Building a 3D Printer
Written by Sebastian Dümcke on
Tags: 3dprinting
My first printer was a cheap CTC i3 clone, that I bought around 100 Euros a few years back. It taught me a lot about 3D printing, but was also a constant source of frustration, especially regarding bed levelling and Z-banding. 18 months in, the heater cartridge of the hotend burned through the control board. For the replacement printer, I chose an Ender 3. It made quite a bit of difference: more reliable, less fidgeting.
Apart from the control board and the hotend I still have all parts from the first printer. Time to turn it into something new.
Build considerations
Here are some consideration: 1. No more spring loaded heatbed. I want the software to take care of levelling. So I will require a 3DTouch or similar bed level sensor 2. Less noise: I want silent motor drivers (and perhaps fans) 3. Direct drive extruder. Since the Ender3 is Bowden, it would be good the have a direct drive for e.g flexible filament 4. all metal hotend for a wider variety of filaments (e.g. PETG)
Another consideration was 12 or 24 Volts. Since the leftover power supply from the CTC i3 is rated 12V, 20A, I will stick to 12V. That will mean longer times to heat up nozzle and bed. Since I do not plan do have dual extrusion at this point a single E motor is sufficient.
Electronics
In the end I ordered the following items from AliExpress, which should arrive in 1-2 months.
- SKR mini E3 board: TMC2209 silent driver, 1 X-axis, 1 Y-axis, 2 Z-axes motors and 1 Extruder motor, support for bed levelling sensor and option to drive Neopixel strip. Only 2 fans. Also some interesting add-ons for the future: relay board to turn off printer and filament run-out sensor. This is certainly a budget option. If I wanted to spend (a lot) more, I would go for a Duet Wifi.
- a BMG dual drive extruder and 12V all-metal E3D v6 hotend clone
- a 3Dtouch bed levelling probe
Motors are from my previous i3 clone, as well as the PSU and mechanical endstops. If possible I will reuse the belts, plastic gears and idlers as well. Total cost: around 90 Euros.
Frame
Next thing to decide is the frame. Here I tried to find a list of all DIY printer designs out there but could not find a comprehensive list. So I will compile my own list here. Here are the main design types along with examples of builds
i3 design
Improvements on the popular open-source i3 design
- All metal anet A8: AM8 and BLV mgn12
- Franken i3
- i3RS32
- OneZ, TrueUp and Vulkan Bot : i3 design with a single z-axis motor and gearing to ensure 0.005mm/step on the z-axis and synchronized axes. Also of interest in the TrueUp is the idea of using nylon bushings inside the rail for the linear motion in Z.
Core XY design
This is a very elegant design, where the head moves along the X/Y axis instead of the bed. The bed only moves downward. This allows for greater print speeds and less issues with layer shifts. Most of the builds have many features and are quite costly.
Other design types
- Crane design: Indie i2, Bad Robin
- Voxel Ox
- a belted printer for infinite z-axis: The White Knight
- linear Rail printer
- Delta printer: Kossel, Rostock and co. I leave these out since I am currently not interested in delta printers
- hangprinter
My frame considerations
Axes
I see more and more builds going towards MGN12H linear rails to improve the linear motion compared to linear bearings on metal rods. I want to give this a try as well. The question remains if all axes should be equipped with rails or just some of them. This is also a cost consideration.
For the Z-axis most builds use lead screws but my CTC i3 was using threaded rods. I want to try combining linear rails and threaded rods. Since the rails will constrain the movement along the same linear axis, I should get better resolution with threaded rods (less advance per pitch) and save on money (plus these are easy to replace). I have not seen any other build doing this.
Build volume
The heatbed left over from the CTC i3 measures 215x215. So this should be the minimum XY volume.
I mostly print flat parts. Maybe restrict the height to a minimum (150mm or100mm)? Would that make the printer more stable and reliable? It would make it a lot simpler to ensure that the threaded rods are straight on a shorter distance.
I will decide on the frame build in the next article from the series.
Revision history:
- 2020-09-05: add more printer designs and conclude article
- 2020-09-14: add more coreXY designs and link to second part from series. Edited for clarity