[wooden ball machine]

Wooden Prototype

With much fanfare the first wooden prototype ball machine was completed on August 16, 2000. It uses 5/8" steel ball bearings running on a pine track with routed grooves. In places where the machine tends to launch balls out of the RBS there are railings made of copper.

You can see a larger image of the machine in action while you listen to a 50 second (439k) mp3 of the machine in action (yes, that's a large file, but this noise doesn't handle compression well). That's the sound of six bearings in flight, reaching the bottom tray and then being hand shuttled back to the top a few times. Thanks to Matthias Wandel for giving me the idea to put up a sound clip. His marble machines make much more interesting sounds.

Moving Parts

There's the inverted-T flip-flip near the top center, which alternates balls between two overall paths. You can see the whole mechanism. It's not very clever, because it's poorly placed and really left nowhere for the balls to go! In the picture, one ball is about to fall in and go to the left.

The Y-on-its-side just below that is both a delay and a flip-flop. It is slightly counterweighted (from behind) to rest in the clockwise position. Balls enter from the left (in the picture that has just happened) and tip the whole thing counter-clockwise and the ball comes to rest against the edge of the exit path. The next ball goes over the top and dislodges the first ball, releasing them both simultaneously.

The dead-end looking path on the left towards the bottom is actually a tipper, counterweighted from behind. Balls roll onto it and it lowers them onto the (uphill) path just below it, where they join balls coming down from the other direction.

Clever Bit

The clever bit was figuring out how to keep the two right-hand paths from the two flip-flips separate. The trick is the ski-jump (big hunk of wood at the middle right). Balls which take the top right-hand path have enough speed to jump all the way to the left-hand track (with the big spiky copper guides). In the photo, one ball is about to make that jump. Balls going to the right over the sideways-Y fall into the middle ramp because they're not going as fast.

Construction Details

The most intresting part is the making of the grooves in the track. I use two router bits, a core box bit and an edge-fluting bit with ball bearing guide.

A core box bit is useful on flat faces, especially on a routing table against a fence. You can change the depth of the cut by adjusting the bit. You can get these anywhere, even in cheap 1/4" shank versions.

The hard-to find bit is an edge-fluting bit. It makes a fixed-depth flute in the edge of a workpiece, guided by a ball bearing at the end of the bit. You can route the edge of any shape. The Amana #54302 cuts a 1/8" deep groove with a 5/16" radius, perfect for the 5/8" bearings I'm using. I got mine from Ballew Saw & Tool. In my experience it seems like it would be easier to use this in a routing table, but in reality that's a great way to launch your small workpiece across the room. It's much easier to control freehand.


[Ben Jackson's Home] [Rolling Ball]