Final Post: The Completed LED Lamp

After many weeks of working and many hours in the engineering lab, our LED lamp is finished! It is ready for exhibition in the Science Center next Wednesday.

Description of the LED Lamp:

The LED Lamp is a lamp that gives off light from three LED-encrusted paper spheres which hang from a stand. The lights respond to the light cast upon them by brightening or moving.

Physical Construction:

The main structure of the lamp is a wooden stand with a triangular central stem and a circular platform. The stand was laser-cut and is held together with interlocking notches (the notches connecting the platform are superglued, since these notches are shallower due to the platform being of a thinner kind of wood than the rest).

The PicoCricket, along with three motors, rest on top of the platform. They are held in place by Legos hot-glued to the wood. The Cricket is lifted above the plane on Lego piers so that everything fits on a small platform, and the overall shape can be covered with a dome.

The light spheres hang from axles connected to the motors. The lights hang from the wires for the lights and sensors. To make the wires wind around the axles in an orderly manner, we made Delrin bushings to hold the wires in a narrow slot on the axle. At first, we made a simple donut-shaped bushing, but realized that the wires would not fit through when they were fitted with male headers. To circumvent this problem, we included a narrow slit in the bushing to take them on and off the wires. The bushings are hot-glued to the axle (to which the wire is taped) to keep everything in place.

The light spheres themselves are painted, hollow papier-mache balls with LEDs poked through them. There are 13 white LEDs on each sphere, which are wired together in parallel and connected to the purple and white wires that the spheres hang on. The other two wires in the ribbon connect to a light sensor which pokes out the hole in the top of the sphere. One of the spheres is held together with strips of glued paper, and the other two are held together with superglue, which we found out too late is actually a better method.

We added male headers to the other end of these wires to make them easy to plug into the ports on the Cricket.

Three light sensors, one touch sensor, and three sets of lights fill up 7 of 8 ports

All of these wires and motors are covered up with a dome made of strips of construction paper (a brilliant idea by Dania Wright).

Programming and Action:

The final program incorporates two modes: static mode and moving mode. Pressing the touch sensor will switch between these two modes.
In static mode, the program finds the average ambient light at the moment it is switched into the mode, and turns the lights to level 2 at that light level. From there, the lights get brighter if the ambient light decreases, and get dimmer if the ambient light increases. Each sphere changes brightness individually, based on its own light sensor.
In moving mode, the motors winch the lights up and down continuously at a speed proportional to the light falling on each individual sphere.
For more detail on how the programming was accomplished, see previous blog posts.

                                                                                     

The Process: Highlights

We started out with a very vague topic of "Art with Light and Shade." After a series of impossible ideas, we settled on hanging spheres of light. Though we originally wanted the spheres to be able to chase each other and move to areas of more or less light, we couldn't think of a good way to do that without derivative control, which would be very hard to implement with the Cricket.

The first thing we tested was using the sensor ports to power lights. With Robbie's help, we could light up and vary the power of the LED's on a breadboard. From there, I wrote a program for the static mode of the lamp as the works-like model. Next, we needed to come up with a program to run the motors. At first, we wrote the methods in the text language, but it produced a lot of inexplicable issues such as certain ports failing to turn on and ending the program in the middle. When Ashley transferred the method into the blocks language, the problems were solved.

Now we had a working program, but no physical device. We started by making the spheres and wiring the LEDs and light sensors into them. Then we had to think about the stand. We wanted a center-stem design with a domed top. At first, we just made the stand itself, which we designed in Solidworks and assembled on the same day. At this point, we were able to put everything on top of the stand and make sure that it would be stable. Then, we attached headers to the wires of the spheres and attached them to the Cricket, finished our winches with Delrin bushings, and had a working model. The only adjustment we had to make to our earlier program was to change the number of counts that the motors would go up and down, based on the length of the wires. Finally, we made a dome out of construction paper and the lamp was complete!

Final Testing:

Leading up to the exhibition next Wednesday, we did some final testing and recorded the results in this very positive and heartening table:

We will be presenting in the Engineering Lab, so we will be able to exhibit its features either by turning lights off and on, or by putting the lamp in a dark box.

Final Impressions and Improvements:

Overall, both Ashley and I were extremely pleased with our final product! A lot of things that seemed like they would be major issues, such as making a balanced stand and being able to winch the spheres up and down, turned out to be very doable. Although the lamp is far from our original vision of light spheres that could react to one another and move to areas of higher or lower light, it is really cool that it is able to react to ambient light in two different ways. It is also beautiful.

If we were to do this again, I think it would be helpful to plan all the components more concurrently, instead of, for example, finishing the LED control before we knew that it was even feasible to wire them inside of the spheres, and finishing the spheres before we had any idea what form the stand or dome would take. On the other hand, our linear approach ensured that all the parts converged at the end in a working manner, and subsequent parts were able to account for the quirks of earlier parts (for example, we could build the winches specifically for the spheres to hang on their wires alone, instead of strings and wires as we had originally thought it would be).

In the realm of improvements, if the spheres were heavier, they would do better at hanging down straight, and would not swing as much. If we added weight we would also have to come up with a better, more durable way to connect the wire and light sensor to the spheres. We would also have liked to adjust the stand so that the switch could have threaded down through the stem and come out near the base. Also, the dome could have been constructed better if we had glued one end of the paper strips at a time, making the dome more symmetrical and keeping the sides from bowing out. It would also be nice if we could come up with another dome material that might be more soundproof than paper, to make the sound of the motors less noticeable. Also, it would be worth looking into the reason why the lights blink briefly whenever a motor reverses direction, which we assumed was an internal Cricket quirk which might be hard to overcome.

Demo videos:

Photos of the completed product (and . . . others):

Acknowledgements:
Ashley Thomas was an awesome partner!
Amy Banzaert was ever helpful!
Robbie Berg knows a lot about Crickets!
Lyn Turbak asked a lot of questions!
Dania Wright is a dome genius!
Cassie Hoef thinks creatively about fire mitigation!