The breakthrough achievement I mentioned in the prior post was the successful manufacture of a 3D printed pulley.
That happened this morning.
As I mentioned in the previous post, with the CNC and more recently with the 3D printer, there’s a very specific design ethos I have with what I’m doing – partly influenced by the world I’m bringing these machines into, manufacturing things for my business, and partly influenced by the world we all live in.
That ethos is as follows:
- Easy-to-Source – something that, where off-the-shelf parts were required, those parts were as few in number and as common as possible, to avoid the inevitable dreaded manufacturing and supply chain interruptions, and to enable easy redesign / substitution when and if necessary
- Easy-to-Build – something that, in being made up of a high percentage of user-printed parts, was designed in such a way that these parts could be printed from common, easy-to-print materials, and designed in such a way that even less-than-perfect(ly calibrated) printers could (still) yield usable parts
- Robust yet Brisk – something that (eschews the en vogue obsession with ultimate speed and lightness, something that) is strong and stable, perhaps at the expense of ultimate speed, but equipped with motors and tools with plenty of power overhead such that adequate speeds can still be reached
The beauty of this has been that not only am I able to more effectively dodge supply chain constraints, but I’m also able to create perfect custom parts – parts that allow a more cohesive and better design than having to create a machine around a bunch of off-the-shelf items.
Being able to create a part that fits in with everything else around it, instead of everything around it having to be designed to fit it – man that’s nice. Saves a whole lot of “back to the drawing board” redesign of stuff when you find out something isn’t available or won’t fit in a given area.
Pulleys and Idlers
One of the most important of parts on any CNC, 3D printer, or any similar machine are the parts involved in motion and movement. For belt-driven machines, such as the 3D printer I’ve been designing, this involves the pulleys and idlers.
Throughout the design process, I’ve spent a lot of time on Amazon, trying to design my machines around what I can get. While there are other places one can go such as AliExpress, and other 3D printing companies, no other single source has as quick of shipping or as wide a selection as Amazon.
The problem that I’ve encountered with sourcing pulleys and idlers not just from Amazon but anywhere really is tremendous inconsistencies.
Some of the listings have dimensions on them, but many do not. Of the ones that do, different ones have different dimensions. They’re all roughly the same, but one pulley will be .5mm higher or lower than another. Or one will read 12.65mm in diameter while another one is 12.75mm.
Not to mention very few of them have 8mm bore, 10mm wide pulleys. These ReliaBot 20 tooth 8mm bore 10mm wide pulleys were the only ones I could find. And unfortunately, no dimensions to know how tall they are, how high one belt will run above the other, what the clearances are. Not until I asked a question, and even then, still not all the info I needed. Anyway…
When you’re dimensioning things out and placing belt runs, running back to the product listing to double check a dimension, and then seeing one page saying “Only 17 left in stock, order soon!” and another saying “Arrives Prime by [a date somewhere between three to five weeks out]”, it’s a little unnerving. This is not good news if you have any hopes of putting together kits, sharing your plans with others, let alone building another one of these in the future.
This was something I encountered first designing the CNC, and I’ve met again with the 3D printer. It’s what inspired the first of my engineering ethos points – easy-to-source.
Printing Stuff
I decided very early on with the 3D printer to create printed pulleys and idlers instead of buying off-the-shelf. My rationale was that these are parts that are hard enough to find as it is, and could easily dry up. Doing this makes it easy to just make more of them if you need them in the future. The challenge, though, is in successfully printing them.
The resolution and accuracy of a 3D printer takes a lot of work to dial in, but it IS possible. These pulleys have very, very fine details to them, and if your printer is the least bit off, it can produce a part that looks good but just doesn’t match up well.
I’ve seen it done before though, so I figured if others have done it successfully, I can too. And luckily, that turned out to be the case.
Several months ago, I took my $400 Amazon Ender 5 Pro printer, installed a direct-drive MicroSwiss all metal hot end / extruder combo on it, did a number of tests, and improved my print quality considerably. It is possible to print really decent stuff with a cheap machine. It just takes more fiddling than some of the others, and had I not done this, I’m not sure I’d be writing this post right now.
Looking back in retrospect, I probably should’ve actually printed one of these pulleys out as a test a long time ago before designing a whole printer around them and facing a potential redesign to accept off-the-shelf parts if it didn’t work out, but there was this voice that said “just do it, you’ll figure out a way to make it work and get it all calibrated”.
Yesterday I shared some of what I’d been doing on one of the 3D printing Discord channels. I felt like there was some positive feedback to it – the printer overall. The commentary around 3D printed pulleys had some skepticism in it, specifically in my desire to manufacture instead of using off-the-shelf parts, and I understand and appreciate that.
There was some question about what would happen if the pulley fails. Truthfully, I don’t see them failing anytime soon – but time will tell. If they do, they’re super easy to replace, and I believe that printing them in PETG, ABS, Nylon, or PC could yield a very robust pulley. I’ve printed these with 100% infill, and due to the shape of them, the whole pulley is more or less a non-stop path of concentric rings of filament, which produces a rather strong part. That, coupled with the fact that there are four of them driving the XY axes (two for X, two for Y) means that they’re each pulling a half of the weight of the gantry in any given direction. Given that I’m using 10mm wide belt, and these are 30 tooth gears, means that there’s a lot of surface contact, spreading out the stress considerably.
I’m happy to report with the first print I did today, everything turned out marvelously. The pulley is designed to accept a pretty standard lock collar with a grub screw to lock onto the motor shaft and prevent rotational play. The grub screw that is in there right now and comes with them will be replaced with a slightly longer one to extend out and act as a key to maintain rotation. It is mounted on a standard 8mm hardened steel rod, which is driven by a NEMA 17 stepper motor at the bottom of the printer.
Next up – printing now actually – is the center cage, a custom part with rather tight tolerances, designed to have four MGN12H bearing blocks mounted inside, and to hold the extruder and hot end.
More soon!