This is an array of the empty dog food cans, each with a single LED light inside and a colored lens on the opening. The LEDs are controlled by motion detectors triggered by interaction from the viewer.
By using LEDs as a light source for each of the cans, the power requirements are low. The circuit uses a small amount of components for driving the LEDs, and this instructable will describe in some detail how it uses motion detectors, transistors, resistors, and LEDs to create the interactive light show.
I am a novice at electronics and have only recently created my first circuit design and built this project with success. Anyone interested in electronics could easily reach my level of expertise as a successful novice by reading and doing, my one bit of wisdom learned is that if you can afford better tools, that is the way to go.
As an artist, I really do not want to promote one store or product over another, but my community does not have the best selection of stores where I can find good electronic components, so I list Radio Shack as just "Shack," replace with your favorite store or supplier.
64 dog food cans (washed)
32 green 10mm Super bright LEDs (www.evilmadscientist.com)
32 blue 10mm Super bright LEDs (www.evilmadscientist.com)
50' hookup wire (Electronic supply, guessing since I did not count usage)
10 cedar panels (Hardware store)
2 aluminum angle bar (Hardware store)
2 aluminum bars 1/16 inch thick (Hardware store)
8 1/4 w 1K resistors (shack)
8 PNP transistors (Shack)
8 DP-001 Motion Detectors (www.glolab.com)
8 Fresnel lenses (www.glolab.com)
5' heat shrink tubing (for a professional end product, coordinated colors are cool)
1 9V 800ma power supply (Shack)
1 switch (Shack)
1 round PCB (Shack)
31 brass #8 screws (Hardware store)
31 brass #8 nuts (Hardware store)
31 brass #8 washers (Hardware store)
32 glass lenses (Original idea was for paper, vellum, and mica or any form a silhouette masking)
1 Extension cord
Hot glue gun (better than duct tape)
Wire strippers (do not rely on teeth, the tool pictured here is the best tool for the job)
soldering iron (don't cheat yourself here, I became better with a better iron)
Heat gun (only necessary if you are heat shrinking your wire solders)
helping hands (optional but highly suggested)
magnifying glass (optional)
Breadboard (optional, but necessary tool for anyone serious about electronic circuit design)
1 39K resistor (sensitivity programming DP-001)
1 2.7K resistor (dwell programming DP-001)
1 multi-size drill bit (a must over a standard drill bit)
1 screw driver
1 hammer (optional, smashing a toe removes the boredom and tediousness of soldering 64 LEDs with 128 wires)
1 Caliper or scale
1 wood glue
2 long screw clamps
Vcc = source positive
Vdd = FET positive, the power supply provides positive to the detector, the NFET transistor on the DP-001 outputs a positive value at the terminal we call this Vdd
Vss = source negative.
As an artist working mainly in oils and recently in more high-tech pieces, I have also wanted to incorporate a little green into my work. I have two pugs and they seem to like eating everyday, which leads to a waste from food containers, so I started saving the cans for some future project I knew that I would come up with when I had a larger collection.
Another artist friend, who works in fused glass, mentioned that there was a juried show that had "collaboration" as the theme, and we decided to work on an art piece together. It was a perfect opportunity to use those dog food cans that were taking up residence in my garage. With so many cans, it was apparent that the piece must take the form of some sort of an array, which lit up by the motion of the viewer. We met at a local coffee shop and I laid out my plan, the name of the piece came as natural as nature itself, an array of light using an electrical charge.
Here is a quick description of the work and process creating this piece of art.
Step 1: Building the Frame
The cedar panels were found at a local hardware store and were designed for lining closets. The cost was inexpensive $23 dollars for 12 planks; they were perfect for the project. They were also chose for color and form with an added benefit of the slight cedar aroma.
First the face of the planks were sanded and coated with a flat Varithane to prevent them from attracting grease and dirt through handling, and to bring out the color of the cedar.
The planks are 3.75" wide and 48" long, perfect for the matrix to fit within the width and height of the planks creating perfect spacing for a square matrix.
The dog food can diameter is 3" and finding a hole saw this size was easy. I measured the centerline of the planks and then the distance between the centers of two planks side by side.
Using this measurement I spaced the holes along the vertical line of the planks to create a square array of cans. This provided me with some space on the top and bottom of the piece, to balance the piece horizontally, two blank planks were added, one on each side of the matrix.
Drill the holes for the cans using the 3" hole saw, sand the hole and test the can in the hole to test the opening. Glue the panels together with a small amount of wood glue and clamp together, let dry overnight.
I wanted the ends of the cans to be even and the base so they would protrude through the back of the panel by only 1". Using stacks of the hole pieces that were drilled out of the planks in order to level the completed panel face down so that each can protrude by 1" through the back. Using the hot glue gun, a bead of glue was placed around the base of each can securing them to the panel.
In order to give the piece enough strength so the panels would not crack and separate when handled, the planks were also tied together on the top and bottom with a flat aluminum bar and a piece of angled aluminum. The flat bar could be left out, but I wanted strength and have been known to over-engineer from time to time.
First line the bar and angle bracket with the edge of the panel, clamp then drill a single hole through the vertical centerline of each plank, one on top and one on the bottom. Tie them together with the brass screws, nuts and washers. To add strength to this application, a bead of hot glue down the length of the bars and the planks. I also put a small hot glue bead at the base of each nut to keep them in place; the frame is ready.
Next prepare the cans. The interior of the cans were a gray color which absorbed the light from the LED, in order to get more of the light to strike the lenses it to bounce around, this was accomplished by painting the inside of the cans with marker paint.
The reason for the choice of chose marker paint was due to its nozzle, which is designed to point down at the ground so the nozzle is straight which makes painting the interior of the cans easy. I also wanted the colors to shift somewhat so I chose a red, green, blue, white and yellow colors; at this time, the look and color was not known to me since my friend was busy making them while I built the frame and electronics.
To drill the holes in the cans, a standard drill created a burr, which was too difficult to clear and also make the hole oblong once de-burred. By using a step drill bit, the hole is clean because this bit will mill the edges of the hole as it drills making a perfect round hole the right size for the LEDs.
Next I measured the diameter of the business end of the DP-001, so I could drill holes in the panel for them to peek through; picked a corresponding drill size and laid out a circular pattern for the holes. This was to keep the consistent similarity with circles.
With all the cans in painted, drilled and installed into the frame, it is time to work on the electronics.
Step 2: Electronic Design
Understand that I am a novice at electronic design, if some of my interpretations about component operations are incorrect, then please post a comment so that the reader can find clarity.
Also the wire stripper tool was the very valuable tool on the workbench, it can save your teeth if that is your habit, and can save your sanity when stripping hundreds of wires; this is an inexpensive tool, but a great tool.
Before we add all the electronics, it is better to create a design and then test the circuit's operation. De-soldering is not the way to progress and you can waste many good parts that way.
First order of business is to calculate the values of the components and define the power requirements of the circuit.
The first component is the DP-001 motion detector, which has a power requirement range from 4v DC minimum to 15v DC maximum, which gives us a nice range to work with. The circuit will be driving 65 LEDs and each LED is rated to draw 20mA of current maximum. 65 x .020A = 1.3A (64 LEDs in cans and 1 for a power light), the current needed for the DP-001 is a low 45 microamperes or .000045A x 8 = 00036A, which is a very low power requirement.
I chose a 12v 800mA DC power transformer, realizing that I was not going to have all the LEDs on at the same time, and none will ever be on very long, this is has plenty of power.
Now that we know what power will be driving the LEDs, we need to calculate the size of the limiting resistors that will prevent the LEDs from burning out while keeping them as bright as possible.
This is a simple task of using Ohms law to determine how much resistance each LED needs to keep cool and bright. The LED's specifications say that the maximum current should not exceed .020A (20mA), you can push this value to make them brighter if the "on" duration is short enough.
Calculating the resistance needed, take the voltage and divide it by the max current value. 12v DC / .020mA = 600 ohms. I wanted to get the most light from each LED so a 470 ohm resistor was chosen.
Remember the lights will not be on continuously, so the danger of burning them out is small, plus 470 is close to 600. To check how much current will be drawn through the LED if we use a 470 ohm resistor we divide 12v by 470 ohms to equal .0255mA, a difference of .0055mA, which is negligible.
The DP-001 motion detectors can only sink 100ma of current, so driving all 64 LEDs from one module would not work, plus they would all go on at once, which would be less effective and somewhat boring.
If we divide the 64 LEDs by 8 and use 8 DP-001 detectors each driving 8 LEDs for a total current of 160ma per detector, it is still too much for the DP-001 that has a max sink value of 100mA.
The 2N3906 specification say that it can sink from 10 micro-amps to 100 milli-amps, but I would rather risk a transitor than the motion detection module.
How I choose a transistor that will work in our circuit:
There are two basic types of switching transistors that we will be looking at, an NPN or PNP transitor. The NPN and PNP designation describes their gates and operation. I chose a general-purpose PNP resistor, the 2N3906, it will not have to dissipate much heat and is well suited for this project.
Transistors have three connectors called the base, collector, and emitter. They are turned on by a voltage sensed at their base, which will open the gate and allow more current to flow between the collector and the emitter.
The difference of operation between NPN and PNP is that the NPN will switch on if the base has a positive voltage of 0.7v or more and will switch off below this value. The PNP is reversed biased and switch on when the base senses a low voltage below .07v and switch on above this value.
The LEDs are switched on by using the terminal out of the DP-001 to switch on the transistor that will allow current to flow through the LEDs. The DP-001 outputs a "high" at the output terminal and will go "low" towards negative when motion is detected.
A quick note on PNP and NPN transistors, I will not get into the construction of these components, just the fact that they behave opposite because they are biased opposite. The NPN transistor will conduct current between the collector and emitter if there is a positive voltage value difference between the base and the emitter, while the PNP will conduct current between the collector and emitter if the base senses a lower voltage between the base and the emitter.
We cannot use an NPN transistor because it is switched when there is a "high" on its base with respect to its emitter. Remember, the DP-001 goes "low" when motion is detected. So I chose to use PNP transistors since they are triggered by a "low" on the base with respect to the emitter, allowing current to flow through the transistor when the terminal of the DP-001 goes "low" with the detection of IR motion.
The circuit below is a simple circuit showing how the system will work, to add another 7 detectors, resistors, and LEDs all we have to do is copy this design eight times.
Here is some of the logic that went into the circuit designed below so that it works as planned and the components do not burn up in a cloud of blue smoke.
We do not need current to flow through the terminal output of the DP-001 and through the base of the 2N3906 transistor, we only need to have a logic switch between "high" and "low," to reduce current through the base of the transitor add a 1k ohm resistor (r1) at the output of the DP-001 terminal and the base of the 2N3906 transistor.
Before tying the LED anode to the transistor, we place a current limiting resistor (r2) with a resistance value of 470 ohms in between the two components.
When the DP-001 is not detecting motion its output terminal will be "high" (Vdd) and this high value will be sensed at the base of our transistor, blocking the flow of current between the collector and the emitter. When the DP-001 senses motion the output terminal will go "low" (Vss) and the transistor will switch on and allow current to flow between the collector and the emitter, lighting the LED, the 470 ohm resistor will limit heat causing current through the LED.
Step 3: Building the Circuit
I suggest investing in at least an average size breadboard, it is the good tool for a circuit tinker. First I tested the simple design using the DP-001, limiting resistors, switching transistor, and LED. When this worked as planned, I built the switching circuit with all eight transistors and resistors and hook them all up for a final test.
The simple circuit was test worked, when IR motion passed in front of the detector the LED lit up. At this point it was time to solder wires to all the LEDs, then wire all the detectors with their positive (red), negative (black), and terminal output (green).
In order to save space on the circuit board, I placed the current limiting resistor (r2) in-line with the wire that was tied to the collector side of the transistor.
The photos below show the "flower" circuit board, notice the yellow and red lines, each have a current limiting resistor (r2) in-line and covered with heat shrink.
Now prepare the 64 LEDs with all their positive and negative leads; this is where the hammer comes in handy to relieve boredom, choose to smash a toe because you need your fingers to finish the work.
Hooking up all eight detectors, transistors, LEDs, I wired them on the breadboard, with a wave of the hand, eight LEDs waved on and then off.
It was time to wire it all together. Since each detector will drive eight LEDs, I created a pattern of LED groups, making sure to spread the LEDs that would light up by any one detector. Tie all the positives leads of a group of 8 LEDs together. Now a take the eight groups of negative leads and tied them all to the common ground on the power supply.
Each LED group was clipped to the collector of the transistors; positive and ground was tied to the circuit board. The emitter side of the transistors were directly tied to Vdd and the collector side tied to the LED's anode through the limiting resistor while the LED's cathode was tied to ground.
The circuit test worked; the next part was to hot glue all the LEDs into their cans, maintaining a orderly routing of wires. The circuit flower was tied to the back of the array panel to a metal bracket with on the back of the panel zip ties.
Next I tied each group of 8 LEDs positive leads to the collector of a transistor on the flower. Next glue all the motion detectors into holes that were drilled earlier, be sure to use good wire management in order to keep the nest of wires from getting away from you.
On the front side of the array panel I hot glued the Fresnel lenses in front of each detector. Once the Fresnel lenses were in place, the sensitivity of the detectors was noticeably increased.
The 12v DC power supply wall transformer was then mounted on the back side of the panel with the positive lead tied to the switch and the other end tied to the circuit flower's positive connection.
The flower and motion detector's ground leads were tied to the system's common ground.
The extension cord was zip-tied securely to the transformer with tie wraps to prevent any pulling of the cord from disconnecting the power. The switch was mounted on the back edge of the panel with hot glue.
I used some pipe routing straps to hang this piece on the wall (first image), they were temporary and were chosen to keep with the similarity of circles in the overall design. They have already been exchanged D-rings and doorstops for standing the panel away from the wall.
This piece is very fun to play with, as a viewer moves around, patterns of light dance with the movement. In the future, I might re-wire this piece by adding a micro-controller and charlieplex the lights making cool patterns when there is no motion for a specific amount of time.