For some years I have an eShapeoko with a 1.5 kW spindle+VFD.

[Menno_de_Graaf]

For some years I have an eShapeoko with a 1.5 kW spindle+VFD. Although it cuts nicely in wood/MDF, this machine is a bit too light for milling aluminium. Also the spindle is a bit too heavy. This time, I decided to make my own machine design at the same time increasing the working envelope.

The aim is a modular ball screw design with supported linear rail based on aluminium 30 series extrusion profile, 10mm alu plates and otherwise mostly standard components:

• Alu profile 30x60 (or 30x90) and 30x150
• 10mm custom alu plates
• SBR16 supported rail with SBR16UU linear bearings (blue)
• SFU1605 ball screws and ball nut bracket (green)
• FK12/FF12 ball screw end supports (green)
• NEMA23 motors mounted with M5 spacers

Since it is a modular design, it can be made in almost any desired size, my own build uses 750mm Y-rails and 1000mm X-rails. This gives a usable working area of about 550x800mm.
Currently I completed the XY parts using MDF plates to see if I got all the distances and holes right. Comments and other ideas are very much appreciated!

[Luis_Diaz]

Cool! Do you have any idea of how much will cost to make one? I have a small cnc and will love to have a bigger one

[Menno_de_Graaf]

Expected that question, so I am currently summing up what I already spent on the prototype. I will let you know, when I am done.

[Luis_Diaz]

@Menno_de_Graaf thanks! That the con of a nice job hahaha

[Josh_Rhodes]

I’d like to know the cost, and how well it ends up working.

[Menno_de_Graaf]

A first rough estimate of the costs for this project, includes already the Z-axis:

• Alu profile: 150 euro
• Ball screws, nuts, end supports: 220 euro
• Lineair rail and bearings: 170 euro
• Small hardware, rings, nuts, bolts: 30 euro
• Motors+mounting material: 125 euro

Plates not included yet, and of course also a power supply, control board, stepper amplifiers, connectors, cabling and spindle need to be added to get a fully functional machine.

[Menno_de_Graaf]

So a quick comparison:
Costs for the mechanical part (without motors) of this project is about 570 euro, still without the custom plates. A same size eShapeoko prepared for NEMA23 would cost about 410 euro.

[Luis_Diaz]

@Menno_de_Graaf worth the money?
missing/deleted image from Google+

[Menno_de_Graaf]

My eShapeoko is 750x500mm and as said for milling wood it is great, but that’s the practical limit. I would not recommend this large a size due to flexibility in the X-axis and Z-axis.

[Luis_Diaz]

@Menno_de_Graaf itc could be used to mill pcbs or acrilic? Not really interested on wood mill (but it could be useful to mill some parts for a arcade cabinet that I’m thinking about)

[Menno_de_Graaf]

The eShapeoko is a good small size desktop CNC, but it remains belt driven. I have used it for milling pcbs, results were ok. For a small size machine I would go for a 6040 machine instead.

[Marcus_Wolschon]

Is the X axis not overconstrained with 2 drives? I see a number of locations where stiffness can be increased.

[Menno_de_Graaf]

Overconstrained?! There are numerous designs with both single and dual X-drive. The reason I decided to go for dual X-drive is that no moving parts are needed at the bottom side and thus no large unsupported distance needs to be covered. This way, the machine can very easily be enlarged in Y-direction without affecting stiffness.

Obiously there are places where stiffness can be improved, but every machine including this one is a compromise. So, the question is: what is the most important place where stiffness needs to be increased?

[Marcus_Wolschon]

I don’t trust the stepper mounts to not twist (a 3d printed or purchased block with a few access holes may be stiffer).
The wooden gantry plates are not very stiff. A thicker Aluminium or POM version may be better (adding recesses for the carriages and bearings allows it too be thicker everywhere else.)
the wooden motor mounts may bend when pulled/pushed (make them L shaped) …

[Menno_de_Graaf]

Thanks, using standard stepper mounts was the original idea. I already bought a couple of those, but figured the M5 spacers would be a simpler solution. Currently the spacers are mounted in the MDF plates, which already feels quite solid. Lets see what happens when the motors really start spinning. It will be straightforward to return back to stepper mounts if it turns out to be needed.

[Marcus_Wolschon]

@Menno_de_Graaf What accuracy, speeds and forces/stiffness are you planning for?
You mentioned Aluminium.
I guess it’s not only soft AlMgSPb and not just engraving but cutting it?.

[Menno_de_Graaf]

And of course, the MDF side plates are not the final solution as you could have read above. I plan to use 10 or 12 mm alu plates.

[Marcus_Wolschon]

Where do you plan to mount the cable chain to run all cables and tubes(if water cooled) to the spindle on the gantry and to the Z-stage?

Did you plan any way to adjust squareness and tramming?

[Menno_de_Graaf]

Cable chain along X-axis at the rear side of the gantry. Along Y-axis at the table.

Squareness can be adjusted with the bolts in T-nuts.

I am not designing for very specific accuracy/speed/forces/stifness. Having said that, a practical goal is milling alu from 6xxx series in 0.5mm steps with feedrate around 1000 mm/min.

To put it simple: I would be happy if I get a much better machine than eShapoko/Shapeoko2 for less than twice the price.

[Paul_Shaw]

I was told when designing my machine that those cheap rails preformed better on the top and bottom of the beam, it makes them further apart and the direction of the force is more rigid. Also high rails reduce the flex in the sides as well as reducing weight that’s moving and keep the bearings and screws a bit further from the dust from cutting.

missing/deleted image from Google+