[telstar505]

Hi!

I am a industrial designer in need of a CNC machine. I have looked around for machines to buy but decided to build one myself. It seems feasible given the vast amount of resources available to do so nowadays. And also fun! I have operated several 3-axis CNC machines in the past and have assembled a Shapeoko for a friend but still consider myself a novice in the area, that is why I decided to post here. I want to share with you the results of my research and hopefully get some feedback on how to proceed.

Machine requirements:

• It will mostly carve 3D shapes, occasionally 2D shapes.
• 5-Axis: It will have to mill undercuts on parts sized 150x150x70mm approx.
• 3-Axis cutting volume: 450x450x200mm approx.
• It is meant for fine detail work so it must be very precise with an emphasis on high quality surface finish. Repeatability precision must be at least 0,01mm, hopefully better.
• I mostly want the machine to mill PU foams(like Prolab65, or this), plastics like Nylon, ABS or Acetal and hardwoods.
• Minimum tool size: 0,2 mm
• Small budget of course

My thoughts so far on how to solve this are:

• In order to achieve these requirements I think the best alternative would be to base the design on the Tweakie CNC and customize it so an AB platform can sit on the machine table. Looking around this one seems like a good candidate. Has any of you tried a platform like that? The platform will have to be removed for larger 3-Axis milling, of course. I can have the frame built locally by a professional metal shop for a decent price.
• I made a quick schematic model for the machine in Inventor. I used the measurements provided by the seller for modeling the AB platform:
  
• The quotes I got for the frame uses 25x25x2mm steel square pipe. The design considers the addition of a plate that will be milled completely straight in order to have a proper mounting surface for the square rails to sit on. Also, oversize holes to allow for fine tuning the alignment of the rails. Do you think I should provide other means in the frame for calibrating the alignment and squareness of the X,Y and Z axis?
• I am thinking of getting these rail+ballnut screw and supports kit. If possible, I would love to know your thoughts on this kit, do you think they will be precise enough for my needs?
• As you can see in the picture I am planning on making an outrunner brushless motor for the spindle. No other reason other than I thought it would be fun to build and maybe can make it lighter that a spindle from Aliexpress. I was thinking on building something like this one, with conical bearings and other improvements. I am aware that I will need high RPMs(higher that 20000RPM) for the toolsize I want.
• My thinking is that NEMA 23 stepper motors will do fine. I have these Wantai 57BYGH218 in mind, they are 2Amps per winding and 1,8º step angle. Will I need 0,9º stepper motors? Do you think 2Amps motors will do for my application? I have used Wantai motors before and the quality is good but I am not an expert.
• Might add epoxy granite slabs to the frame for harmonics damping. In these areas(red):
  
• My plan for building this is to buy the components first to make sure all measurements are correct and then have the frame built. That way I can minimize fit issues between components.
• Toolchain that I plan to use is:
  - Toolpath and Gcode generation: Inventor HSM
  - Machine controller: Arduino Due(running g2core)+BigEasy Driver
  - Control Software: Chillipeppr

I would LOVE to get some feedback from you. Thanks!

[awerby]

A machine like that might be able to handle the PU foam, but the hard plastics and hardwoods won't have a smoothly cut surface. The frame design is too flimsy for that. Since it's a fixed bridge, you can build it much heavier than you've drawn it, and it will work better. I'd suggest steel tubes at least twice the dimensions, and twice as thick. The granite slabs would help, but all the steel members should be stronger as well, and braced in more directions. The "Tweakie" design Tweakie, Tweakie.CNC, CNC, router, mill, engraver, vinyl cutter, laser. is okay for what it is, but it's a very light-duty machine intended to be cheap and easy to build. As the person who designed it says: " The gantry style construction of Tweakie has large gaps between the point at which the cutter contacts the work and it’s supporting bearings, and as a result it’s capabilities are limited but it’s good point is that the working area is reasonably large." You've exacerbated its inherent problems by adding that trunnion table, which just adds more of those gaps and weak points. I'd suggest trying this out as a 3-axis machine first, and if its stability exceeds expectations, adding the 4th and 5th axes later. That will give you time to save up for the software you'll need to construct your 5-axis toolpaths.

It's difficult to get the surfaces of a welded steel structure flat enough to mount linear rails to. The Tweakie design avoids the issue by not supporting the rails at all, just stretching them over a few cross-beams. This is not the recommended way of doing it, since you don't want these rails to flex. If the ground plate you mention (but don't show in your drawing) goes under the table and the X axis rails mount on it, that would be better. You also need another one on the bridge to mount the Y-axis rails on. Making all the holes oversize does make it easier to adjust to squareness, but it also makes it easy for it to go out of adjustment. There's nothing more frustrating than realizing that the last batch of parts you made were out of square because something shifted.

I don't think you'll get anywhere near the .01mm repeatability you're talking about, but it's hard to believe that it would matter for the foam or even wooden parts you're making. But if you're really going for high precision, using the cheapest Chinese ball screws you can find isn't going to get you there. You can read what one customer had to say about these things on Amazon: https://www.amazon.com/SFU1605-Balls.../dp/B073VHJQWM . Holding .01mm over the distances you're asking for would require expensive precision-ground screws, not these.

Going for super-high RPMs on the spindle is good if you're using it for engraving, but it takes expensive bearings to hold up to that. And a spindle like that will be fairly useless for other purposes - it would just burn wood when you tried to cut it with a normal-sized tool - if it actually had enough torque to spin it. You might consider using two spindles; one for general purposes that went slower with more torque, and another for engraving and micro-tools.

The motors you're looking at will probably work if you can give them enough voltage and amps. On this chart (Datasheet) 57BYGH218 PDF - Promoco : Hybrid Stepper Motor (1-page) , you can see they have an inductance of 4mH, which means that 64 volts would be optimum. I doubt your Arduino drivers would handle that, though.

It's usually a good idea to build your frame first, and then find the electronics and motors that will make it move. You seem to want to reverse that , but the frame design can (and should) evolve in the process of building and rethinking, and you don't want to be stuck with electronics that won't really work for what it turns out to be.

[wizard]

Im on my cell phone at the moment so this will be very short.

1. The frame is too flimsey for what you want to do.

2. You will need to review your dimensions to make sure that when using the fourth and fifth axis, you have the required clearances. I didnt look at this closely but your numbers had me concerned. In other words when you rotate the trunion up will you actually clear the gantry, the Y axis slide and any tooling used.

3. Doing your own spindle is non trivial if you want high speeds and dimensional control.

4. Building a machine to hold tight tolerances gets to be expensive. It is doable but fair warning is required. You wont be able to bolt linear rails to a weldment and get acceptable behavior. You will need to either machine crtical surfaces or use other surface leveling techniques often discussed here.

5. If you have a local machine shop that you have a good working relationship with i would suggest that the base frame be built out of large steel tubing with welded in place pads for the linear rails and gantry supports. Ideally that frame would be stress relieved before machining. The idea here is to set your self up with a good base upon which to set everything else.

6. Remember the gantry beam is only supported at the ends. As such it benefits from being a large square tube.


In any event six quick suggestions. 0.01mm isnt impossible but that is a lot tighter than most machines built in this forum. The workspace volume you are asking for isnt huge so a reasonably stiff machine can be had without resorting to huge structural pieces. The real advantage with respect to small machines is that it is fairly easy to find small machine shops with the capacity to machine your components.

[louieatienza]

I'm pretty sure your configuration is A-C for the 2-axis table... A-B would be a gimbaled unit. A revolves around X, B around Y, C around Z.

Like the others have mentioned, the uprights and gantry could be built a little sturdier with larger tubes, and larger tubes overall on the build will make alignment easier and add some stiffness. Also you should consider your gantry height. Normally we'd recommend the lowest gantry height possible, but in your case you'll need clearance for that turntable when the trunnion is at 90 degrees, and at zero if your work needs to be mounted on a riser or 3D vise. So whatever clearance you need PLUS the length of the tool and THEN rapids clearance.

[telstar505]

Yeah, I also thought 25x25x2mm frame members would be too flimsy, will try to double that.

Its also a good idea to start with 3-axis and add a the trunnion table after everything else has been sorted out.

I wasn't planning on mounting the square rails on top of the steel tubing. I meant to weld a plate on top of the tubing for each axis, then have it machined so the rails can sit completely flat, like this:


@louieatienza Thank you for your point on the A-C mount! I actually did not know the correct way of referring to these rotary axis.
I understand the gantry height should be as slow as possible so I will only add enough for the trunnion table+the estimated workpiece size.

Questions:
- Can you recommend better rails that would be more suitable for my purposes?
- Do you think a pillar type design would be better?

Thanks a lot for your comments!