Source Code. Projects. Nerd Stuff. Art Stuff.

Building the Machine


Posted on October 29th, by amoeboar in classes, pcomp. No Comments

After last week’s play tests, we realized that the users needed more response from our device so that they could know how they were controlling the machine’s output. We added a start/stop function so that people can stop the drawing if it doesn’t look the way they want it to and start it up again after re-recording their voice. We decided on the following workflow:

 

Prototype 1

We wanted to keep as much of the motor setup off of the support arms, so we opted to give our machine a rotating base. This would correspond to the rotation of the large arm on our drawing sketch that was mapped to volume. We looked for a way to create a two-axis drawing mechanism on top of the machine (for freely rotating x- and y- values) so we thought of a pivoting pair of scissors to hold our pen:

We built it out of cardboard and balsa wood to test the mechanism, and before attaching any motors to it, moved it with our hands to see what kinds of shapes it would make.

 

We realized that this design only drew “donuts,” as the scissor mechanism did not allow the pen to reach the center of the paper.

 

Prototype 2

We got rid of the scissors, and substituted it with a sturdy arm meant to hold the motor. However, although we were reaching the center, this iteration only created “flowers,” as there was no second axis of movement for the pen to add a variable dimension. Also, our DC motor wasn’t enough to turn the base, so we opted for a gearhead motor.

 

Prototype 3

Our motors worked well independently. We decided to add a rotating disc (and a second motor) to the drawing arm for the variability we were looking for (we still needed to map values from out mic input to a third drawing axis).

Our next step was to begin construction of the working model. We opted for acrylic because of its sturdiness and ease of fabrication via lasercutting. We ripped rods out of a printer:

 

We are close to finishing our model. What remains is to get the motors working together, and to get the software talking to them.

 





Leave a Reply