The bootstrap CNC uses rails like that. I did try that for a few experiments, but ended up not liking it much for CNC use
My experiment that way ended up using printed parts to hold the bearings instead of drilled angle, because I had a bit more control that way, but the tolerances ended up being too picky, and making the rail ended up being a sticking point. I wanted to count on the tablesaw to make the rails straight, but they weren’t straight enough to keep play out of the system. 1/64" was too much out of square over 6’ and I ended up with an overcomplicated adjustment rig to keep the rails parallel. In the end I scrapped it all. I know other people have built it and it works, but maybe they weren’t as picky as me for tolerances, or maybe they were making shorter rails, or a combination.
The only issue I see with that is hurt feelings. How in the world do we put up one or two ideas, but not all of them or what happens when we remove one.
Let’s use 4th axis for example. There are a few designs out there. They all have strengths and weaknesses. So is the best way to do a little write up “here is what a 4th axis allows you to do, here are a few ways to do that, and then what? Do we add some links or feature one?” and do we do it in the docs and then some sort of list somewhere, might get to be a bit of a long list?
So I was going to test this sort of thing. The main issue I see is trying to align the linear contact patch of a roller bearing (many bearings) and a flat bar stock seems like a recipe for failure. Now change that to some wheels point contact patch and misalignment seems like less of a issue. Same principle as a round bar and bearings. Flat on flat can steer, point of flat has no steering.
Or are you saying aligning the rail was an issue? If that is the case, were you aligning one rail or more than one?
For what it’s worth, I put together a set of those “bearings on angle” rollers for doing a bootstrap CNC years ago, and they sat in a drawer waiting for me to get to the next step for a looooong time.
Those are the bearings now running in my Burly MPCNC.
That is where I see some sort of peer review/rating system comes in. If there are 4 implementations of a fourth axis and one has a 200 thumbs up from users and the others have 7 thumbs up each then you know how to sort/filter them in the catalogue. The goal would be to make it 100% hands off for Ryan (or others updating the documentation) for the “User Mods”. The catalogue would simply show like a report on the user provided data. The only time Ryan would need to get involved would be if he wants to endorse or feature a mod.
I was trying to think of a way to make this as easy as possible to build for the V1 website. I came up with an idea that may be garbage or gold. Lets say we use a website like prusaprinters.org to hold all the models, images, instructions, comments etc… Then on each of those listings that users want to have added to the V1 User Mods Catalogue. They need to add a specific Tag (search key word). Then the V1 catalogue would just need to do a live search of the website for anything with that tag and show the results sorted by the number of hearts that model has received on that website.
For example. If a user opens the catalogue and clicks on MPCNC - Burly then it would do a live search of all the models on prusaprinters.org that have a tag of “MPCNCBurly”. Then it would show the results with a thumbnail and the 150 character summary of what it is. You can see the number of hearts a print has and sort by that.
This is also good for a mod that works for more than one version of the machine. If a specific mod works for more than one machine then the user just needs to add the additional search Tag. The catalogue could have instructions for users to submit their own mods.
The only thing this idea doesn’t do is have a way for V1 to endorse mods/upgrades. But I suppose that could be done by manually adding links from the documentation.
Very obvious problems with this idea would be if random people start adding random tags to unrelated models/pages. Or what if prusaprinters.org changes format etc.
Basically it’s all an issue. This is a stone b***h to get aligned smoothly.
The 4 wheels, if you attach them to angle, the drills for axles has to be super precise, or else there’s no way to make all 4 contact properly. I tried 3 and 3, and it wasn’t constrained enough to keep from tilting, and was really weak against twisting forces, which it would generate on its own.
I did finally manage to get 4 sets of those made by making a jig on my drill press to locate the holes at the same distance from the angle edge… Then made another jig to center punch them first. and still ruined at least 6 pieces of angle, and several drill bits.
Then trying to get a rail with adequate precision to keep contact with all 8 bearings while sliding, and I gave up in despair. If your material doesn’t keep perfectly tight to the table saw fence for the whole rip, you’ll get a loose spot, where the wheels will lose contact with the rail when it’s done. That’s with 1" by 1" by 1/8" steel angle.
The Bootstrap has some sort of pressure adjust to allow for some imprecision in the build, but it’s still assuming that the rails are perfectly parallel, more than the real-world material that I had actually was.
Yeah, I did get a set of the bearings that I was somewhat satisfied with, then tried to get it aligned to a rail, which is where I scrapped the bootstrap CNC project. I understand that people get those things made, but I couldn’t get the X or Y axes to my satisfaction (6’ rails each) I think that the X/Z part is still in the garage somewhere.
Cool, That is kind of exactly what I thought. I have a solution I want to try. I really want to show you smart people my sketches to filter out any dumb ideas I have up front. That feels like it could be a bad idea though.
Well I think you solved it. Directed search. Printables or Search Thingiverse - Thingiverse If someone wants to put their name in the hat for a mod all they have to do is include the key phase. In this case, “mpcnc needle cutter”. That would probably need to be more like “Mpcnc Primo needle cutter”
You are right though people could abuse the key phrase but liking projects would make the cream rise to the top.
I think dust buildup might be an issue with bearings on aluminum rails? The benefit of sliding rather than rolling would be that it pushes debris out of the way instead of rolling over it and lifting the z axis.
I wonder if this has actually been an issue for the home made guides/rails though
Count me on that as well. Ill test/R&D as well. Especially something to do with that aluminum angle stuff. I have some of that already and want to experiment/ build another machine.
The table is just a large workbench I seem to be using for other purposes too.
However, I often push against the GT2 belt because it´s on the side of the table.
Would it be an option to mount the Y stepper shaft horizontal instead of vertical, so it pushes the GT2 belt below the table instead? That way the belt is rarely touched?
So one could use the worktable while CNC is not in use?
Agreed, First I will try to get it really close to the side.The belt on the top can build up with junk, on the bottom fixes the required table thickness. But yes, I want to move the belt.
