Build a 3D Printer

Building a 3D Printer: Goals

So the first part of the build (the 3D printing of the structural parts) is now complete. So that’s good. I’ve been going through the BOM (bill of materials) and starting to figure out the stuff I need to order and what kind of timeline that’s going to exist on (budget is monthly, not up-front). So far, building a 3d printer has been a pretty mundane experience — the fun begins when I start getting everything put together. But in my last installment, I promised I’d expound on the goals of this project and what I’m hoping to get out of it, so without further ado…

Project Goals

For overall performance, I’m wanting to exceed the print volume and equal the resolution of a Stratasys uPrint SE Plus. I also don’t want to spend a ton of money, either. So my goals are:

  1. Spend less than $500.
  2. Get a build envelope larger than a Stratasys uPrint SE Plus (8″ wide x 8″ long x 6″ high).
  3. “Reasonable” z-layer thickness.
  4. Ideally, I’d like to be able to print with ABS, PLA, and nylon.
  5. Learn a lot so I can apply that knowledge to an even cooler follow-on project. (Future plans for building another 3D printer will be discussed in the final installment of this series.)

So Far

  • Cost: $0. (Have not started ordering from the BOM yet. Printed parts cost me nothing.)
  • Build envelope exceeds baseline goal (384 cubic inches). Final build envelope will be 1179 cubic inches, which is 307% of goal. Rawk.
  • Layer height — the stepper motor is a 200-step setup tied to a threaded rod that’s got an 8mm lead. That means a single step will move the build plate 0.04mm (+/- %5), or about .0015 inches. That’s almost an order of magnitude better than my baseline printer. DISCLAIMER: I may be talking out of my ass on this one — this is based on mechanical movement of the plate. I am uncertain as to how the z-layer thickness is determined, and that may be a function of the extruder head.

Establishing a Performance Baseline

To make sure I have a fair point of comparison, I decided to go ahead and select three items off of Thingiverse, and to go ahead and print them. Those three items are a set of 5mm calibration steps, a spirograph kit, and a Penrose triangle. Each was printed with a different amount of infill (as detailed below), and with a layer height of .01 inches.

 Time (minutes)ABS (inches3)Support (inches3)ABS Volume/Rate
5mm Calibration Steps40.22.07.005 in3/minute
Penrose Triangle134.75.36.005 in3/minute
Spirograph1361.41.05.01 in3/minute
Benchy177.97.44.005 in3/minute
Regarding the baseline objects, here's how they were printed:

5mm Cubes: .01" layer height, high density fill
Penrose Triangle: .01" layer height, low density fill
Spirograph: .01" layer height, solid fill
Benchy: .01" layer height, solid fill

(More information here, covering performance, things like support material, etc.)


Building a 3D printer is an interesting process — you might not discover some things until you’ve already committed to a build. As it turns out, the system is designed around printing with PLA. The heated bed can only maintain 60°C, which isn’t hot enough for ABS plastic. I wanted to be able to do ABS, but I think that in the interest of budget, the plan is going to be starting with PLA, and then making a future upgrade to a hotter bed (and possibly moving it into an enclosure so I can control the heat of the atmosphere around the print).

Time, Cost, and Materials

As mentioned elsewhere in this post, thus far, all that has been done is the 3D printing of the parts required to build this monstrosity. For now, though, let’s look at some interesting stuff with the printing:

 Time (minutes)ABS, in3Support, in3Cost
Cable Chain20104.353.41$0
Box for RAMPS 1.4
Pack 1241510.223.27$0
Pack 2366912.056.14$0
Pack 315345.891.98$0
Pack 4254611.233.15$0
Pack 510586.411.75$0
Pack 65073.560.99$0
* All materials printed with a Stratasys Dimension Elite at .01" layer height, with solid infill. Exception is the "Optionals" pack, which was printed on a Stratasys uPrint SE Plus at .01" layer height with solid infill.

So that works out to 243.6 hours of printer time — it could have gone faster had I not used solid infill for the parts, but I wanted to ensure that I was going to have something super-durable.

The amount of ABS plastic I used for this is roughly equivalent to 3/5ths of an ostrich egg, or 1.7 venti-sized drinks at Starbucks.

Katee Sac
No, not this Starbuck.

What’s Next?

The BOM. Got to start ordering all those physical things that I can’t print, like V-slot extrusion, nuts, bolts, etc. I’m going to do some legwork on the new Arduino board I need for this and the RAMPS 1.4 board — doing so in such a way that I keep costs down.

Building a 3D printer is time-consuming, y’all.

Image Credits: mebner1/.

Let's Discuss...