Keith's Electronics Blog

Background

Steve Wilson is a graduate student involved in WSU’s CRATEL (Center for Research in Art, Technology, Education, and Learning) program, and his video synthesizer was sponsored for further development in last year’s BETA (Bridging Entrepreneurship, Technology, and the Arts) competition.

One of Steve’s challenges for his video synth, and an interest in general, is creating expressive interfaces to electronic instruments. He describes it in a post to the Technology: Art and Sound by Design class mailing list. Basically, it comes down to the idea that acoustic instruments offer an immense amount of control over the sound that comes out through physical manipulation, and electronic instruments tend to offer much less control, and through mechanisms mainly like knobs and sliders.

While brainstorming with Steve about his synth, several of us came up with the idea of an interface like a plasma globe, which you control by moving your fingers over the surface. A cheap videocamera inside could watch what was happening and control the action.

Apparently a group of engineering students have been working on the globe subproject for six months without much progress to report. Although their difficulties seem to be mainly lack of organization and lack of time, others of us had talked about how clear an image might be obtainable on the camera with the globe lighted from the inside or outside.

Finally it occurred to me (while driving home from work) that gloves with LEDs on the fingertips would make lighted spots that should be very easy to see from the inside of the globe. Different colored LEDs on different fingers would allow the driver software to identify the different fingers; and pressing the LED against a translucent surface will produce a sharper point than holding it further away, so it might even be possible to determine approximately how near a fingertip is to the globe’s surface. And if made properly, the glove could be worn without impeding other use of the hand–just adding lights for the globe.

Time to prototype.

Gloves

I was hoping to find a cheap lycra glove to make something like the original Dataglove. (Yes, I have a Dataglove system. No, you can’t play with it.) What I found instead was certainly cheap, but knit cotton instead of lycra. Heck, for 96¢, it was good enough.

When I put the gloves on, I could feel that they’d been assembled somewhat sloppily. Inside-out, it was easy to see that there was a lot of extra fabric inside the fingers outside the stitching:

So I trimmed it a little closer and more evenly around the fingertips, for better fit and comfort.

LEDs

I wanted to use surface-mount technology (SMT) LEDs, so they’d fit flat against the glove surface. The ones in my stockpile are out of broken digital office phones, used to light up beside the line appearance and feature buttons.

I wanted to be sure I knew the polarity before I started attaching things and soldering them together, so I tested first with my meter. The LEDs actually lit with the small amount of current it supplies, which was kind of cool. Curiously, the green LEDs have the anode on the end with the cut corner, and the red LEDs have the cathode on the end with the cut corner. On each LED that I identified, I marked the anode end with the color of the LED. (The ones on this sheet are tinted to indicate their color, but the ones I’d already removed weren’t.)

The phone flex-PCB has a 300Ω resistor in series with each LED, so I tried out a couple there on the strip. Nice and bright.

Soldering the LEDs

My plan was to solder wire the LEDs, then sew the wire through the glove to the back side, where I wanted to attach a battery holder. I got out some fine wire, scraped off the enamel insulation, wrapped it around the fly leads on the LED, and soldered it up.

Sewing

After I had both wires soldered to the LED, I threaded one at a time through a needle, held the LED about where I thought I wanted it on the fingertip, and started sewing. The goal was to get the wire from the tip of the finger to the back of the hand without leaving any large loops that might snag on anything, and without reducing the stretchiness of the glove. The latter factor called for a zig-zag stitch, which with a single thread looks like a running pattern of slash marks.

Once I got the wire stitched around to the back side, I followed along the back of the finger and then curved toward the center of the back of the hand.

Glue for Protection

I needed something to protect the LEDs on the fingertips so they wouldn’t snag. When the wires were safely stitched around out of the way, I got out the hot glue gun, put on the glove, put a dot of molten glue over the LED, and pressed the glue into a perfectly-shaped dimple in a USB serial connecter that I lifted from the CRATEL lab. It soaked into the fabric a bit, adhered very well, and made a nice smooth bubble over the LED.

Well, mostly nice and smooth. Hot glue tends to stretch like pizza cheese, and I made a little boo-boo.

Battery Holder

For now, I settled on prototyping just two fingers, because I have only two colors of SMT LEDs, and because it seemed like enough for a proof of concept. I steered the appropriate wires onto the same paths and ran them up to where I wanted the battery holder.

Then I clipped the wires short, scraped the ends, folded the glove over to get room to work, and wrapped the wires around the leads of a surface-mount coin-cell holder salvaged from a PC motherboard.

I soldered them on, then folded the battery holder leads underneath so they wouldn’t poke through the glove and scratch up the hand. Laid back out, it looked like this.

The Lights

I didn’t bother with a switch on the prototype–just pop in the coin cell and the LEDs come on. And because I’m using an old, small battery, I didn’t bother with current-limiting resistors–the internal resistance of the battery seems to be enough to limit the current to the LEDs.

Due to the idiosyncracies of my digital camera, the green LED on the index finger doesn’t look very green in this picture, but it really is. It’s a lot dimmer than the red LED, though; and I’m not sure why that is. Before going beyond a prototype, I’d want to pick current-limiting resistors to more closely calibrate the brightness of the different LEDs.

I don’t have a translucent globe to use, but I borrowed a hemispherical Tupperware bowl from my friend Lawrence’s daughters. (Don’t ask my why the girls have their own set of Tupperware–I really don’t know.) It works well enough to demonstrate the idea:

You can clearly see where the red LED is touching the bowl. Now all we need is a videocamera, some fancy software to track the lights, and an interface to a video synth.

Seriously, all of those things are Steve’s domain. I just wanted to see whether I could throw together a working glove, and I think this’ll do.

Oh–and when you want to shut it off, you take out the battery and drop it into the thumb.

This entry was posted on Sunday, April 1st, 2007 at 03:44 pm by Keith Neufeld and is filed under Miscellaneous, TechArt. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.