Andrew Burks

General Concept

With the current direction for the Vibraphone design, the notes are played by dropping a steel ball onto the keys.  There are several ways to actuate the balls, but considering the cycle times we want to achieve and the cost of each device (since we need 30 total) using a small motor and a wheel to push balls off of a queue and into free-fall seemed like the best idea.

The club has a few sets of tiny motors that have been donated, and each set has as least 30 motors of that kind.  The two sets we investigated are the tiny DC motors from the handheld fans, and some tiny bipolar stepper motors.

There are two trains of though for what to put at the end of the motor.  One possibility is to put a circle with a squishy perimeter on the motor, and use friction to pull balls through the mechanism one at a time.  Another concept is to cut ball-sized notches into the perimeter of a plastic circle, acting like a sprocket on a row of queued balls.

We have tested both concepts on the fan motor, and they each have their advantages and disadvantages.  After testing them both on the stepper motor in a more controlled way, we should have a better sense for which type of wheel will work best for us.

Here is a photo of the notched wheel Mike Ornstein and I milled in the roboclub CNC mill the other night.  The notches fit the balls great, but a consistent problem we had with the fan motor was that balls would jam if they tried to fill an empty queue.

Fan Motor

The fan motor is easy to control.  If you put a voltage difference between the two wires, it will spin.  If you give it a low voltage (1V) it will spin fast.  If you give it a higher voltage (3V) it will spin VERY fast.

From my observations, there just didn’t seem to be enough torque on the motor to handle the balls we were giving it.  Also, when testing with the notched wheel, the fan seemed to lack all braking ability.  Obviously with a friction wheel instead of a notched wheel the motor will be able to resist back pressure, so I look forward to seeing how that performs once we get a nicer friction wheel made (Plastic circle with a notch for an O-ring).

Stepper Motor

The club has a box of 150 tiny stepper motors with 4 wires coming out of each of them.  Starting the process knowing absolutely nothing about steppers, I eventually determined that our stepers were Bipolar Stepper Motors.  Basically, there are two pairs of wires and I need to follow a cycle of powering and releasing the pairs in different directions in a certain order so I can get the motor to ’step’ 1/48th of a rotation.

By using an H-Bridge configuration on each pair of wires, I could independently control the direction of the current in the wires with digital logic.  20 lines of C++ and an Arduino later, I had a great test rid that let me step the motor at whatever speed I wanted!

I was very pleased with the initial performance of the stepper motor.  There seemed to be a lot of torque behind the motor, despite its size.  The biggest advantage in my eyes though was the powered braking.  By setting only one set of wires in only one direction and leaving it there, the motor was in a powered lock.  This should help with the back pressure issue we were facing on the fan with the notched wheel.

Here are some photos of my final Arduino/H-bridge setup.  I was very happy because both the circuit and the program worked on the first try!  That never happens!

A completely axially symmetric vibraphone robot would be awesome.  We decided to move away from a big row of keys and towards a round plate of keys.  Here is a quick render of the key mounting structure and how it incorporates the ball retrieval and distribution system:

Structure

The large round plate is actually a 30-gon not a circle.  It is inscribed in a 32″ circle, and is 1/4″ thick.  There are 60 unique (thank you design tables!) plastic supports that slide onto notches in the aluminum.  Each plastic support has to be unique because of the awkward hole spacing in the individual keys.

There are already notches in the plastic for clips that should hold it into the aluminum plate (aka “Megaplate”).  However, depending on the design of the ball deployment mechanism, the retaining clips for the plastic plates should be incorporated into the support for the mechanism.  Here is a close up of the plastic supports:

Distribution

Finally, here is a close up of Mike Ornstein’s ball collection and sorting mechanism.  It uses brushes from the bottom of doors to pull balls up an archimedes screw into a paintball-style hopper.

Overview: We are officially crazy

Animusic is a group that makes great computer animations involving “impossible” instruments playing great music.  While considering actuation mechanisms for the RobOrchestra Vibraphone project, somehow we decided it would be a good idea to do something similar to the instrument that takes center stage at 1:07 in Animusic’s “Pipe Dream”:

Details

Right now, 3/8″ diameter stainless steel balls are looking very promising.  Mike Ornstein, Dan Shope and I have subconsciously split up the work into 3 sections.  Dan is working on the mechanism to take the balls and dispense them onto the keys quickly and with a short reload time.  Right now, it appears that this will be accomplished with a group of DC motors.  Mike is working on the mechanism to lift the used balls back up and dispense them to queues leading into Dan’s mechanism.  This is most likely going to be done with an Archimedes screw and a paintball gun style dispenser.  I have been focusing on the structure of the whole mechanism and collecting the dispensed balls and funneling them to Mike’s mechanism.

The biggest problem I am facing with this design is the awkward hole arrangement in the keys.  I basically have two very awkward hole lines I need to support for both the naturals and the sharps.  A string pulled taut needs to go through the holes in the keys and the supports to hold up the key and let it vibrate naturally.  My initial concept involved about $60 of waterjet-cut 1/8″ ABS.  Here is a render of this initial design:

This concept was that with angled plates in front of the keys sloping back toward the keys, as well as slopes over top of those angled toward the center, I could funnel all of the ball bearings into a channel between the two sets of keys.  Unfortunately it takes up a whole sheet of plastic.

Future Concepts

Moving forward, I want to find a way to eliminate all of the unnecessary material in all 32 of those vertical supports.  A bar or two mounted along the path of the key mounts could allow me to build much smaller plastic mounts for each key.  Look forward to another post with more designs, and watch my friend’s blogs for updates on their portions of the project!

My new workstation is now humming along perfectly.  To review, the specs are:

• Intel i7 920 Quad-core @2.66GHz (currently not overclocked)
• ASRock X58 Motherboard
• Nvidia FX Quadro 580 workstation graphics card
• 6GB RAM
• 1TB Samsung HDD
• 650 Watt Corsair Power Supply
• CoolerMaster Hyper 212 CPU cooler
• CoolerMaster 335 Case

Build Observations

Putting together a computer was quick and easy.  Everything went super smoothly and was very straightforward.  I never really had to read the directions (although Mike Ornstein was guiding me heavily).  The hardest part was making sure all of my components were designed to work together.  I spent more time researching individual parts than I did assembling the whole system.

Overclocking seems unnecessary right now.  I have Solidworks running in RealView, and I can spin large models with no lag.  The biggest improvement is my rendering ability.  1920×1080 renders of complex geometries in Photoview used to take more than an hour, or just crash my laptop.  On the new computer, it takes only four and a half minutes.  This makes sense, considering that my Windows 7 Experience index raised from a 3.1 (Limited by graphics score) to a 6.9 (Limited by HDD score; Graphics and processor are highest – tie at 7.9).  I successfully played an HD movie over the weekend, and am very pleased with the results.

Software

With one computer that i use in my room (or remote into) and another I take around with me, it is important to make sure they play nice.

I found a file synchronization tool, FreeFileSync, that I really like.  I have a “SYNC” folder on each computer, with everything I want to have available to me on any computer (School/Project/Personal files, Music, etc).  FreeFileSync matches this folder from each computer against a backup I set on my external hard drive.  So I have three sets of identical data in three separate places.

I was having trouble syncing my iTunes playlists.  The way iTunes handles its playlists and libraries is very weird.  I decided to convert to Windows Media Player, and I’ve never been happier.  All of my music (4-5GB) and playlists are synchronized now.  However, because WMP playlists are just xml data with absolute mp3 file locations, syncing the playlists would make it not work on one computer.  I solved this problem by drawing on my 15-123 PERL skillz and writing a quick script to convert the absolute file names into relative file names in the playlists (which are identical between computers) and now my playlists can synchronize too!

I have set up remote desktop and have given my friends an account so they can render on my machine.  I’m curious to see if i start remoting into my workstation when I’m on campus, or if I’ll just keep using my laptop most of the time.

I look forward to installing Synergy which should allow me to control the workstation with the keyboard and mouse on my lenovo (which I love) as well as using my laptop and workstation screens side by side, as if they were one computer.