Attribute the scare quotes to the fact that that there is a *lot* of electronics that still need to be inserted. But, the “device” (sans microcontroller, power, and little intangibles like “programming”) is living happily on the guitar. It’s exciting. The last few days have been a sometimes tortuous exercise in trying to mill precise little wood blocks. The biggest stumbling ‘block’ (hah) was the main support that keeps the solenoids assembly a certain distance away from the neck. Because the strings still have to move, the clearance has to be substantial. I ended up cutting out the mounting holes into slots, so that it can move back and forth and be adjusted properly. See picture below.
|
||||||
|
Since the laser cutter is down, I had to see if I could figure out how to reuse my existing brackets. Using the drill press, I made a series of small holes to accommodate the solenoid lead wires. Then I carefully pushed the wires through and put the solenoids in the masonite mounting board I had cut previously. A couple of the holes broke. I traced the old laser cut template onto another piece of masonite, and used the drill press to create another supporting board. As expected, being exact here is pretty important, and the holes for the carriage bolts aren’t good enough. Thankfullyu, I can just drill these better and use washers. Sanding Sand paper bends one way better than the other. Used the band saw to cut lines to a circle that I drew with a compass. Used .5″ dowel to make “pads” for the solenoids to strike the strings. Too thick. Since the mounting holes I drilled in them aren’t completely straight, some of them rubbed together. So, I switched to .375″ dowel. Much better. The whole assembly is built, but getting it onto the guitar without messing up the instrument has proved pretty tricky. Turns out there is a metal rod that runs down the shaft of the neck — can’t screw directly through it, then. Cut and sanded wooden blocks to hold the guitar from the back, but I still need to have something on the top of the neck for the assembly to rest on. It needs to be exact, and without the laser cutter, that’s been hard. I made five cuts, and none of them really work. Time to head home for tonight. I’m disappointed that I didn’t finish the physical build today, so that I could do the circuits tomorrow.
This post concerns my progress on my Physical Computing final project– an assistive device that will allow a musician with limited dexterity in their left hand (from, say, accident or illness) to again play the guitar. A series of buttons underneath the fretboard will, with the aid of the Arduino microcontroller, trigger a combination of solenoids to depress the strings to create a chord. ULN2003 – $1. ULN2380 – $1.75. Finding the right transistor array, one that will be sure to be able to supply enough current to all the solenoids I’m going to be using — priceless. I hadn’t thoroughly investigated the amperage requirements of the solenoids I am using until now– and boy, was I in for a surprise. Even though they’re 1/2″ in diameter, and barely an inch long, they like to suck on about 400 milliamps between 9V-12V– that’s a lot of juice, and we’re not talking about tropicana. Damm. See, those transistor arrays I was looking at CAN put out about 500 milli amps — *IF* you’re using just one circuit. By the time you’re running juice to two or three solenoids, that amperage is going to drop by half. And my design calls for the possibility of ALL of the solenoids staying on for, possibly, an extended period of time. Previously, I was going to order up a triad of ULN2003′s, but they don’t meet the power requirements. I looked at another, beefier transistor array that I found through this but I realized I needed to do some solid math. 400 milli amps x 18 solenoids = 7.2 amps. Sheee-it. And I thought I was going to have some cute little battery series on board the guitar fretbox. Ain’t gonna happen. We’re talking something akin to a motorcycle battery. I could be off a little, but it’s the same order of magnitude. I’ll probably look for some kind of computer power supply, but either way, this is going to be tethered instrument. The other issue, of course, is heat. With this consideration in mind, I think I’m just going to use 18 individual transistors. It’ll be easier for the heat to dissipate, I think. And I won’t have the power problems that plague transistor arrays when they have to fire multiple circuits concurrently. Hah– con’currently.’ Continuing to work on my guitar “fretbox” assistive device. Used my cheap digital calipers ($15 on Amazon) to measure the distance between the guitar frets and strings, drew up a quick guide in illustrator, and then drew circles over the strings whose diameter corresponded to the diameter of the solenoids I’ll be using. I was pretty excited (see smile above). But what I didn’t take into account is how much extra wood the laser cutter would eat up. The solenoids don’t exactly fit snuggly. I measured the interior diameter of the holes, and subtracted the difference between what I told the laser to cut and what actually was cut. I used this factor to redraw the circles… and now, they’re too small. This means that this relationship is probably non-linear. I’m so close though, that I think I can just measure the difference, and make a fudgy “close enough” calculation. I bet I can get it pretty darn good next cut. Tomorrow! Found a solenoid that is strong enough to depress a guitar string– and at 1/2 an inch diameter, it seems possible that I’l be able to find a layout that could accommodate six of them, side by side, in a vertical row on the fret board.
Nick Yulman suggested that I stagger the solenoids– it’s a great idea, and it allows me to use lager (and more powerful) component than I otherwise would be able fit in. Turn-of-the-century player pianos, he mentioned, used a staggered similar layout to fit a number of plungers side-by-side. Next step: create a model on paper (or maybe Google Sketchup!) Excited that my hunch about solenoids seems to be paying off. The definition of “art” came up in class today, and I was reminded of the the educational goals of the American Visionary Art Museum. While they probably don’t serve as a totally inclusive definition of art (I doubt anything does) or encompass all of our intentions in designing interactivity, I see a mirror for many of my aspirations– and yours. It’s worth a look: 1. Expand the definition of a worthwhile life. Thanks to Jason Eppink for pointing me to this list. Following Owen’s great overview of our progress so far, I was inspired to try my hand at keeping some kind of running documentation of our next steps. If you haven’t read his post, go [there] first. OK. In order to make the clicking sounds, Owen created a very clever and efficient bit of code, using the Random() function in a For() loop to vary the number of times a simple Tone function would be quickly implimented: for (int i = 0; i < random(0,10); i++) { Simple and smart. Linking an additional behavior to the clicks (and by "behavior", I mean a change in our organism's lighting, its movement, its temperature, etc) such as doing a PMW write to an LED strip, however, is tied to the repetition of the For() loop. Flashing the lights in exactly the same manner as the clicks just doesn't look very organic. The oscillation of different behaviors needs to be related, but not congruent. Enter the millis() function. We'll need to create a couple of counters, probably one for each of the behaviors we want to implement. I'll start off by creating a timer class, so that we can call different timer objects. Rune, a 2011 ITP resident, helped me with this kind of code yesterday. [insert processing code] Well, *that* didn't work. Turns out the way that the Arduino language handles objects is a far cry from the way processing works 'em in. (Probably an overstatement). But, we'll try to just get millis functionality in Arduino without getting classy. Err, complicated. Meaning, no classes. For now. void loop() { duration = random(100,1000); if(millis() - lastMillis > duration) Weird. This code appears an even clicking, like a metronome. But after giving the random function a broader range– duration = random(5,1000); – a nice, syncopated clicking emerges. Great. First, a quick visual summary of the “raised fist” of solidarity and resistance: https://docs.google.com/present/edit?id=0AVvVpJApPj-LZGM0cTJjcWRfNjdmNWMyYjRjNw&hl=en_US Looking at the #occupy movement, you’ll see that it’s not particularly “branded” — it’s expressed. What I mean by that is that the imagery that represents the movement is imagery that individuals within the movement have chosen at particularly times to express particular Ideas and feelings. As much as the movement is about “we” it’s really about preserving the worth of individual expression. In this way, banding, or logo design, could conceivably be understood as antithetical to #occupy. I think the issue is more subtle. What’s important, I think many folks will agree, is about using consciously the powerful semiotic tools we see codified in good branding. Logos can say a lot. Sometimes they can say little, and prevent a lot from being said. It’s this particular spectar of homogenization that rightfully concerns many protectors and their allies— any brand that aims to unite us all will surely leave many of us out, and render diverse perspectives unrepresented. I offer the following image in tribute, in admiration, not in summary.
Nothing more, nothing less. The only thing that catches me is the idea of representing myself as a ‘digital anything’ on simple paper. There has to be some simple medium that isn’t too showy, but is also playful and preserves some of the playful mystery that the card suggests. For comparison, you can view a recent card below, a lucky collaberation with Lindsay Eller and Justin Aubuchon for which I’m very grateful! |
||||||
|
Copyright © 2013 Justin Lange - All Rights Reserved |
||||||
Recent Comments