edit SideBar

Main / GlowClouds


The goal of this project is to build an adorable, super nerdy, and interactive tiny piece of tech to share with friends. The aim of this project as far as it relates to the PIA project goes is to get some practice in really rudimentary 3D modelling, by designing and printing a pedestal for a custom lamp to sit on. The pedestal will hold the chip, power source, a switch, etc. to run the rest of the project.


  • Ordered a roll of 1.75mm, glow in the dark blue, standard PLA filament, that will work with one of the 3D printers at the Hacklab, as well as the printer that we have at home.
  • December 23: calibration object on Type A using new filament
    • downloaded Cura for Type A
    • Calibration object
    • Hacklab standard printer profile (Note: change temperature from 225 to 180 for the Type A]]
    • Open Thingiverse file in Cura, adjust profile, export to GCode, upload into Octopi (Quick Start Instructions for the Type A), print
    • Note: print bed is adjusted a little high at default, make sure to lower at start of print so there's space for extrusion. Near the end of print, things got a little goopey and messy as well, so I adjusted back down again smidge close to the end, and that seemed to help.
    • Note: print turned out well but did not cool layers enough at the top, ended up messy/goopey
    • Stuck well to new tape placed before print
    • Print
  • December 23: rabbit on Type A using new filament
    • rabbit
    • attempt 1
      • same process and adjustments, adjusted setting for minimum layer time to 20s
      • added supports touching buildplate
    • attempt 2
      • reduced minimum layer time to 15s
    • printed for about 5 minutes, printer stopped extruding unexpectedly with no apparent cause
    • determined that the print head had probably cooled, as the gears were turning but grinding through the filament, rather than feeding it through
    • for next test, reduce minimum layer time and keep a close eye on extrusion
  • January 1 - 4: printer maintenance of the Type A printer (see details here
  • January 10
    • completed first draft OpenSCAD design for 3D printed base for glow cloud
    • printed a test print on the Type A printer that came out perfectly calibrated, but with some gaps in extrusion (rabbit, from above)
  • January 17
    • another series of printer calibration tests that proved that Type A works very well with the new default presets on the updated version of Cura for Type A, and then also that there are no problems with the filament that I purchased being used in the Type A - no extrusion gaps in the calibration objects
  • March 1
    • finished final touches on the glow cloud base based on measurements from the actual button and power cable parts, exported, and printed a one-third size prototype
    • results - double-check the height of the middle layer (does this need to be shifted? right now is cutting off holes for power cable and button), add a hole through the middle for wires to go through (4 mm in diameter), find a way to easily get rid of support material (Emailed a colleague to ask if he knows of a miniature version of an orbital sander - suggested a profile sander, which can be found in a grey case in the shop at Hacklab)
  • March 3
    • repaired design flaws discovered in previous prototype
  • March 4
    • printed new prototype which worked out perfectly except the middle layer appears to be too thin. This should fix itself once the design is scaled to actual size.
    • design is now ready for a full-size print
  • March 5
    • attempted first full-size print, but 2.5 hours in (print was curling a little but otherwise going well), the eeepc running the printer bank froze, freezing the website interface and the printers themselves and causing the print to fail
  • March 14
    • after the first full-size print failed, I realized that the reliability of the 3D printers that I have access to for a project this size is probably questionable (ie, if I waste 2.5h of time on each of eight prints needed to complete this project...), and decided that it was worth investing the time to re-model the glow cloud bases in OpenSCAD to be laser cut from acrylic rather than 3D printed
    • It took almost no time at all to re-design in OpenSCAD, but I realized that I had never used the software with the Hacklab laser cutter
    • after talking with another lab member, I learned that I would need ImplicitCAD to be able to translate the models that I had made into vectors that the laser could interpret
    • potentially important note at this point: I use a Mac OS
    • Because the cloud version of ImplicitCAD isn't currently working, I had to use the git to install the program on my computer
      • this involved, first, installing XCode
      • order to download the Haskell platform
      • and finally, I followed the instructions on the ImplicitCAD git (under "Try ImplicitCAD!") to install and test ImplicitCAD
        • I used the Homebrew code in the command line, aka Terminal to instal GHC (the Glasgow Haskell Compiler and Caball
        • installed the latest release
        • and used the Haskell ImplicitCAD test to test it - this should result in the creation of a file called "test.svg" in the directory that Haskell was started from. In my case, I had no idea what directory that would be, so I used spotlight to search for the file - which was in my home directory.
  • March 23
    • went to Plastic World today and bought an acrylic rainbow to laser cut the glow cloud bases out of, after some basic math to calculate the quantity of materials I'll require

Page last modified on June 13, 2016, at 09:42 AM