[Williambeuk]

I'm planning on building a diy cnc router for use primarily on wood and have made a cad design. I would appreciate any suggestions on how I could improve the design or any considerations that I might not have thought about. The constraints to keep in mind is I am intending this to be a fairly cheap build, and I already have some components (round 500 and 200mm rails, and leadscrews, nema 17 motors, scs8uu slide bushings, pc power supply, arduino + ramps, etc.) so upgrading to better parts like ballscrews and large solid aluminum plates isn't in my price range right now.

Most of the machine will use aluminum extrusions 30x60, and 30x30, with 6.35mm 6061 aluminum plates as well. I'm also considering mounting the whole thing on a concrete base to reduce vibrations and make it more solid. In the cad design I have a dremel in a 3d printed mount, I will change to a more powerful spindle later, but I intended to first test it using a dremel 300 that I have.

The Y-Axis gantry, Z-axis, and dremel combined will weigh 13lbs. The whole machine will weigh about 28 lbs.

Full machine:

Side view:

Z-Axis:

Y-Axis Gantry:

Base/X-Axis:


Thanks!

[ger21]

That gantry will be very weak, where it attaches to the bottom crossmember, and the hole gantry will be somewhat unstable.

[awerby]

I doubt those NEMA 17 motors are going to be strong enough to push all that weight around, not to mention overcome cutting forces. The tilted gantry would make more sense if it had a vertical crossbrace, but as it is, Ger's right - it's adding weakness, not strength. Those unsupported round rails won't help with that at all - they will vibrate like guitar strings. And you'll just kill your Dremel tool if you try using it as a spindle. If you want to test the machine, mount a Sharpie marker and try drawing with it. If it passes the test, a trim router is the lightest thing that will serve as a spindle and last for an appreciable amount of time.

[cyclestart]

scs8uu slide bushings, so 8mm round rails. Good enough for a flat bad scanner but not for supporting weight over any kind of span.
Sorry but that's how I see it.

[ger21]

I missed the unsupported 8mm bearings. If you were to push on the spindle with 1 finger, you might see up to an inch of movement. That's how flexible those 8mm shafts will be.

[fretman_2]

Wow...you are going to have a lot of flex in your gantry supports. The unsupported rail is going to be a big issue too. In my opinion, this design is better suited as a 3D printer. You can make a few design changes and the machine would be much more rigid.

[Williambeuk]

What do you mean by vertical crossbrace? There is already a 3060 beam going across the gantry parallel to the rails, is this not a suitable crossbrace or are you speaking of a different section? Thanks.

[awerby]

That beam parallel to the rails would be horizontal (unless you're mounting your machine to a wall). A vertical one would go from the top of your gantry to the base where it seems to be riding on a single truck. Adding another one on each side coplanar with that truck would help with stability.

[Williambeuk]

Ah I see, thanks.

If I was to use wheels with the extrusion like those on openbuild, could I combine them with the 8mm linear rail to get the best of both worlds, the wheels supporting more weight and the rail keeping things straight? Or do they not combine well?

[wizard]

Just a few cell phone observations.

1. If you want inexpensive and as stiff as possible keep the mechanical design simple.
2. Aluminum extrusions are not a cheap path to a rigid frame.
3. Unsupported linear rails are for the most part a very bad idea. The smaller the rail diameter the worse the problem.
4. NEMA 17 motors are on the lite side for most routers.
5. As already mentioned, Dremels make terrible spindles! Since you are looking at small steppers I’m assuming a lite duty machine or engraving but even so Dremels die fast on a CNC. Sooner or later you will want an upgrade and that is where the NEMA 17’s could become a problem.
6. If engraving with very small cutters, runout on Dremels and cheap trim routers becomes a significant issue.

[skrubol]

8mm unsupported rails are completely useless in a router build that's not tiny (like < 100mm travels.) They're an order of magnitude too flexible for a decent router of this size. Probably close to 2 orders of magnitude too flexible if you wanted to cut metal. Also, even if you had 25mm rails (about 30x stiffer,) you would want 2 carriages per rail or at least long carriages, as the twisting forces (like the off-center gantry on the lowest rails,) will hurt accuracy and wear out the slides very quickly.

[ger21]

and the rail keeping things straight?

They're not strong enough to keep anything straight. Even at 20mm they'd have a fair amount of flex.

[Williambeuk]

[...]
2. Aluminum extrusions are not a cheap path to a rigid frame.
[...]

What would you recommend as a cheaper option?

[awerby]

Welded steel?

[wizard]

What would you recommend as a cheaper option?

Well Steel was very cheap relative to aluminum. You can weld it together or implement a bolted together design.

However in the case of this design you could go with wood (wood products) framed machine and fully supported rails and end up with a better machine. There are a number of examples in this forum some better than others. One of the more interesting machines is built with bamboo laminates.

In any case, sticking with Aluminum or going to another material the goal is a mechanical arrangement that results in a stiff machine. The goal should be large box sections, assembled in to a well supported frame.

Now all of that stated Aluminum Extrusion does have some advantages. For one you can buy kits pre cut, of a proven design. This is great for people starting out with no shop tooling to speak of. You basically end up with an erector set like collection of parts. However that convenience comes at a rather high cost relative to using raw materials yourself.

Perhaps the most important thing is that whatever you build is suitable for what you intend to use the machine for. This can be tricky as people often underestimate what they will want to use the machine for. It is real easy to build yourself a machine that doesn't do what you need while being perfect for somebody else. I'd nail down your needs before going too far.