus the magnetic field, which should make them a bit stronger.

The second design has the coils built into the track itself.

This design has the coils a lot closer to the marbles, and probably will work much better.

The electronics will be figured out later, but will probably be controlled with mosfets and possibly some sensors to detect when the ball is in the right position.

12/19/2025 8 AM - Redesign

After printing the parts, I realised that the coils were way way to small to actually move the ball. I then did a redesign, having larger coils that completely surround the track, which should work a bit better.

The new track should work much better.

12/19/2025 9:01 AM - Printed redesign

I printed the redesign and wrapped the coils.

Aluminum tape was used to turn on the coil when it got close, a very janky solution but it worked. This variation moved the marble, but it didn't get enough energy to move up on its own.

On a whim, I tried a small 5mm spherical magnet

I found that this worked much better, and it could even shoot out with only one coil.

I then redesigned the coils again, so they could be reused on different tracks without needing rewinding.

12/19/2025 9:02 AM - Printed redesign #2

I printed the 2nd redesign and wrapped the coils.

These coils were able to shoot out the small 5mm balls very quickly, so now I just need to figure out how to make a track. A benefit, is that I only need one coil to move it to the top of the marble run, so I can have multiple upwards segments, but I am not too sure what I will do yet.

Hopefully, I will use hall effect sensors to detect the location of the ball, so it knows when to turn the coil on.

12/20/2025 - First track design

I started off by designing part of the track. This is to see if my coils are strong enough to move the ball to the top, and so I can figure out where the landing to the jump will go. I was too lazy to design supports, so I will just leave the printer's support structure on and use it as is.

1/1/2026 1:50 PM - Track Testing

I printed out a couple variations to the track, and tested the marble run on them. I eventually broke apart the run, to test ramp angles so the ball can still move up.

I also used a L298N stepper driver to control the magnet, and an arduino nano to pulse the coil.

Now, I need to make all the ramps in the design roughly 45 degrees, because that is what I tested to work the best.

1/1/2026 1:52 PM - Finished Track Layout

Using what I learned from testing last time, I finished the CAD of the track layout, and made the track.

3D sketches in fusion 360 are honestly the worst thing I've ever worked with.

Next, I will need to print out the track, and see if my track works.

1/1/2026 1:56 PM - Designed Supports

While waiting for my parts to print, I designed the supports for my run. Not all of them are finished, as some are just sticks that don't connect up to the track

The coils were also redesigned, which means I need to rewind them...

1/16/2026 8:43 PM - Redesign

so uhhh... the design didn't work.

I printed out the parts, but it was honestly too tight and such so it didn't end up working. What I did instead, was I took random parts, and arranged them until I found an arrangement that worked.

This worked decently, and was pretty simple so I chose to use this version instead.

1/16/2026 8:47 PM - CAD redesign & build

I put the redesign into cad.

Honestly, it worked better than expected. I later added some paper walls to prevent the ball from falling off, and covered the ramp with a sheet of plastic(tape) so the ball wouldnt get flung off.

1/16/2026 8:49 PM - 3 more designs

I iterated 3 more times to figure out exact dimensions.

Some key features I changed:

• Turned the track concave to prevent the ball slipping out
• Decreased the angle of the ramp to make going up easier
• Increased the radius of the initial curve to prevent getting stuck
• Changed the profile of the curves before & after the track to prevent the ball getting stuck

1/16/2026 8:52 PM - Final Print & Design

I designed the final iteration & built it.

This was a collection of all previous designs and iterations, and I am quite proud of it.

Additionally, I soldered together the electronics, so they would fit under the case.

Building the track did not take quite long, but soldering that perfboard was really really really terrible.

Here is the wiring diagram: