Design Story and Advice - How not to design a 3D printer (A flawed mini 3D printer)
We calculate the final price from your selected file. You review and accept it before payment.
Saved on this device.
A learning experience for me, designing and building a 3D printer from scratch. Core XY motion system, klipper firmware.
- Rating
- 0 / 5
- Likes
- 9
- Downloads
- 11
- Makes
- 0
- Views
- 224
- Reviews
- 0
Updated at source:
Model information
- Source
- Printables
- Category
- Other Printer Parts & Upgrades
- License
- GNU General Public License v2.0
Preface (scroll to the bottom for issues and TL;DR)
Grab yourself a cup of coffee or tea because there will be a lot of reading going on. This is an article, not a description lol. This is just a heads-up before we start. If you simply want to read the description, go to this GitHub link. Alright, now enjoy reading!
Introduction
Hello there! Thanks for stumbling across my printer design. I’ve always wanted to achieve a goal I set for myself: designing and building a 3D printer from scratch. This has been a learning project—there are many imperfections, no doubt, but I hope you’ll learn something and avoid the mistakes I made if you ever decide to design your own printer.
Mistakes/Problems
Although the printer functions, there are issues with it. Firstly, there seem to be tolerance issues that might be caused by the design of parts. This has caused artefacts on the surface of the prints. I noticed this while building it, but no matter what I tried, the issue persisted. It’s most likely due to material shrinkage since I used ABS for the parts. I’ll provide a video showing the problem. Other than that, there are some minor imperfections, some of which might not have been discovered.
So, with that in mind, please don’t build this printer until the issues are resolved. (You can attempt to do so, CAD files on GitHub)
How it all started…
When I first had the idea to design a 3D printer, my plan was to build an E3D tool-changer-style printer inspired by the Voron belt system while keeping it budget-friendly. Everything seemed cool until my GCSE were coming up, and I had to put the project on hold. But I didn’t want to give up—I thought I could juggle both school and this design project. What could possibly go wrong?
To keep it going, I integrated this project into my DT coursework. The contextual challenge I received was perfect:
Contextual challenge: How can products be used to encourage participation by students in STEM activities or clubs at school?
So I was technically fixated because if the contextual challenge wasn't this, I would've justified my way through and made it fit. But it did fit perfectly in this case, and I started brainstorming immediately for different products and decided that this was the most suitable. With that, I started designing right away as I only had less than a year to finish everything up, and this was with the pressure of other stuff as well. So here comes my first mistake: not planning sufficiently. I was juggling between a small form factor, low-cost voron 0-style printer and a tool-changer style printer, which would've been 5-10 times more expensive. See the problem? Yes, the problem here is that I am not fixated enough on scale, whilst being extremely fixated towards wanting to build a 3D printer. So I wasted quite some time deciding, and when I decided to go with the cheaper option due to time and complexity constraints, I didn't have much time to plan. So, I started with the CAD straightaway and designed the printer after scrolling through web pages and selecting a few components that are more commonly found. I wanted to combine voron 0 with the Rook. I liked the Rook for its simplicity, but wanted it to be enclosed and have an open front like voron 0. But here comes my brain again: “You wanna make this modular(small parts, no tapping needed, easy sourcing), so people without a big printer(or tools) can build it as well”. I have to say, the idea was good, but the execution was not.
So, after going through the design, I decided that it was good enough and started production. At that time, I was even thinking of designing cable tie channels for cable management. But soon, I realised how much “thought” I had ignored. Mistake two: not listing out the features and specifications of the product and checking if all criteria have been met. If I had done this project the traditional GCSE way, thinking like a product designer, starting with drawing, analysing and repeating before going into CAD, I might've avoided quite a few of these mistakes. But it would've taken me way longer than one year to do this, and the thing is, I have other subjects to study as well and do not have the time. The reason I said I could've avoided this mistake is that a lot of things are not listed in the specification. Who would've thought that you needed a limit switch for the Z axis? (Hmm, couldn't be me.)
As I started development and production, I realised issues and had to come up with solutions for them. Some of them were surprisingly good, but others weren't as robust as I had hoped. And when I was assembling the frame, I realised something bad. Real bad. The linear rods for the Z-axis were pushed in as you moved the bed.
AHHHH! Materials can shrink; I knew about this, but I didn't think 2% could make such a big difference. This is what happens when you think like a carpenter and not an engineer. I tried to square the frame again and again, printing a few of the bed frames, but there was nothing I could do.
That is the biggest issue with the build. In the end, I was advised to move on or else I wouldn't be able to finish the project in time.
I finished up the build, designed the enclosure and z-limit switch at the last minute, and booted Klipper. And tell you what, Klipper was a PITA. The SD for RPI corrupted, then it was Klipper hating me for 2 days, and then it was me hating on Klipper. But after all that, we became friends. Real. I was fine-tuning Klipper at 4 a.m. before the deadline. And guess what, my classmates didn't finish their coursework and had a good night's sleep, so the deadline got pushed back. Yay, more time to discover problems, and I wasted my liver's HP for no reason.
And after testing, as you would already know from the top. It was able to print, but the print quality was far from perfect. I am a little sad because, unlike everyone else in my cohort, I spent most of the time making the product itself better, not the coursework slides/presentation, so even if my product was more complex than everyone else's and even functioned, I didn't get the highest grade. You can say that I am dumb for not focusing on the slides since that's what's graded, not your physical product, so even if your product didn't work, the examiner/moderator wouldn't know or care. However, I believe that in the role of a product designer, we should prioritise the product and the user more than anything else, so no point in making good slides but a product that doesn't work. More importantly, I did this project to challenge myself, not to please the education system.
The moral of the story, plan properly and blame everything on ABS when things don't work out. Okay, just kidding. I'll list the issues with this printer and sum up a list of advice. But please don't give up on your project, it'll work out one day. (Maybe that's why, like everyone else, I have a bunch of unfinished projects… but hey, at least I didn't give up… not yet at least)
Issues
- X-carriage is rubbing against the enclosure panels. It was colliding, but after modifying the part, it's only touching. (still a problem)
- Z-axis movement is not smooth. As said, this is due to Bed Frame shrinkage. (The bed frame is the part holding the bed and linear bearings)
- The enclosure door is moving too much, this is totally my fault. I got lazy with designing, so I just took someone's double-hinge design off Thingiverse. Should've used a single hinge. (Completely avoidable and not an issue if anyone is building a new one)
- Tolerance might've been an issue; printed parts connecting aluminium extrusion are given 0.2mm tolerance as they are powder-coated and not anodised. This might've caused an issue, and maybe no tolerance would've made things more accurate.
Advice (TL;DR)
- If you have decided to design your printer, please remember to plan sufficiently. Draw out your design and ideas.
- List out everything you need (requirements definition) so that you don't miss out on parts when designing, even on obvious and simple parts like the limit switch holder. My print head was a last-minute design because I wasn't sure of which part to use, but in the end, due to time constraints, I designed an extremely ugly one.
- Take advice from others, whether they are experts or not. I would ask my DT teachers and peers for advice. Even though they have no expert knowledge of 3D printing, they would have their own experience in problem-solving. By communicating my issue to them, they can suggest solutions, and more importantly, I get to visualise the problem better for myself. Often, the solutions they provide won't work, but they can inspire you to come up with one that does.
- When encountering problems mid-build, try coming up with a suitable alternative; don't fixate on your original idea. I was gonna do a cut-out in the acrylic panel so that the limit switch fits flush and no wire would be seen, but I have adapted and installed it on the 2020 extrusion with no problems.
- When designing something like this, start simple; the printer bed is a good starting point. Don't start with something overly specific or insignificant, like an enclosure panel or a joint between two extrusions. Get a general dimension set first.
Although I have been 3D printing for quite some years now, I am by no means an accredited product designer or engineer. But I am happy to be able to share my advice with anyone who is designing, so if you disagree with anything, please let me know, as I am here to learn as well.
For BOM and Description, go to GitHub.
Thanks for reading, and happy printing!