I started working on the fireworks burst patterns and then ran into a chicken-and-egg paradox. "Chicken-and-egg" is a metaphoric adjective describing situations where it is not clear which of two events started first. What do I need first the pattern or a working RGBLED OFC (RedGreenBlueLightEmittingDiodeOpticalFiberCable)? I put aside the fireworks burst pattern design to construct a working RGBLED OFC. This would provide me the knowledge of OFC strand count per LED. The strand count allowed me to determine the number of pattern holes and density (i.e. how many strands) to determine size.


{gallery} My Gallery Title


bom.png: plastic styrene tubing and RGBLED


building_LED.jpg: cut the tubing, insert the OFC strands, drop of superglue and push in the RGBLED


led_testing.jpg: Using the E14 swag breadboard to confirm the RGBLED OFC's are working after assembly


patterns.png: patterns generated in Inkscape, Preferred choice number 1


do the math.jpg: three rings, four RGBLED's (two for outer ring), add a button and a trigger output to get GPIO totals


The RGBLED OFC were easy to construct. I just filled the styrene tubing with strands, cut the strands from the lamp stand, pored some glue in the tube and stuff in the RGBLED.  Two sizes of tubing were used. This enabled a strand count reduction while still being able to fit the RGBLED. I've thrown a few pics up of the materials used during the build.


My own self criticism is that the assembly process could have done with a bit more rigour. I used what was available and didn't spend a lot of time perfecting a design. Light coming out the end of the strand was the goal. Trust me working with superglue and plastic has left me with no finger prints for at least the next few days.


With the RGBLED OFC strand holders built, I went back to trying to determine a pattern. I used Inkscape to generate hole patterns drawings. Pattern 1 was constructed by hand placing each dot. It became the design of choice.  Using this pattern 1 and the OFC holder strand count the following was decided. Three rings of strands, covered by four RGBLED OFC strand holders. Two of the strand holders would be placed in parallel to provide coverage density of the outer ring.


I grabbed a sheet of graph paper and did some math. The Pi has limited GPIO's. I adjust the number of RGB LED pins that could be used to ensure there were two GPIO's remaining to accommodate a launch button and the LED shooting sequence trigger. That gave me a total of 27 GPIO's. To light all RGBLED OFC would require 4 LEDS X 3 pins per X 3 rings 36 GPIO plus two addition 38. That exceeds the Pi maximum GPIO count. The pin count reduction reduces the eight colour combinations available from the RGBLED.


While assembling the RGBLED OFC's are started exploring a connection method. Looking through the old parts bin a possible solution came to mind. It minimizes the effort to establish a connection and also aids in testing and troubleshooting. I would like to do some further exploration so I will save the connection/electronics for another post.


Pattern decided and RGBLED OFC's assembled. An added bonus it rained here today. As a result, one of the farmers dropped because he can't harvest with the crop being wet.  The project has generated some interest. He has offered to assemble a display panel, drill the holes and mount the RGbLED OFC's. He tells me the rain will keep him off the fields for about three days. That allows me to focus on the electronics and the assemble can happen in parallel.