Volumetric 3D Display


A 3D image floating in air.  Built with something from an Evil Mad Scientist.  Using both coding and saber saws.  "Sign me up!" was our reaction to all of this.  We hope you enjoy our little project as well. See it in action on YouTube here. Just remember, it's really just this:

Peggy2LE kit + a box fan = 3D awesomeness (no charge)

There will be a quiz.  In 3D.

First things first, lets cover how it works.  Imagine a cube (as close to the one at right as possible).  Now, imagine slicing that cube radially, like making knife cuts across a cake, each going through the center and the full diameter of the cake.  On each slice, record where the cube crosses it.  Make 128 slices and you can capture a good amount of detail of the cube.  Now, "play" the slices one at a time, rotating a little bit with each just as you did when making the slice.  Tada!  You've got a 3D volumetric display.  Automate your imagination and you've got this project.

Here is the end product standing still so we can see it better. 
On the bottom, is a standard box fan.  The blades have been removed along with one side of the plastic grating.  (Pay no attention to the exercise weights used to hold it before balancing.) 

In the middle, is a Peggy2LE kit from EvilMadScience.com It has 625 LEDs in a tightly packed 25x25 matrix.  We're showing it here with some mounting hardware and a chunk of plastic to both balance and shield it.

Surrounding the Peggy are bearings, a cross brace, and some aluminum rails supporting the whole lot.

Barely visible, but highly essential are some magnets that the Peggy uses to know where it is.

The red lump on the right is a variac for finely adjusting the fan speed.  Who really wants to settle for the typical 3 speed fan anyway?

General Build Steps
The first step is technically to acquire the parts.  But, the parts you'll use depend on your local sources, experience, and personal take on the build steps.  So, we start here with steps and list parts second.

Assemble a Peggy2LE and verify that you can program it.  Hot glue down the large capacitor on the board.


Attach a Hall Effect sensor along with a pull-up resistor to the Peggy2 using a 3-6 inch cable.  We used a lump of servo motor extender cable because it has a nice 3 wire socket for the sensor.  Using a cable will give you flexibility in mounting the sensor to its final location, but still lets you test the code and sensor now.

Find a box fan about 20 inches across.  You're looking for a good sized motor and variable speed adjustment.  The motor size is essential to ensure you've got enough torque to throw the board around even if not perfectly aligned.

Remove both gratings from the fan.  On the side where the box fan mounts to support rails, find the main power cord and move it.  I like it routed to near the speed adjustment knob, creating a "back" side where the controls are located.

Measure and cut a section of aluminum angle stock.  In our build, we went side-to-side across the fan body, perpendicular to the fan's support braces.  This part is pretty fun, but watch out for aluminum chips that melt into your saw blades.

In the center of the angle stock, attach a bearing using a plastic insulator between the metal bearing and the stock.  Use non-conductive nylon bolts to hold the assembly together.


Mount a shaft coupler to the motor shaft.  This is technically just 1/2 of the coupling.  The other half will be on the shaft attached to the Peggy.

Attach the other 1/2 of the coupling to a threaded shaft, leaving the threaded end open and available to the Peggy.  Snap that onto the motor's coupling.

Align your bearing/insulator/bar assembly perfectly over the motor's shaft and onto the fan side walls.  We used some spare plastic as shims to get the desired height since the fan walls were not tall enough.

Once perfectly aligned, drill holes through the fan's steel case and bolt down the assembly.  This and the prior step are good points to practice colorful vocabulary.  Run the fan often to check the alignment.

Create a second bar with a bearing and insulator assembly.  This will be used for the top.

Next, we need to bind the cross brace with the bearing to the fan because the side walls will flex greatly.  Cut a strip of aluminum stock and bolt it to both the cross brace and one of the fan frame rails the fan itself is bolted onto.  In practice, this takes some cutting and bending of the new brace.  When drilling mounting holes, keep them tight to the bolts as any slippage here will shake the display.  When tightening the mounting bolts, take care not to bend the frame rails holding the motor.  Check your bearing alignments after this.








Cut 2 rods of square PVC or nylon.  They should be the width of the Peggy2.  These will bolt with spacers to the Peggy2 and to the shafts.  Countersink a hole halfway through each for the shaft, giving a very snug fit.  Put a complete hole through the bars as well so the threaded hole on each shaft can be used to bolt the shaft to the bar.

Cut a plate of plexiglass the same size as the Peggy2. This will be bolted in front of the Peggy2.  It serves several important purposes: counter balance to the Peggy2, safety shield in case anything comes off the Peggy2, and structural support so the rotation doesn't bend the Peggy2.  Leave a notch open over the programming pins, unless you're really, really good.  I found also that I kept nudging the small capacitor near the programming pins, so I hot glued the capacitor in place.

Mount the plexiglass, bars, shafts, and Peggy2LE together.  Key to this arrangement is the fact that the Peggy2 is offset from the center and actually faces inward when mounted yet the LEDs are close to the axis.  This means that while spinning, parts are actually forced back into their sockets.  You'll need a couple more mounting holes on each bar to make this all work and some nylon spacers to get the bottom bar above the electronics.  Pictures are below, showing balancing weights.












Finish building the surrounding frame and supports, using the top bar and bearing assembly.  We put our supports all on one side so there would be a "front" easily visible for friends at the Maker Faire.



Mount 2 magnets on the cross bar just below the Peggy2 as it rotates.  Mount your Hall Effect sensor so it passes over each.  It isn't really important which one is which as long as two different polarities are presented to the sensor.  You may notice these magnets are just labeled "A" and "B".  Also, we double-stacked these small magnets to get them as close to the sensor as possible for the cleanest signal.  It looked like a good idea during debugging but didn't really make a difference since the magnets are strong enough singly.

Get power to the board. We used simple twisted wire as brushes, one pole to each of the 2 axles.  (Now you see why we insulated the bearings from the mounting rails?)  It isn't pretty but it works.  The Peggy2 has that very large capacitor so even if the power gets jittery, the board runs well.










Parts, Roughly


A box fan.  Get one with metal walls for stiffness, variable speed for tuning, and large motor for good torque.

A Peggy 2LE kit  We used the blue LEDs.  I think the red ones have a larger viewing angle and are brighter, but blue aesthetics beat the geometry argument.

2 power supplies for the Peggy2LE.  Why 2? One for your bench.  One to cut up to feed power to the mounted board.

Time.  Seriously.  There are 625 LEDs to solder onto the Peggy.  Time.  Seriously.

2 threaded shafts.  We used these








2 miniature bearings.  We used these.









A shaft coupler and related hardware.  We use this system because it comes apart easily during testing and tolerates misalignment well.  (Did we mention the great opportunities here for practicing your colorful vocabulary while aligning everything?)  Be sure to get both sides of the coupling and the disc that binds them.




A Hall Effect sensor, latching.  Get a sensor that latches on when it crosses a particular magnetic pole then latches off when it crosses the other pole.  By latching, we ensure that no matter what our code may be doing, it will eventually get to notice the magnets that tell us where we are.  You'll need a pull up resistor too.

2 small, strong magnets and mounting hardware.  Not wanting any form of eddy currents, I mounted these using hot glue and corks.  It was extra convenient that the corks were just the right height.

Nylon or PVC square rod.  2 pieces, each the width of the Peggy2LE are used, so get about 2 foot to give yourself some practice stock.  Be gentle when drilling into this as the drill bits grab quickly and these need some countersunk holes.

Plexiglass, thick, the size of the Peggy2.

A variety of nylon bolts, spacers, and washers.  These are used because they are non-conductive.

A variety of bolts, locking/star washers, and nuts.  We used a lot of #6 bolts - small enough to make drilling through metal reasonable, but large enough to handle the mechanical load.

Small balancing weights. If you add these afterwords, as I did, I highly recommend nylon reinforced strapping tape to ensure they can't fly off.  After all, this thing is spinning.

Programming

The 3D cube seen in the demos required a bit of calculation.  We wrote a program using Processing to do the LED/cube/slice intersections.  It's been posted to OpenProcessing.org here.

The Peggy2 is really an Arduino at heart and is programmed in C.  The source code is below in the attachments (MakerFaireShow.pde).  Be warned, I'm changing this right until show time.

ċ
MakerFaireShow.pde
(17k)
Wesley Faler,
Aug 1, 2010, 4:11 PM
Comments