CONNECT is a kinetic sculpture which produces chaotic patterns. There is no main program outside of the sculpture and no digital connection between its elements. Each module has its own simple program logic and they just start to react to each other because of the physical connections built.
I am still working on the documentation of this project (Code, PCBs). Nevertheless you can have a view on the already existing images and video here:
That's pretty trippy. Is there any signal at all between the modules? Or is it only sensing the change in weight and inertia when the magnet bar connects them?
To be honest there is a one way serial connection between the boards. But I use it just when everything is started to overgive some parameters from a controller board outside of the sculpture to each module. Otherwise I would have to programm each board again when changing some main parameters
I use a piezo sensor and a lowpass-filter between sphere and stepper motor. So the programm on each board can measure the amplitude and the frequency of the swinging sphere. This is just a relative measurement, and in the beginning each sphere has to swing to recognize if the bar is connected (you can see this in the beginning of the video). If the bar is connected, the average value of amplitude is less than if not connected, because the connected bar interrupts a regular swinging. This was one of the hardest parts of the project. I also tried to use the frequency as an indicator. But I didn't get very good results.
Nevertheless the frequency measurement is used to drive the stepper motor up and down at the right speed, so that swinging of the sphere is amplified as good as possible. I did this because I like the idea of feedback mechanisms to regulate a system. The module stops driving the motor when the bar is disconnected. This is also recognized in the average value of the amplitude measurement. Then it waits as long as the bar connects again. This again is also recognized in average amplitude of the swinging. So you see the amplitude is always the indicator for a connection built or not.
Another hard part was to find the right magnets for the bar. They must not be too strong so that the connection cannot break and not to weak, so that the bar is falling to the ground when swinging. I also thought about using solonoids but in the end I like the idea of staying analog in the lower section of the sculpture.
So the bar is responsable for the behaviour of each element. It's the indicator if a sphere has to swing or not and he also carries the impulses from one element to another.
I hope all the information does not confuse too much. At the moment I have too finish another project but after that I want to put the code and the layout for the PCBs online because they are nice stepper motor modules for other projects too.
Wow, that's very impressive! i like it a lot! so you could detect when the magnets would attach to one another? did it complete a circuit between the two? its really neat! any problems with it tangling? =P
There is no circuit completed between two. Each module just measure the oscillation of its sphere and there is difference in the average of the amplitude if the bar is connected. Completing a circuit maybe would be easier, but I wanted to stay analogue in the bottom of the sculpture.
And yes there is sometimes a tangling, but when you leave it after some while it always gets resolved, because of the gravitation
ah, that's really neat! (my dad saw that and claimed he "figured it out," but apparently he was wrong)
Very cool. I like the way the bar may or may not move from sphere to sphere. Sort of a nomad.