computer-moved levers and fake digital info (long)
Posted: Tue Feb 03, 2004 3:08 pm
Hello all--
I'm an artist/sculptor/teacher learning the basics of robotics and computer-controlled thingys in order to one day make sculptures that move. I have been reading books ("Practical Electronics for Inventors" by Paul Sherz, and "Robot Builder's Bonanza" by Gordon McComb) and think I have just enough understanding to start asking wildly premature questions.<p>(This email is basically asking two question in the same forms: Which electronic components might be used to do such-and-such things? Later, I'll worry about *how*.)<p>1. Is governing a bunch of levers something that an old IBM/PC (pre-1994?) can do through its printer port? Will each lever need its own motor? Electromagnets (called solenoids?) seem involved in elaborate machines. Is this an option for a tight budget and an old computer?<p>[Explanation: I'd like to control any combination of several dozens of mechanical levers (though usually fewer than a dozen simultaneously) on a machine that is a very simple pipe organ. There's no need here to get into organbuilding, but briefly, the keys on an all-mechanical pipe organ are actually levers that connect the organist mechanically with the valve that allows air to flow through that particular pipe, which equals a lotta noise. The travel of the key/lever is in the 1-2 cm range. Organs that are not completely manual have electronically assisted switches that govern the instrument's functions. The functions are basically linear movement stuff: opening valves, sliding a wooden piece a few inches back and forth, and so on. Contemporary organs will also often have the brains to be programmed: "When I flip this master lever, please flip levers 2,3,7,11, and 27." Companies that build this stuff seem all to have dedicated proprietary circuitry and software that is complicated and absurdly out of my reach financially. I'm looking for basic systems that might approximate the same thing. I'm not an organist, but a sculptor who views the organ as a cool kinetic sculpture that makes noise.]<p>2. Can I paint black marks (vertical thin stripes) on a white surface that can be read by a sensor as information? Is this the job of an IR sensor that detects light reflectance? I've read about line-tracking 'bots but my application seems different. If I wanted to use those marks to be encoded instructions for a lever machine (like the one above) and move the levers in real time as info is being read, how should I start thinking about electronics components? For example is there a limit to how quickly the information can sweep past the sensor?<p>[Explanation: I'm fascinated by how information is stored mechanically or digitally, as in audio CD technology and also by the paper punch methods once used to encode info for computers and still used for musical instruments. Note that this is not a project about using existing technology ("Why don't you just burn a CD?"), but re-interpreting the processes as part of a "kinetic sculpture" or "robot" or "weirdo musical instrument". I'd like to have a way of encoding information "digitally" as 1s and 0s --black lines or white background-- to send information to a system of levers in a project like the one above. So far what I understand is that on CDs, the data are bumps that encode binary info in a microscopic spiral, read by software to deal with error correction, compensation for the increased speed of the CD surface as the laser reads bumps closer to the edge, and stuff I still don't understand about processing. In MY simplified version, I'm thinking of a tall white cylinder (PVC pipe or a smooth ceramic surface) with opaque black paint. Off the top of my head I'd say I have to spin the cylinder at a constant rate, mount a sensor on a frame that tracks properly along the surface, and the software to figure out what it all means. As for printing the black marks, the only idea I have is to coat the surface in photosensitive emulsion (with black paint in it) and rig up a tiny UV light to expose (and make permanent) the correct intervals and shapes, which is obviously done by computer. After the unexposed emulsion is washed off, the remaining exposed marks that are now a permanent part of the surface are what the sensor detects as 1s.<p>That's it! If anyone would like to comment on ANY aspect of this creative process, please do so. If you've made it to the end of this email, I apologize if I have come across as a newbie with ideas bigger than his head. I appreciate any input.
Thanks,
Marc
I'm an artist/sculptor/teacher learning the basics of robotics and computer-controlled thingys in order to one day make sculptures that move. I have been reading books ("Practical Electronics for Inventors" by Paul Sherz, and "Robot Builder's Bonanza" by Gordon McComb) and think I have just enough understanding to start asking wildly premature questions.<p>(This email is basically asking two question in the same forms: Which electronic components might be used to do such-and-such things? Later, I'll worry about *how*.)<p>1. Is governing a bunch of levers something that an old IBM/PC (pre-1994?) can do through its printer port? Will each lever need its own motor? Electromagnets (called solenoids?) seem involved in elaborate machines. Is this an option for a tight budget and an old computer?<p>[Explanation: I'd like to control any combination of several dozens of mechanical levers (though usually fewer than a dozen simultaneously) on a machine that is a very simple pipe organ. There's no need here to get into organbuilding, but briefly, the keys on an all-mechanical pipe organ are actually levers that connect the organist mechanically with the valve that allows air to flow through that particular pipe, which equals a lotta noise. The travel of the key/lever is in the 1-2 cm range. Organs that are not completely manual have electronically assisted switches that govern the instrument's functions. The functions are basically linear movement stuff: opening valves, sliding a wooden piece a few inches back and forth, and so on. Contemporary organs will also often have the brains to be programmed: "When I flip this master lever, please flip levers 2,3,7,11, and 27." Companies that build this stuff seem all to have dedicated proprietary circuitry and software that is complicated and absurdly out of my reach financially. I'm looking for basic systems that might approximate the same thing. I'm not an organist, but a sculptor who views the organ as a cool kinetic sculpture that makes noise.]<p>2. Can I paint black marks (vertical thin stripes) on a white surface that can be read by a sensor as information? Is this the job of an IR sensor that detects light reflectance? I've read about line-tracking 'bots but my application seems different. If I wanted to use those marks to be encoded instructions for a lever machine (like the one above) and move the levers in real time as info is being read, how should I start thinking about electronics components? For example is there a limit to how quickly the information can sweep past the sensor?<p>[Explanation: I'm fascinated by how information is stored mechanically or digitally, as in audio CD technology and also by the paper punch methods once used to encode info for computers and still used for musical instruments. Note that this is not a project about using existing technology ("Why don't you just burn a CD?"), but re-interpreting the processes as part of a "kinetic sculpture" or "robot" or "weirdo musical instrument". I'd like to have a way of encoding information "digitally" as 1s and 0s --black lines or white background-- to send information to a system of levers in a project like the one above. So far what I understand is that on CDs, the data are bumps that encode binary info in a microscopic spiral, read by software to deal with error correction, compensation for the increased speed of the CD surface as the laser reads bumps closer to the edge, and stuff I still don't understand about processing. In MY simplified version, I'm thinking of a tall white cylinder (PVC pipe or a smooth ceramic surface) with opaque black paint. Off the top of my head I'd say I have to spin the cylinder at a constant rate, mount a sensor on a frame that tracks properly along the surface, and the software to figure out what it all means. As for printing the black marks, the only idea I have is to coat the surface in photosensitive emulsion (with black paint in it) and rig up a tiny UV light to expose (and make permanent) the correct intervals and shapes, which is obviously done by computer. After the unexposed emulsion is washed off, the remaining exposed marks that are now a permanent part of the surface are what the sensor detects as 1s.<p>That's it! If anyone would like to comment on ANY aspect of this creative process, please do so. If you've made it to the end of this email, I apologize if I have come across as a newbie with ideas bigger than his head. I appreciate any input.
Thanks,
Marc