From Hive13 Wiki
|HIVE13 Tribute to Galileo|
Galileo's Finger is HIVE13's maker technology tribute to Galileo Galilei, the father of modern Physics. The original idea evolved into an interactive obelisk/kiosk/display thingie that we can take to Maker Fairs to publicize our Hackerspace. It is a multi-media art piece with a steam-punk theme.
<<< Challenge to other Hackerspaces - What is YOUR thingie? >>>
The project started back in March 2010 when HIVE friend Warren Trefz and Cincinnati's River City Works group decided to gift the HIVE with one of their one-of-a-kind works of custom blown glass. Warren is the instructor and project director at this unique glass blowing facility in Cincinnati.
The piece is essentially an egg-shaped glass goblet with a glass lid containing a glass replica of Galleo's finger. It turns out the remains of Galileo's real finger are contained in a similar goblet in a science museum in Italy. Why? Well, Galileo's invention of the glass telescope in the early 1600's was beginning to provide hard scientific proof for the Copernicus theory (that the earth revolved around the sun and not vice-versa) in a time when the powers to-be in Italy (the Catholic church) did not approve of such heresy. The whole thing makes a very interesting story that needed to be told.
The Museo Galileo in Florence, its presentation, and a close-up of Galileo's real finger.
Our goblet and its lid, how they fit, and how we assembled them.
So in the glass piece Warren and the River City Works folks made for us, the finger kinda looks like a glass pickle, the lid is not attached to the goblet, and the piece needed a bit of finishing before it could be safely displayed. HIVE member Jim took on the task of making an appropriate presentation for this unique artifact. The result has been an evolving/expanding active project that exhibits some of the HIVE's woodworking, brass forming, laser etching, dumpster diving, and Arduino-based electronics skills.
The above photo shows the egg-shaped, blown-glass goblet containing the green glass finger. There is a mounting that holds the glass lid onto the top of the glass goblet. The effect resembles a vintage hot-air balloon. The implications of lots of hot air are entirely appropriate. The goblet/balloon sits on top of a SteamPunk rocketship. The cockpit is half of a scavanged hall chandelier having a six-sided brass frame and beveled glass panel inserts.
The hex-shaped wood body is stained and varnished mahogany and the six feet resemble rocket nozzles like the whole thing is launching off. Laser etched panels on the sides explain the Galileo story in the original Latin and with an English translation. The third image is a top down view looking inside the body to show the location of the power supply.
Above is friend-of-the-HIVE Clyde Kober and his Epilog laser. He is showing the laser etching he did on one of the three sign plates for this project. Clyde makes his company's laser-etching services available to the HIVE at a very favorable rate. Check out his company website at the following link. http://www.cin-deescrafts.com/
Thanks, Clyde. They look great!
Above is what one of the signs looks like mounted on the body. Click on the photo for a close-up. It was spray-painted black and then the paint was sanded off flush with the top surface. The black paint stays in the laser-etched cavities of the border and lettering and has a really sharp look.
Click on any of the three above figures to see the artwork for the three signs; in the original Latin, the English translation, and the background for the story.
We decided this bizarre thingie needed to sit on something to be appreciated; but just any old table would not do.
We made a custom hex-shaped wooden base that makes the connection to HIVE13, Cincinnati, Ohio, USA. The above photos show just this base. We inserted a 36" round glass table top between the base and the rocket ship goblet/obolosk/thingie that sits on the glass table top and the base structure.
So this is the final guy currently parked at the HIVE. The custom base and glass table top ended up much nicer than the generic table we would have used shown in the background of this photo. Everything is just great, except... this unique obolisk/kiosk thingie can't just sit there, it has to do something. Enter the Arduino LED add-in to the project. The idea is for it to have some autonomous interactive function that would attract the attention of the casual passer-by.
The above picture shows the hex-shaped structure and the six approach vectors (two red, two green, two blue) radiating out from where the six distance sensors would be located. The objective is to have some distance sensing ability when a person approaches from one of these six directions. The desired range is roughly 36" down to 6", more or less. These are analog inputs to the Arduino. The Arduino outputs an increasing intensity to the red/green/blue LED outputs. A shorter distance equals a brighter intensity. Once up close, you could wave your hands in front of the sensors to dynamically change the rgb values of the merged LED light.
The left image above is the prototype using an Arduino hooked up to a MaxSonar EZ1 ultrasonic range finder as an input to pulse the set of six red, green, and blue LEDs. The pulse rate increases as you approach the unit. It is only a prototype at this point, but needs to be seen live to be fully appreciated. The right image is another prototype using a Boarduino (Arduino clone) to drive three TLC5940 chips and 18 tri-color LEDs. The animation sequence starts with LED1 switching from red, to green, then blue, then the same at LED2, LED3, etc. up to 18 and back down again.
Above is a poor resolution image of the schematic being developed for the real deal. There will be a more complicated arrangement with three sets of 6 LEDs. The intent is to use three AD5206 digital pots (each with six channels) per the Arduino example http://arduino.cc/en/Tutorial/SPIDigitalPot.
Well, the AD5206 digital pot chip turned out to be a bad choice, or a 'dry hole' as a well digger would say. Several HIVE yodas (thanks IanD and JamesS) have pointed out that changing voltage on LEDs is a poor way to control their brightness. PWM (Pulse Width Modulation) is the better way to go. While the Arduino itself has a few PWM capable outputs, it does not have nearly enough for what we want to do here. However, in the magic world of Integrated Circuits there seems to always be some other chip that can do what you want to do. Enter the Texas Instruments TLC5940 16-channel LED driver chip with DOT correction and grayscale PWM control. There is also an Arduino playground example on the web at http://www.arduino.cc/playground/Learning/TLC5940. The chip is only $3.50 retail at Digikey or Mouser and TI will even send you a limited number of free samples if you only ask. The following links are for the Texas Instruments TLC5940 chip as the LED driver.
For smaller size reasons we chose to use the PWP form factor over the NT (DIP) form factor for this chip. The bottom of the PWP form factor chip serves as a thermal pad requiring it to be surface-mount soldered. This was another advantage as it forced us to get practice with our hot plate surface mount soldering techniques described later.
The first photo above shows the rescued ATX power supply that will provide a regulated 5.0 VDC at 10.0 amps for the arduino and as many LEDs as we want. It will be mounted in the wooden base. The green 'enable' wire makes an on/off switch. The second photo shows the mounting for the three MaxBotix MaxSonar(R)-EZ1 sonar range finders. The third photo shows CAD cross-sections for a multi-part stack of acrylic disks with cutouts configured to mount the top six LEDs. These cross-sections show five different acrylic disks (at .093" and .220" thick) with cutouts that position the LEDs and connectors.
The photos above show the laser-cut acrylic stack for the pancake that holds the top six LEDs. It turned out really nice.
The 3-D image on the left above is a partial view of just the lower LED and brass rod structure. Only one of the six wings is populated with rows of brass rods and LEDs. The eighth (top) level of LEDs is built inside the core stack at the top as can be seen in the following cross-section. The center image above shows top and side views of the new design for the inner core. This structure has 48 tri-color LEDs arranged in rings of six on eight vertical levels. Each LED location is typically powered to be individually red, green, or blue, yielding 144 primary color addressable locations. Note all three (RGB) inputs can be simultaneously powered to yield a rainbow of color choices at each LED location. The core structure exists to route all the wire leads from where they are on the PCBs to where they need to be for the LEDs. Below the core is a matrix structure of brass rods that brings power out to each LED. Four rods are needed for each LED; one lead is the common/shared (+) anode and the three leads are the color cathodes (red, green, blue). The view on the right shows the detail for the cut out pattern for each unique acrylic layer in the core stack.
The above views show the cross-section of the core structure and two different 3-D views of the bottom LED structure. The core exists to route the 48 red, 48 green, and 48 blue wires from the six side-mounted custom LED Driver PCBs, to the structure of brass rods and ultimately the 48 LEDs. The core structure is a multi-layer stack of about 42 individual acrylic layers. Each layer is different, typically dedicated to routing six wires for one level and one LED color.
The above three views show the design of the top core stack of acrylic layers. The Lilypad Arduino mounted onto the acrylic layer that is colored purple. The base of the goblet sits on top of the tan layer and is constrained by the top blue layer. The two cross sections are cut 30 degrees apart to show the different internal features that include the tie rod hex nut pockets and sensor ribbon cable passages.
The above views show separating and preparing the numerous laser cut parts, and the setup at the start of the assembly. The anodes (+) for each LED are common and can be ganged together. The third photo shows how the eight brass rods from each of the six wings, are gang soldered and a jumper wire from each gang is then soldered to a brass disk washer installed on the central all-thread. This simplifies the first 25% of the detail wiring.
All the brass rods are inserted into the structure, then the appropriate red, green, or blue hookup wire lead is soldered onto the top of each brass rods. The hookup wire is organized for the addition of each layer in the core stack.
These views show the 48 red, 48 green, and 48 blue wire runs as each layer was added, threaded, and assembled in the core stack, one-by-one.
These views show the core.
These views show the top of the core and the I/O wiring to the Lilypad Arduino. The Arduino is hidden under the base of the goblet when it is all assembled.
These views show the assembly and the mirror effect.
The above sketch is hard to see, but it shows the wiring of the Arduino LilyPad and the PCBs. The other two views show the LilyPad from AdaFruit and the custom circuit boards built for this project. The core needs six boards, but we made nine to have some spares to practice on. Two of the custom boards in the above picture have been populated with seven remaining to do.
The above images are from EagleCAD. They show the wiring schematic, the board layout, and the "cream layer" that was used to make a solder mask stencil to reflow-solder the surface mount components for the boards.
Use this link for a detailed description of the reflow-solder operation on the PCBs.
OK, so the whole thing is put together and working now. It has been running live at the HIVE for several weeks. There is still a problem with the second blue circuit board to fix. Also, the three distance sensors are not obvious to the casual observer. We've been casting about for a way to highlight that they are there.
We started with an idea to fab some Maker-Bot parts. These would be a horn (or funnel) cone shaped part that fit over the sonar distance sensors. We made these in individual red, green, and blue colors, but they did not turn out great. Also the green horn seemed to affect the green distance sensing.
The new and better idea is to laser-cut some acrylic parts as shown in the following images: