It Breathes

I've continued work on the Deep Listening Device, which is now able to breathe with a set of bellows that I've constructed out of millboard and a wind-resistant fleece.


I constructed a motor-driven reel at the bottom of the device to pull the ropes through a pulley system which raises the bellows.


The motor currently installed in the device is not powerful enough to reel up the bellows, so I will need to find another suitable motor. In the meantime, the pulley system works well when the rope is pulled by hand.

I installed one solenoid to pull one of the accordion valves open to show how it can be controlled. The solenoid plunger is attached to the valve through a throttle cable, which seems to work well.





I have the solenoid attached to a BEAM head robot circuit, which activates the solenoid when the circuit reads a difference in light levels between its two light sensors.



Here is a video of the device in action:


The squealing heard in the video is caused by one of the support wheels. It adds nicely to the evil/creepy atmosphere of the device.

More Building

I've been busy working out a few things with the accordion and building a frame for the bellows. I am building a larger set of bellows to ensure that the accordion gets enough air for it to sound.

I experimented with a throttle-type cable to connect the solenoids to the accordion's valve levers. This will allow for easier placement of the solenoids.



I built a wood interface that will let me build onto the accordion's treble casing, using the same pin-and-socket joint that is used to attach the existing bellows to the accordion.



I've started building a frame around the accordion interface that will support the accordion and expand along with the bellows using a system of pulleys. If it works, it will allow the whole device to grow in size as the bellows breathe in, and sink with gravity as it breathes out. I still need to add parts to stabilize the bottom of the frame and attach the motor that will drive the bellows. Some small wheels will be used to give stability to the top of the bellow mechanism.



I'm also thinking about the sensors that my Evil Deep Listening Accordion will use to sense its environment and inform what notes it will play. The sensors may end up becoming independent little creatures that "sense" evil or appear to be evil/suspicious. They could send signals back to the accordion, telling it of approaching evil and/or other changes in its environment.

Building and Testing

I constructed a set of canvas bellows that seem to work well. Millboard was cut and glued to the inside of the bellows to provide a structure.





Here is a video of the bellows in action.



I still need to work out a way of sealing the ends of the bellows and attaching a tube to carry air to the accordion. I hope to build a few sets of bellows that will operate in concert to provide air to the accordion. Having several seperate bellows will provide a way of visualizing what the accordion is doing - giving it a strange set of lungs that show how the Deep Listening Device is reacting to what it hears.

I plan to use stepper motors to operate the bellows because they can be precisely controlled. I started working with the Arduino microcontroller, using it to drive a stepper motor. The stepper motor is connected through a Darlington transistor array to provide enough current to power the motor.



Here are some sketches of what may become the Deep Listening Device.

A Very Flammable Lampshade

I set out to build a test set of bellows today. I found a website that describes how to build a set of bellows for a camera. I modified the dimensions to produce something close to a pyramidal shape, and constructed a test model using Bristol paper for the ribs and vellum for the outer material.



I believe that the difference in dimensions for each end of the bellows that I chose is too extreme for this method. Going through the graphic calculations, the design I attempted required 7mm and 3mm ribs. The 3mm ribs are too tiny for the bellows to function properly.



It does however make a nice, extremely flammable lampshade.



The method of building seems sound, so I will attempt this again with end dimensions that are less different from each other.

Music and Harmonics

Here are some panels I put together showing a little bit about the relationship between harmonics and music.



The musical scale we use today is not as natural as we might think it is. Our scale, known as 12 Tone Equal Temperament, is slightly out of tune when compared to the natural harmonic scale (known as Just Tuning or Just Intonation). This website has a nice example showing the difference between the two tuning scales.

The device I hope to create is informed by Pauline Oliveros' concept of "deep listening." The device, based on my accordion, will listen to the sounds in the room and attempt to "sing along" with them. Sensors in the room will determine what tones are produced by the accordion in response to what it hears. This device would produce a music of observation - a music formed by watching and listening to what is happening in the area. Some behaviours might be built into the device playing on the concept of the "evil" accordion. It may become nervous and angry when it senses many people near it, producing music filled with dissonance and minor chords. When the deep listening device thinks it is alone, it might produce beautiful harmonies with what it hears. A light sensor placed in the window could be used to detect the weather and time of day, which would further inform the accordion about what notes it should play.

There will be quite a bit of work involved in automating the accordion. I will need to find at least 24 solenoids to operate the white keys of the accordion (plus 17 more if I am able to operate the black keys as well). I will also need to build a new set of bellows that could be operated by a motor.

The Accordion is Evil

I've been doing quite a bit of research for the past week, and am in the middle of compiling it now. I've discovered quite a few references in "pop culture" that associate the accordion with hell and/or evil. The "Devil's Dictionary" defines accordion as "an instrument in harmony with the sentiments of an assassin." I've put together a collage linking some of these references together, with hell being inside of the accordion.



I've also been studying the work of Pauline Oliveros, a composer/performer who experiments with electronic and improvised music using the accordion. She developed a concept called "deep listening" - that is focusing and listening to everything that one could possible hear at a moment in time. I've created a couple of panels that summarize some of her work and ideas.





I'm also researching harmonics, the musical scale and basic chord theory. Drawing from all of this research, I propose to build a "deep listening device" that would listen to sounds in the local environment and attempt to sing along to the sounds in harmony or dissonance using a "mechanized" accordion. More details to come...

3-D Sketching

I came across this link today: Sketching in Thin Air

Using video cameras and motion capture, the Front design group is able to "sketch" in three dimensions. The captured sketches can be fed into a digital prototyping system to produce a physical object.

To Montreal and Back

We had a great trip to Montreal. RPI brought one huge pneumatic wall to which we interfaced our electronics. The installation became a whale-like creature, which breathed with the help of shop-vacs attached to each side of the wall.



A light sensor controlled the rate of breathing (this was the "breathing eye" of the creature). The whole structure would curve and lean to one side depending on which side of the wall was being inflated. Contacts placed between the air cells of the wall activated my electric typewriter and Chelsea's "Let's Pretend Elmo" doll. As the contacts were activated, the typewriter would record of the movement of the wall as keystrokes. A camera pointed at the typewriter sent video to a projector that projected the creature's writing onto the installation space and triggered its "breathing eye". Dana's magnetic tape head was attached to the typewriter and moved along a strip of magnetic tape as the carraige moved back and forth.



Elmo was embedded into the end cell of the structure and would spontaneously cry out and wiggle as its contacts were activated. Kyle's optical theremin was attached to Elmo, causing its sound output to vary in reponse to Elmo's movement and the movement of people in the space.



Seen as a whole, the installation could be seen as a strange beached whale, desperately calling for help using the typewriter, but not knowing the right keys to press. The creature generated cryptic whale songs using the magnetic tape head and optical theremin, while a partially digested Elmo wriggled inside.

The Montreal Proposal

As an FYI for those who don't know, we're going to Montreal to meet up with another architecture school (RPI from Troy, NY). We will combine our work with RPI's pneumatic structures to create some theatrical installation. Our studio has been broken into groups to work on developing a response to RPI's structures.

I am working with Chelsea, Dana and Kyle to develop a response to the work of Erica, China, Barbara and Matt at RPI. They are proposing to build 3 darted/pleated wall sections. Each wall has two air panels. The wall would curve and "stand up" as one of the air panels is inflated. Deflating that panel and inflating the other one would cause the wall to curve the opposite way.

I propose using the electric typewriter as an output device that would record the motions of the whole wall system. The keyboard of the typewriter would be interfaced with switches placed throughout and around the walls, so that as the walls move, the switches are tripped and send keystrokes to the typewriter. The typewriter would type out the actions of the whole system (in gibberish). The typewriter could also take sensor input from Chelsea's Elmo doll to record its actions as well.

Possibilities

I worked out a couple of possibilities for turning the electric typewriter and cheque imprinter into devices that can interact with RPI's pneumatic structures.

The first is the "Equilateral Typewriter." This mechanism would provide an articulation in a pneumatic panel surface which changes over time and in response to people walking around the panel.



This mechanism is designed to interact with RPI's equilateral triangle pneumatic panel, which has the possibility of configuring itself into different hexagon patterns depending on which air cell is inflated first. A touch sensor, possibly triggered by someone pushing aside a plastic curtain, would partially deflate the panel by opening a valve. The typewriter's keyboard contacts would be placed into floor mats around the panel. A paddle or some other device that could constrict an air cell would be attached to a stepper motor (activated by the floor mat sensors). The paddle would be used to constrict one of two cells, depending on which floor mat sensor was activated. It is hoped that this constriction control would allow the panel to alternate between two different hexagon patterns. The panel would inflate again after some period of time, into the hexagon pattern determined by the action of someone walking over the floor mat sensors. The mechanism could be modified to interact with other types of pneumatic structures as well.

The "Paymaster Enigma" would be used to scramble sensor inputs and redirect them to various mechanisms. This device could take input from any sensor in the entire installation and redirect it to one or more devices. This could add to the theatrical quality of the installation, since the outcome of a sensor interaction would not be easily predictable.


The mechanism is inspired by the Enigma code machine used by Germany to encrypt communications during World War II. The numerical levers of the cheque imprinter would be interfaced with switches or photocells. A circuit would be used to scramble the sensor inputs based on the positions of the numerical levers. The crank of the cheque imprinter would be used to reset the levers.

A variable resistor could also be mechanically attached one of the numerical levers. This could be used by a circuit to control the intervals between when certain pneumatic structures are deflated or inflated. For example, when a lever is at its lowest position, a structure could be set to deflate every 20 minutes. With the lever at its highest position, it may deflate every 20 seconds.

My accordion is here at the bus depot, waiting to be picked up. I will begin to experiment with it tomorrow. I'm hoping that tubes could be attached to the accordion to capture the air moving in and out of the instrument. The air could be redirected to a small air cell that could move a pneumatic door or panel. It might be possible to hook up tubes to many parts of the installation, so that the accordion could be used to "play" the architecture.

Electric Type

Here's a video of the naked electric typewriter in operation:



I've documented the "engineering" of the cheque imprinter and the electric typewriter. Click the images below to download the entire documents.




The accordion won't be shipped out until Wednesday or Thursday. Here's a picture of it in the meantime.


Air being pushed out of the accordion bellows could be used to inflate a pneumatic device. It may be possible to use the accordion keys to "fine tune" the flow of air.

Naked Machines

I continued the vivisection process with the cheque imprinter and the electric typewriter.


The inside of the electric typewriter is better than I hoped. It contains three stepper motors, a touch sensor, several switches, gears, a nice contact keyboard and possibly a thermistor (think heat sensor).

Vivisecting Junk

I've been doing some research into the predecessors of my electro-mechanical junk. Most of it has led me to the inventions of the typewriter, adding machines, calculating machines and analog computers. Much of that technology shares similar mechanisms, such as the keyboard which is present in the typewriter and also in adding machines.

I started to take apart the Paymaster X-900 cheque imprinter. The mechanism inside has some very complex movements. Here is a video showing the internals of the imprinter during operation:



Good news! I was able to procure an accordion today. It is on its way from Saskatchewan.

Precious Junk

We were to find 4 electronic or mechanical devices that we will tear apart and integrate into our work. Here are pictures of three of my items:

Paymaster X-900 Cheque Imprinter

Brother EP5 Electric Typewriter

Sawyer's 550A Slide Projector

Not sure if I will use the slide projector or not. It does have a nice rack and pinion for focusing the lens.

I'm on the lookout for a cheap accordion for the project as well.

The Machine Continues. . .

I've been scouring through the thrift stores the past couple days looking for electronic/mechanical devices that we will be integrating into our projects. I picked up an old computer, an electric typewriter, a slide projector and a Paymaster X-900 cheque imprinter. The cheque imprinter is a nice piece of equipment - made of solid metal and has a great hand crank that powers it. Pictures will be posted soon.

Continued work on the "Actuator Theatre" today. I finished constructing the pulleys and designed crankshafts to power the Artoblevski machines.


Pulling the braided fishing line attached to the pulleys causes the whole contraption to spring to life.



Here's a video of it in action:

Drawing Machine and Beyond

We had some preliminary reviews of our paper machine/drawings in studio. We are now allowed to use bass wood in our designs if needed as we elaborate on our work. I think it is more interesting to continue working just with the paper, so I'm going to avoid using wood for now.

Today I worked on improving my curve-generating machine. I re-layered the parts a bit to improve their function. I rebuilt the 90 degree slider with a larger "foot" so that it wobbles less.

I hope to make the machines operate simultaneously and with a single input of motion. I'd like the drawing mechanism to work autonomously, so I experimented with a few different types of pencils and pens to see how well they draw without using pressure. Pencil hardness's down to 2B didn't mark the paper much at all. Pens were not much better. Conte sticks seem to work well enough that the line becomes visible, so I modified the machine to make it able to hold the Conte stick.


I'm now building some larger pulleys that will be wound with string. Pulling the string will turn the pulleys, activating crankshafts attached to the machines. Or so I hope.

A Slightly Larger Paper Machine

Here is the "Hain Linkwork Tilting Crawler" mechanism I constructed yesterday. It works fairly well. The pulleys were constructed from a strip of paper with tabs that is fitted between two paper discs. The rest of the links were cut from three layers of paper laminated together.

A Tiny Paper Machine

Here's a picture of the test gears from yesterday.


This morning I built paper versions of the gears that seem to work well. Folded tabs on the teeth of the gears help them to make better contact with each other.


I continued drawing out and cutting parts for the drawing machine. The parts ended up being really tiny. I would not want to go any smaller than this. The pin joints work well. I still need to revise the 90 degree sliding joint and the "T" joint to make them operate smoother.


The mechanism works as is, but it's a bit "fiddly." It draws only one portion of the curve. I think I would need to build more of the same mechanism in different configurations to complete the curve.

Studio Begins!

All set for another year chock-full of studio goodness!



Project One for Patrick Harrop's "Hacking Studio" begins with a drawing machine drawn and built out of paper. Two machines from Illich Artoblevski’s Mechanisms in Modern Engineering Design will be integrated into the project. I selected one mechanism that generates a fourth-degree curve and a "Hain Linkwork Tilting Crawler."

Today was spent figuring out how the fourth-degree curve generator might work and designing the gears that it requires. Using simple circle geometry, I constructed a set of gears with three elliptical lobes. They almost worked (and probably would work well if the gears could be constructed absolutely perfectly). The lobes mesh together well, but slight variations in the lobes cause the gears to lose contact with one another.

I constructed another set of more conventional gears that seem to work well (at least in millboard form). I'll try them in paper form tomorrow.