Making the Maker Hub: more 3d printers
Editors note: Elon is opening a makerspace, called the Maker Hub, this fall. In preparation, student staff member Connor TeVault is spending the summer testing and configuring equipment that will go into the space. He’s keeping track of quirks and lessons learned in this series of blog posts.
Since the last blog post we have received another 3D printer, the Fusion3 F306. This printer has a massive build volume of a cubic foot, dwarfing the Wanhao 4S. It also uses a different geometry scheme, called CoreXY. CoreXY printers use a series of cables and pulleys to move around the printhead, reducing the amount of moving mass and allowing the printer to move faster and more accurately. The difference is noticeable immediately when comparing parts from the two printers, as well as in the level of involvement required to get a successful print. Thanks to the speed and quality of the F306, the Wanhao has not been getting as much use, but it still has its advantages. One of the other big differences in the two printers is the firmware they use. The Wanhao requires files to be loaded in .x3g format where the F306 uses the traditional .gcode format. Because of this difference establishing a workflow for the two printers is very different, and currently I am tackling several incompatibilities in the software we want to use to control printing.
Raspberry Pi and Octoprint
I have been working on setting up a Raspberry Pi computer as an Octoprint server. Now that this is finished, users can access the Octoprint web interface at http://makeelon.pagekite.me/ (if this link doesn’t work just try again later, the server is still undergoing frequent revision) to monitor and control the new F306 printer. On the site, click on the ‘control’ tab to see a live stream of the 3D printer, which often runs all day and through the night. Under the ‘timelapse’ tab you can download an automatically generated timelapse video of the build process! This web interface also allows users with the proper credentials to upload files remotely and begin printing them, a feature we intend to explore as a solution to printer workflow in the Maker Hub. Unfortunately, as of this writing, the Wanhao is not supported by Octoprint thanks to the firmware onboard. The developers of Octoprint are working on a solution so I’ll revisit this issue in the near future.
RepRap: 3D printing a 3D printer
The RepRap Project was started in 2004 with the goal of developing a low cost, open source 3D printer that could fabricate the majority of its own components. Since then it has expanded from one man and a single 3D printer design to thousands of contributors worldwide and hundreds of design variations. These printers are all designed with 3D printed parts, their plans and specifications are freely shared, and all run open source firmware. By using the collaborative power of the internet, the RepRap Project has been able to reduce the high costs involved in design and research and evolve its products at a tremendous rate.
Several years ago I built a RepRap Prusa i3 using a laser cut wood frame and 3D printed parts ordered online. The Prusa i3 was and is one of the currently supported/under development RepRaps, and is commonly considered the entry model given its low price and relative simplicity. The experience of building a 3D printer from parts was incredible, and took several days. Calibrating and dialing in settings took longer on top of that, but after the work was put in I had built by hand something that could build nearly anything! Over the past few years a lot of advancement in the field of 3D printing has happened, and after spending so much time working with the Wanhao and F306 I have found myself desiring a more advanced personal 3D printer. To this end I have decided to use the non-printable parts of my old Prusa i3 (motors, nuts and bolts, electronics; also called ‘vitamins’) in the construction of a new printer for myself.
I knew starting off that I wanted to build a “delta” style printer. I was drawn in by the clean design, flashy maneuvering, and characteristically high print speeds and quality. The easiest way to understand the kinematics of a delta printer are by watching a video of one printing:
This video showcases two different sizes of the Kossel RepRap variant at work. The Kossel itself is an evolution of an earlier prototype delta printer, the RepRap Rostock.
I have chosen to build the Griffin Open Source 3D printer, a variant of the Kossel currently under active development focusing on frame rigidity and maximizing print speeds. My first step is to print out all of the necessary parts, then to strip apart the donor Prusa i3. I will be posting updates to the blog over the rest of the building and calibration process so be sure to check back!
In the last blog post I mentioned that I was looking for alternatives to Solidworks, a very expensive professional 3D design software. We need a 3D CAD/CAM program to use in the space so students can design their own 3D objects to print, but justifying multiple copies of a program that costs thousands of dollars is hard given the use case. To this end I have found an excellent solution: Autodesk has recently released a version of their Fusion 360 design software (roughly equivalent to Solidworks as well as some killer cloud-based features) and is providing it for free to students, hobbyists, and startup companies! After using the Fusion 360 suite for several weeks I think it has a very powerful feature set for mechanical design as well as artistic modeling, and will be a great tool for our space. I am especially impressed by the set of 3D printing utilities included in the program, allowing users to double-check their models for structural integrity and 3D printing compatibility.
Acorn Award Update
After a lot of brainstorming and trial and error, I finally have succeeded in reproducing the Acorn Accolade Award. The key to the solution was accepting the inherent limitations of 3D printing as a fabrication technique, and embracing alternative methods that take advantage of 3D printing’s capabilities instead of just trying to design around flaws. Because of the layer by layer construction of 3D printed parts, the orientation of a design makes a big difference to how the final product looks. Text looks great when played down flat, but building it up layer by layer leaves artifacts on the edges of the letter, making printing it in the final orientation impossible. The acorns, on the other hand, must be printed standing straight because the foot of the base is larger than the rest of the part. By splitting the award up into two pieces, a base and a snap-fit faceplate, the award can be printed with the desired quality and features. Another change that had to be made was cutting down on the amount of text present on the faceplate. Although it is possible to print very small text, it becomes illegible very quickly as the smaller features of the text become too much for the printer to handle. This was especially evident when using a font with a serif, which does not print well. Despite having to make some changes, I am very happy with the way the awards came out!