Andrew Burks

Autonomous Harmonica-Playing Robot

I have wanted to add a diatonic harmonica to the RobOrchestra since I joined the group in my freshmen year.  The Carnegie Mellon Robotics Club recently acquired a MakerBot 3D printer, so my dream has become a lot more feasible than before.

Harmonica Interface and Control

The concept for the robot is simple: A 3D printed plug (with complex inner geometry) routes the 10 square holes of a diatonic harmonica to 10 NPT fittings.  The NPT fittings connect to 10 different solenoid valves, each corresponding to one hole on the harmonica.  This allows for individual control of the air going through each hole.

Achieving both “Blow” and “Draw” Notes

The solenoid valves all connect to a single manifold, which is connected to two other solenoid valves.  One of the two valves is connected to positive pressure, and the other is connected to negative pressure (vacuum).  Activating one of the two solenoid valves at a time can simulate a blowing or drawing, while the other 10 valves select any number of holes on the harmonica to play.  Unfortunately, like humans, this robot will not be able to blow in some holes while drawing from others.

Harmonicas of Different Keys

The best part about having one generic plug to route the 10 holes to 10 NPT fittings is the modularity it provides.  Diatonic harmonicas come in all sorts of keys, but they all have the same shape.  Because of this property, harmonicas of different keys can be easily switched in and out of the mechanism.

Funding

The estimated cost of this robot is at $500.  This project is applying for RoboClub project funding, but will not be applying for a SURG.  The largest cost is the 12 solenoid valves at $20 each.  More updates and design will follow if the project gets funding from somewhere.

The Problem

After several iterations of the wheel-style ball dispensing approach, I decided to try a gating mechanism instead.  The main drawback of the wheel approach was the low throughput.  Only about 4 balls per second could be dispensed.  Xylobot, however, can play over 20 notes per second.  While Xylobot’s speed isn’t fully necessary, at least 8 balls per second is a reasonable goal.  The wheel mechanisms do not meet that goal.

The Solution

The basic idea of a single gate mechanism is that the stream of balls is free to fall out the end of its containing tube whenever the gate is open.  The trick is opening the gate for a short enough period of time to consistently allow only one ball to pass through the gate per activation.  In our prototype, the time required to let one ball through was about 40ms.  With delays to account for the return motion of the gate and the settling of the balls, the prototype gate mechanism could achieve a rate of 11 balls per second.

Design

The most important part of this design is constraining the balls to one dimensional travel.  This was achieved by basing the gate around an aluminum tube with an inner diameter of .5”.  The gate would be entering perpendicular to the direction of travel, through a hole in the aluminum tube.  When the gate is up the balls roll through, but when the gate is down the balls are unable to pass through the gate.

The gate itself is a cone of plastic.  The conical shape is important because any non-angled surface has the potential to jam by clamping on the top of the ball.

The plastic cone is attached to a pull-type solenoid with a spring return.  The built in mounting bracket on the solenoid is a great advantage, but the 3-4 amp power draw (at 12 volts) is a major disadvantage.  The gate is normally closed, but when powered the solenoid pulls the cone out of the way of the balls.

In the initial prototype the mechanism made a large amount of excess noise when releasing a ball.  Quiet actuation is important for a musical instrument because the sound of the ball hitting the vibraphone key needs to not be overwhelmed by any other noise.  The addition of an o-ring to act as a hard stop when the gate returns to the closed position should help to muffle the noise produced by the prototype.

The final integration of this mechanism (actually 30 of these mechanisms, one per key) will have the tube nearly vertical, with the gate at the bottom.  With balls feeding into the top of the tube from Mike’s hopper mechanism, the gate will dispense one ball at a time, allowing them to fall onto the key and play a note.

Renders

Side View

Isometric Render

Isometric Render with Transparent Tube

Design

This summer I have had access to the waterjet at NREC where I had my internship.  The current pedal design calls for a series of 2-dimensional parts that I knew I could cut on the waterjet very easily (it can cut through stuff faster than 10 inches/minute, so it’s preferable to a CNC for this application).  Unfortunately I put off cutting these parts until the last week of my internship.  At 10PM on the last Tuesday night of the summer, I realized that I needed to prepare the .dxf files (files that define the shapes of a 2-dimensional part) I would need to cut some parts out after work.

When I went to make the .dxf files, I realized that I could make the entire pedal assembly out of welded steel plates cut from the waterjet.  That night I started making a new version of the pedals (version 9) and had the rough outlines of all of the key parts by the time I went to bed at 3AM.  I also never made the .dxf files I wanted for the next day.

Wednesday night I made the .dxf files I needed for the old pedal design, then spent the rest of the night optimizing the brake pedal assembly for weight.  Here are some renders of the design, as well as some FEA results:

Stress Plot

Factor of Safety Plot

An Exploded View Animation of the Assembly Process

Fabrication

On Thursday night I stayed after work to cut my parts on the waterjet off-hours.  It took about 5 hours to cut three sets of parts for the old design and two sets of the experimental pedal.  Though the brake pedal hasn’t been welded yet, the tabs were easy to press together to get a good sense of what it will look like in real life.  I was really impressed that I could go from concept to completion in only 48 hours, thanks to SolidWorks, NREC, and the waterjet.  Mike was kind enough to take some glamour shots of the final product:

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.