Modretro Public 3D-Printer -- Interest Check

[rebreok]

Hi all. I've been working recently on a couple different projects and keep hitting much the same roadblock for each one, which is manufacturing troubles. Because of this, I have decided that the project I must complete before any other will be a 3D printer.

I have a couple of prototypes going already, and I'm working with some college buddies on programming the microcontroller. It's still in the early phases right now, but I think it'll really pick up and go once everyone starts pitching in.

The goal here is actually to produce a 3D printer design which can be easily replicated in any home with the most basic tools and a shoestring budget. With the help of Modretro however, I intend to create a design that is an easy, cheap, and fast build, that quickly evolves itself into a high-quality printer of modular size. Secondarily, I intend to open mine to the public, selling parts for nothing more than technical costs and making a donation-based profit, if any.

The way it would work is there would be a single devoted thread, where the main team (I) post the prototype specifications, needed changes, and existing code, along with any suggestions, observations, etc. Then anyone who wants can work from the existing, post their data, or even complete divergences from the existing path. It matters little just how we do it, I just want to see it done!

If you think this is a good idea, or better yet if you'd like to help, leave a reply and any thoughts.


Looking forward to working to together,
K

 

[vskid3]

I think its a good idea, I just question whether or not you'd be duplicating something that already exists. No reason to reinvent the wheel. However, if your idea is better than existing 3D printers, go for it.

 

[Abyss]

This project sounds really promising. Definitely keeping an eye on this!

 

[laingsoft]

Ponoko offers a service like this, and they are pretty cheap.

In case anyone wants to get something done.

 

[Zero]

Re:

This project sounds really promising. Definitely keeping an eye on this!

 

[DuHasst0]

This sounds very interesting and would love to contribute.

 

[rebreok]

Alright then. Here's a quick rundown of the major points:

-controlled only by an Arduino. No shields, no add-ons of any kind, only basic circuitry (switches, transistors, etc.)
-must use only a normal, DC motor, the kind you could pull out of any old appliance. No steppers.
-$100 budget, at the most. Ideally we should aim lower.
-Any parts that need to be custom made, like the frame, should be able to be cut from any chunk of wood with basic tools. This means no laser cutting, no plasma cutting, just what can be done with a circular saw, jigsaw, a drill, and some sand paper.

What you should say to yourself is, "how can I build this without paying for anything?" After the fact, what I expect to happen is one will make their own basic framework, then 3D-print better parts to gradually increase the precision of the machine. So far, my minimal-budget mechanism ideas are as follows:

Stepper Motor Alternative

so I was thinking about how I could just make a really nice, small step motor with nothing but a DC motor and a switch, and something something got this idea. The green part is a VEX lemit switch, and the blue part is one you'd have to make yourself. Now I know what you're thinking: "how the flagship am I supposed to make something that small and precise?!" Weeellllll, it just so happens that when I made the part, I discovered that you can put in all the necessary constraints, but still make it as big as you like. Just pick a radius, print a profile, staple to your plywood and you're in business.

Plate Rests

This one's a little more intimidating, but I assure you it can be done. I imagine one would first cut the top and bottom out of 1/2" plywood, then glue them together. The hole can just be an interference fit for a 7/16 nut, plus you could fill the edges with hot glue or something. (not in the tool list, I know, but still common enough.) Both of these design use 1/4" allthread, and you can get more than enough of that on Mcmaster-Carr. (6ft for $4) This does mean that leveling your plate would mean individually spinning each corner around and around until it's just so. Also I realize there are a few parts that I left open-ended here, like how to attach an encoder wheel to allthread, but I think you're all clever enough for that. So by my calculations, if one uses 1/4-20 allthread and the 8-lump encoder wheel shown here, you can get a precision of 1/320th of an inch, or just under 80 microns. For comparison, a MakerBot Replicator2 goes down to 20 microns--a tolerance that we can beat with a simple 3/13 gear train, printed from the initial machine. Obviously there are plenty of other ways to do it, but that's the one that came to mind first.


I have found myself to have bitten off a bit more than I can chew trying to convert stl files to a step-code that this design could use, so I'm thinking of using Slic3r or one of the many existing slicing engines out there, then working from their .gcode. That'll be my next contribution, though it may not come for quite a while. Until then, I'm sure everyone can keep busy with this little springboard here. Good luck.

K

 

[ttsgeb]

You do a magnificent job of making this look less intimidating, though.
I approve of the position detect thing.

To make it easier on yourself, try breaking it up into stages. Don't think printer, think cnc table. Get that working, then worry about the print head.

write me table code already.

 

[vskid3]

Have you considered using optical encoder wheels like ball mice and N64 joysticks for rotation sensing? The switch idea seems like a point of failure and doesn't seem very precise. An encoding wheel placed on the motor shaft before the gearing would have a much higher resolution.

Are steppers really that big of a deal? It seems like they're just as or even more common in electronics you would salvage from. If you do use regular motors, it would probably be much more reproducible if you use motors that are commercially available and have salvage ones an option for people to hack in.

 

[rebreok]

Have you considered using optical encoder wheels like ball mice and N64 joysticks for rotation sensing? The switch idea seems like a point of failure and doesn't seem very precise.

Yes, but these things aren't quite so common, and require purchasing/ordering online. The bump wheel can be made from scrap wood, and just about any limit switch can be used with it.

An encoding wheel placed on the motor shaft before the gearing would have a much higher resolution.

I'm aware, however it would only give you the position of the motor, not the allthread, and that's the important part. Technically it would be no different, but all of this is done, again, under the assumption that you are using parts hand cut from cheap wood. Gears like that might not have the precision needed.

Are steppers really that big of a deal? It seems like they're just as or even more common in electronics you would salvage from. If you do use regular motors, it would probably be much more reproducible if you use motors that are commercially available and have salvage ones an option for people to hack in.

Personally, I've never found a stepper in any product, except one that came from a disc drive, and it would be far too small. And what's important for this is to alter the things most easily changed, the home mode parts like gears and input power. It's a lot simpler to lower the voltage before a motor controller than it is to order a new motor.

 

[vskid3]

An encoding wheel placed on the motor shaft before the gearing would have a much higher resolution.

I'm aware, however it would only give you the position of the motor, not the allthread, and that's the important part. Technically it would be no different, but all of this is done, again, under the assumption that you are using parts hand cut from cheap wood. Gears like that might not have the precision needed.

What I'm getting from that statement is that you're not going to have a gearbox, just the motor connected directly to the allthread. Is that correct?

I realize I may seem like I'm trying to shoot you down, which I'm not. I'm just trying to add some constructive criticism.

 

[rebreok]

What I'm getting from that statement is that you're not going to have a gearbox, just the motor connected directly to the allthread. Is that correct?

I realize I may seem like I'm trying to shoot you down, which I'm not. I'm just trying to add some constructive criticism.

For the first iteration, until you can start printing more precise parts, yes. After that, it shouldn't be an issue. And you're fine, it's good to cover all the bases of why A and not B, that's why I asked for input.

 

[rebreok]

So I made a lot of progress on this project before summer started but lost the notebook containing it. Now I am part of the Minio team, in charge of tutorials, and soon I'll be working on their portable 3D printer project as well, combining my cleverness with their teknowledgie. I think it is still open source, so this shouldn't hinder any current progress. Once they set up a forum I'll be sure to make a 3D printer thread and post a link for it. In the meantime you should go see their website here and my description of minio here.