Building a 3D printer from cheap components: Part 2
This is part 2 in an unbounded series of articles describing the adventure that is building your own 3D printer from scratch.
Taking incredible amounts of inspiration from Tom Sanladerer’s Dolly project: A Prusa i3 mk2 clone YouTube series, I am continuing to journal my progress loosely following Tom’s directions.
It’s been around 2 and half weeks since I last made an article and in that time I have received a few more components from Banggood and Aliexpress.
Organisational approach
It turns out there are a lot of parts to keep track of and I decided to create a spreadsheet to help myself out.
I’ve been able to track just how cheap this 3D printer is actually going to be and so far the amount of money I’ve spent on all the parts has come to a whopping total of £204.70.
Currently a Prusa i3 Mk3 will go for £699, so I’ve spent 30% of what I would have spent buying a printer brand new, though I should mention, the printer I am making is a Mk2.
I have also started printing the parts that will be used on the printer but made a little mistake (more on that later.)
New arrivals
The RoyalMail has been taking one for the team recently. Since last time, I have taken receipt of:
- LM8SUU Bearings
- The wire bundle (kind of…)
- Power supply unit
- LCD Panel (almost…)
- 3 PIN IEC320 connector
- Inductive proximity sensor Z endstop
- 5 meters of teethed belt
So a fair few interesting things, but the most interesting is the PSU.
Having a functional PSU allows me to start testing things like motor movement, heating the hot-end and also opens up issues like cooling.
Minor setbacks
It wouldn’t be right to ignore some of the cock ups I’ve made.
The wire bundle I purchased was meant to simplify connecting all the components of the printer to the RAMPS board. However… the cables that I purchased, although advertised as for 3D printers, had a slight issue when it came to the cables that would be used for the stepper motors. Stepper motors have a JST 6 pin connector, but the RAMPS board (and at least the CR10S) use a 4 pin connector. The cables I bought came with a 4 pin to 4 pin connector meaning they won’t fit into the steppers (boo!).
The LCD panel I purchased is exactly what I need, but I cannot connect it to the RAMPS board yet, as I am missing the breakout board that will make connecting much simpler (double boo!).
Beyond what I have received not being exactly what I anticipated, the other minor setback I have experienced is to do with metal rods used to create the frame.
There is a requirement for 12 rods in total:
- 2 on the left side of the base
- 2 on the right side of the base
- 2 on the front (with 1 being a 5M threaded rod)
- 2 on the back (with 1 being a 5M threaded rod)
- 2 for the Z axis motion
- 2 for the X axis motion
Each are in a varying size (330mm, 350mm and 370mm). I’ve currently got 4 rods and I have ordered another 4 leaving. As I have Z motors that come with threaded rods connected, I only need to get two 5M threaded rods for the last part of the base.
3D Printed Parts
As there is plenty of time to get busy printing the parts that make up the printer before all the ordered items arrive, I started printing out the parts from the Prusa Github repository, specifically those listed in Tom’s Dolly guide.
I did however, make a slight mistake whilst carrying out this process.
The parts that hold anything that heats up, really shouldn’t be printed in PLA.
PLA as a material has some really beneficial qualities; mainly that it is strong.
However, PLA suffers from a few drawbacks that will end up causing me issues in the long run.
One of the issues is that PLA enters it’s glass transition phase at a relatively low temperature, making it malleable.
The other issue with PLA is that it will deform over time when under constant pressure. This could be an issue for some of the parts I have printed, but the main issue is definitely one of temperature.
Stepper motors can reach temperatures of 100 degrees Celsius during operation, but, PLA enters its glass transition phase at around 60–65 degrees Celsius.
This means that I need to print any parts in close contact with a heat source (the hot-end or a stepper motor for example) in a different material.
Hot material choices
The obvious choice in material would be ABS. ABS enters its glass transition phase at 105 degrees Celsius which makes it a slam dunk for the application at hand.
The downside though is that to print ABS, you need an enclosure for your printer to prevent warping due to temperature changes, but, the main issue is the many reported safety concerns with regards to vapours that are released whilst it is printed.
Avoiding ABS on the grounds that I want to maintain my health might be a bit of a poor excuse, given there are things you can do to print it safely, but, it turns out that Prusa recommend printing all their parts in another material.
Enter PETG filament
PETG is a filament that sits between PLA and ABS in terms of its properties and ease to print.
With a glass transition temperature of around 80 degrees Celsius, it promises to handle temperatures substantially better than PLA.
I’ll re-print all the motor mounts and hot-end holder using PETG following CNC Kitchen’s tips for calibrating my printer to get the best out of PETG.
One other side benefit of PETG is that it’s hydrophobic and some filaments are even advertised as food safe!
Testing the parts that I currently have
Now that I have a PSU, I can test the motors and the hot-end!
I eagerly connected the IEC connector to the PSU using a couple of paddle connectors I had laying around from 10 years ago (wow!) and hooked up the PSU output to the RAMPS power input.
I did some quick research to figure out what power output from the RAMPS the hot-end would use (it’s D10 for the RAMPS 1.4 board) and decided to fire up Pronterface and set the temperature to 195 degrees Celsius to see if the hot-end lived up to its name.
The result speaks for itself! Although the photo is blurry, you can see the hot-end heated PLA sufficiently for it to come out of the nozzle uniformly.
Great start!
How about some motor movement then? Well, looking back at the first article, the photo of the RAMPS board and the ATMEGA2560 board show that I eagerly connected all the stepper motor drivers.
When I used Pronterface to move the X, Y or Z motors, the result I received was that of a stepper motor vibrating whilst not actually turning. Initially I thought I may have gotten a duff stepper motor driver, in fact, most of the first hits online seemed to point at duff hardware. The reports of bad stepper motor drivers however were always coupled with overheating.
Troubleshooting the issue then became a game of gathering potential problems I could be facing an ordering them in a list of probability. It was definitely possible that the stepper motors were bad, but they definitely hadn’t overheated, not during my ownership at least.
The motors themselves definitely worked, as plugging them into my CR10S and controlling them functioned perfectly for all the motors involved.
This is when I started to realise it could be some really simple.
After looking up the RepRap RAMPS assembly guide, I realised I had put to one side a bag of jumpers that came with the board. It turns out, as it is possible to use either 16 teeth (16T) or 20 teeth (20T) gears on the stepper motors, you must configure the board for which setup you have chosen.
After jamming all the jumpers in the places they needed to go on the RAMPS, I fired up Pronterface once more and attempted to move the Z axis…
The motors work and are controllable!
I’m just missing a heated bed and I’ve got a deconstructed 3D printer!
Next steps
Whilst waiting for the remaining parts to arrive I’m going to be investigating the best way to connect fans to the RAMPS board so that I can run air over the board and over the PSU to make sure the system has the best chance to not overheat.
I’m also going to print out the motor mounts and hot-end holder in PETG and also explore enclosures for the RAMPS board itself and the PSU. The PSU currently is far too exposed for my liking. I cannot overstate how dangerous it is having mains voltage exposed as much as it currently is, so safety is paramount!
When part 3 comes out, I’ll link it here! If you got this far, thanks for reading.
