This project uses electroluminescent wire (a.k.a. "EL wire") to create a glowing, flashing, spinning piece of eye candy that could be used as decoration, a disco light for a dance party, or just for taking cool photos. This is definitely a work in progress.... It started with some strands of EL wire that were leftovers from a project I took to Burning Man 2002 (the Jellyfish Bike -- but that's another story). I started to play around with this stuff to see what I could come up with. I ended up with some very interesting pictures. Folks on Make and Flickr started asking me how they were done, so here it is.
Step 1: About EL Wire
Electroluminescent wire (trade name LYTEC) is manufactured by Elam company of Israel. It's available from sources such as CoolLight.com, coolneon.com, and many others.
EL wire is thin and flexible, can be bent, wrapped or even sewn into clothing. It runs off of high-voltage, low-current, high-frequency AC, which is typically supplied by a battery pack with an inverter, also sold by the same companies. EL wire will eventually "burn out", depending on how hard you drive it. The wire itself has a central core coated with phospor, wrapped with two very tiny "corona wires".
My EL wire came in convenient 6-foot lengths from CooLight.com; each length came pre-soldered with a connector at one end and a non-conducting alligator clip at the other for securing the "tail" end of the wire to anything handy. EL wire can be soldered, it's a little tricky, but there are some good instructions here. The connectors can be any basic 2-conductor variety. Locking, hooded connectors are probably best, to reduce the risk of accidental shock. I got the connectors from CooLight, but it looks like these connectors from AllElectronics.com are pretty much the same thing.
Step 2: Powering EL Wire
EL wire is powered via batteries and an AC inverter. I got my inverter from CoolLight.com, but it looks like this exact item is no longer available. Look for an inverter that matches both your preferred power source (e.g. 1.5v or 9v batteries) and the length of EL wire you want to drive. My collection of wire totalled about 45 feet, so I got an inverter that runs off of 9 volts and can drive 50 feet of wire.
For longer battery life, I used two 9v batteries in parallel with a little switch. For convenience, the inverter output goes through a connector that matches the EL wire connectors.
Step 3: Winding the Tube
The original idea was sort of a "barber pole" of spinning, glowing EL wire. I used a section of 2" ABS pipe I had lying around (PVC would work just as well) and braided the wires around it. I wound 3 wires (all red) in one direction, and two yellows plus a green wire in the other direction.
Conveniently, all of the other parts fit inside the tube -- the batteries, switch, inverter, and wiring harness -- and a balled-up sock stuffed in the tube held it all in place.
Step 4: Making It Spin
I collected some motors from various surplus stores; finally found a nice sturdy DC job with low revolutions -- perfect! The motor "mount" is actually just a metal junction box in which the motor is suspended. Pretty crude, but slightly more high-tech than the sock.
I made a power supply out of an ATX computer power supply, following insturctions similar to these. This works great, because in order to change the motor speed, all I have to do is change which plugs I use. A variable power supply would be best. Double true!
Step 5: Test Runs
The first few runs resulted in some major wobbling, because the tube was suspended off-center by a longish, too-flexible link. Still, the photos were pretty cool, which encouraged me to keep tinkering....
Not all the strands are illuminated in these images -- I unplugged one or more of them to see how it looked. The second part of this project (not completed yet) is to build a sequencer that I can program to turn on only the wires that go in one direction, or do other cool patterns. For now, I have to stop the motor and manually plug or unplug individual wires' connectors.
Step 6: Happy Accident
I have seven stands of EL wire, but only six were used for this project. One evening, I wanted to check the "leftover" blue stand for brightness, so I plugged it in to a connector on the tube. It occurred to me to turn on the motor. The results were very interesting.
Step 7: Unforseen Consequences
I freed the rest of the wires from the ABS tubing, thinking I could get even more "happy accidents" of the same kind. However..... My first thought was to construct a sort of umbrella structure out of coat hanger wire. This was totally unsatisfactory. With the flexibility of the motor suspension and shaft connection, any imbalance led almost immediately to twisting & shimmying.
So, next I tried to create a more stable platform by cutting a hexagonal piece of wood. I thought the gyroscope effect would help. Also devised a (mostly) rigid connection between the motor and the rotating part. Still, it didn't help. Either the whole motor/armature package need to be rigidly attached to something, or the armature and wires need to be perfectly balanced. One thing that seems to help is increased weight at the bottom.
Step 8: A New Approach....
Instead of a circle, I tried a bar for attaching the EL wires, thinking it would be easier to balance a straight bar rather than a circle (hexagon). It did seem to work better, especially with a heavily weighted bottom bar, but there was still a problem with instability at higher speeds. At low speeds, though, there was a nice "concentric column" effect that seemed pretty stable. I still need to figure out some way of rigidly attaching the motor -- I think that would help with the instability
Step 9: The Sequencer (design)
The 8-channel sequencer will switch the wires according to programmed patterns. It uses a Basic Stamp II microprocessor. The design is based on Mikey Sklar's el pants & bag and Greg Sohlberg's Rhino-8 sequencer. I used the Basic Stamp II for the processor, and went with Greg's suggestion & used a 9-pin connector, with 8 HV outputs and one "common", instead of individual 2-pin connectors for each of the 8 EL wire channels.
For my first attempt, I used triacs for the EL output. However, this turned out to not work right -- the triacs were triggered all the time. I'm not sure what went wrong, but having so much voltage so near the Stamp made me nervous anyway, so I redesigned the circuit to use opto-isolated triacs. These come in 6-pin DIP packages and consist of an LED next to a photo-sensitive triac, so that the low and high voltages can be kept separate. I used MOC3031M's from Mouser. The schematic is shown below. The MOCs are actually used as triggers for regualr triacs. Just wiring the HV to the MOCs won't work.
To create the board, I used my homemade PCB technique, explained in detail in my instructable here.
(1) Basic Stamp II (plus separate programmer board -- comes w/ BS starter kits)
(1) 24-pin DIP socket, 0.6" (you need to be able to remove the Stamp for (re)programming)
(8) 330 ohm, 1/4 watt resistors
(8) opto-isolators, 6-pin DIP package, MOC3031M or similar (I used Mouser #512-MOC3031-M)
(8) triacs, 400v or higher, TO-92 package (I used Mouser #511-Z0103MA)
(1) 9-pin connector (I used CAT# CON-90 from allelectronics.com, but anything similar would work)
(3) 2-pin locking connectors (I used some that were left over from an earlier order to coolight.com, so they already matched my inverter/battery pack inputs & outputs, but it looks like allelectronics.com part #CON-240 is the same thing)
(1) 2-pin header type connector (optional -- for the aux input -- I didn't use it on my board)
A note regarding the connectors: I designed my sequencer and other parts to be easily repurposed for other projects. So, all the main parts (battery pack, sequencer, wiring harness, inverter, and wires) are separate pieces that use the same kinds of connectors. That way, I can plug the inverter output directly into a strand of EL wire to test it, or use only a couple of sequencer channels instead of all 8, or not use the sequencer at all. All inputs (HV into the EL wires, 9v into the sequencer board, 9v into the inverter) use female connectors; all outputs (9v out of the battery pack, HV out of the inverter, HV out of the wiring harness) use male connectors. The only exception is the 9-pin connector that I used to organize the HV outputs from the sequencer board. That connector lets me reconstruct the wiring harness according to the needs of a particular project, without having a mess of connectors sprouting out of the sequencer board. You might want to use a different type of connector for the HV side for safety, and you might want to use a different arrangement/system of connectors entirely. Other sequencer builders (Mikey) use ribbon cable for outputs; that's a good idea too...... whatever works for you!
A note about the controller: I used the Basic Stamp II for several reasons. First and foremost, my co-worker had one he loaned me, along with the programming board, so it was free. Also, I'm totally new to controller programming, but learned BASIC years ago, so the BSII seemed very easy to learn -- and it was. Finally, the BSII has its own onboard voltage regulator, which simplified the circuit design. You could use almost any kind of programmable microcontroller, like a PIC or whatever. Obviously the pinouts would be different, and you'd have to include a voltage regulator in the design.
Step 10: Sequencer (construction & Programming)
Here's the final sequencer board. To create the board, I used my homemade PCB technique, explained in detail in my instructable here.
The microcontroller is programmed via the Basic Stamp Editor using simple Basic language commands. Programming the stamp is done with a separate board with a serial port for connecting to my computer. Once the stamp is programmed, it can be removed from the programming board and inserted onto the sequencer board, ready to go.
I wrote two BS2 programs (so far) to run the sequencer. SEQ1 uses the random number generator to select from a fixed set of patterns for turning the output pins on and off. Each of the 20 patterns comprises a single byte. The leftmost six bits control six outputs (pins 2-7). The rightmost two bits define the duration of the pattern display: 00 = 5 seconds; 01 = 10 seconds; 10 = 20 seconds; 11 = 40 seconds. None of this is truly random, of course; there are only 20 patterns and they are predetermined.
SEQ2 is quite different. It first runs a series of "chase" patterns -- outputs 1-6 are turned on sequentially in one direction; then two adjacent outputs are turned on & chased, then three, etc. After all the wires are lit, the chases repeat, with descending numbers of lit wires, in the opposite direction from the ascending chases. Next, a series of steady illuminations of 1, 2, 3, 4, 5, & 6 adjacent strings, followed by the same in reverse order. Then the whole thing repeats in a big loop.
The two videos show the sequence running without the tube spinning.
The sequencer can of course be used for other projects besides this one.....
Step 11: Structural Changes
For the final design, I used a piece of 7" 24-gauge steel flue pipe. This pipe is nice and solid, fairly heavy & powder-coated black. Very attractive, but a little difficult to work with. I drilled 1/4" holes on either side, top and bottom, for threaded rods. The rod on top also passes through a large 32-oz yogurt container, which holds the batteries, inverter, and sequencer. I stuffed old socks in to secure the electronics.
There are four nuts near the center of the top threaded rod, which can be moved and tightened to fix the location of the suspension point. The seam down the side of the flue pipe adds weight to one side, unbalancing the pipe, so I needed to be able to adjust the balance. I also affixed some heavy washers along the bottom rod with wing nuts, so those too can be moved to adjust the balance.
Step 12: Done(?)
Well, it works now and looks very cool -- but the photos can't show what it really looks like in operation. I'll try to add some video clips..... Some of the chase patterns are really hypnotic. For example, at one point, as the wires spiral up, the lit wires shift downwards at about the same apparent speed, so that it looks like a single wire flashing through the whole range of colors while remaining motionless.
Watching the "drain" end of the tube is also interesting.... the angle of the wires decreases (with respect to the tube end) near the ends, so there is sort of a (difficult to describe) "trailing" effect as the glowing wires reach the end of the spinning tube. It might also be an optical illusion; I can't tell for sure.
The tube wobbles substantially while spinning up to speed, but then settles down to a tolerable level. I don't think I can eliminate all wobble.
One possible direction for future development would be to add a magnet to the motor and a magnetic pickup to the top support rod, so that I can time the sequencer changes to the rotation of the pipe. Any suggestions?
The sequencer itself could be improved by adding a serial port so that it could be programmed without having to remove the Basic Stamp from the board.....
There are a few quicktime videos attached that give some idea of how this looks.
Step 13: But Wait, There's More.....
I was still not satisfied with (a) the excessive wobble and (b) the overall crudeness of the setup, with several different subunits to assemble every time. So, I went back to PVC pipe for the main tube. The motor is now enclosed in PVC fittings, with a 3" end cap on top mounted securely to the motor casing. the motor shaft is connected to a short piece of thin-wall 3" PVC drainpipe. The "bell" or flare on this pipe is just larger than the diameter of the motor housing. There's a 3" connector between the motor assembly and the main tube, which is removable. The sequencer and EL power supply are now located at the bottom of the main tube, which is capped by another removable 3" cap with a hole for the switch.
This new design is much more self-contained and attractive -- it's now a single unit (except for the separate power supply for the motor). The motor assembly can be detached as needed, and the motor itself is completely enclosed. Best of all, the rigid structure virtually eliminates wobble, so now I can run it at almost any speed.