View Full Version : BUILD LOG: "Henge", 3 Axis Cnc, Fixed Gantry, Moving Table

18-03-2017, 11:38 PM

I have previously had designs on building a cnc, but various stuff cropped up & diverted me, so they never got past the design stage.
I'm hoping to re-boot with a smaller, simpler design that i have been researching for quite a few months.

I am immediately breaking the rules as over the years I have collected various mechanical components and I have designed a machine around what I have “in stock”
I know is the wrong way to do it, but that's just the way it is with this one, cos i'm too tight to chuck owt away!

I want to build a Fixed Gantry, Moving Table Cnc. The working area will be constrained (by the parts i have) to 350 x 350mm x 200mm(Z).
Mainly for cutting Aluminium, but also wood & composites such as fibreglass.

Parts & Materials I already have :
6x 20mm Misumi linear rails 580mm long
3x 16mm, 5mm pitch NSK ballscrews with a travel of 350mm
6mtr 200 x 75 steel C-Section
The C-Section is parallel flange and is sturdy stuff (6mm thick with 12mm thick sides)which weighs in at about 23kg /mtr
I realise box section is better & stiffer, but i got 6 metres of this stuff super cheap ages ago & i'm sick of falling over it!

Here are some basic renders of the initial design done with my first steps in fusion 360 (which is a great program)
I have not gone into great detail in the cad as I don't think it's worth rendering every last detail, just the general outlines. The cad is excellent for finding interference parts & testing movement etc.

Construction of the frame will be tack welding, then lots of drilling & bolting, with brackets where needed. I only have basic hand tools & a drill press, and very little mechanical experience.
The orange plates are 10 or 12mm ground steel gauge plate & the white parts Aluminium tooling plate. The red indicates the cutting area.

The Ballscrews are 5mm pitch so I'm planning on ditching the motor mounts & using belts & pulleys to gear 1:2 on X&Y , at the moment I have Z at 1:1
Motors will probably be 3 or 4Nm Nema23 and AM882 drives
2.2kw Chinese water cooled spindle & VFD.

I'm planning to do Axis alignment on the Y plate using epoxy putty & location pins. Then bolting.

Any comments, criticism, suggestions or ideas gratefully accepted.

19-03-2017, 10:02 AM
Hi Greeny,

Looks good and should do what you want it to.

Yes, box would be stiffer but I think this will be plenty good enough as it is. I guess you could plate over the back of the C-section that runs across the top between the two sides if you have a bit of plate spare.

Also the Y axis will be limited in stiffness by the small L bracket which holds the AC bearing block - I would reinforce this or look at a stiffer way to mount it.

19-03-2017, 11:01 AM
Hi routercnc, Thanks for your input.

I wiil look in to adding some braces as you suggest. Easy to do and extra strength is always welcome.

Great point on the bearing block brackets, i will look in to beefing them up.


Robin Hewitt
19-03-2017, 11:38 AM
Looks around to see if Jazz is waiting to pounce :dread:

Coast looks clear, time to make a radical suggestion :topsy_turvy:

If the linear blocks on the Y and Z axes did not line up with each other or with the router itself, then with a bit of jiggery pokery the Y and Z rails could almost touch, the router could sink into the plate behind it, the overhang greatly reduce and the X axis bearings could move even closer to the gantry thereby stiffening the whole assembly :victorious:

It is a matter of thinking in 3D. :beer:

This message will self destruct in 10 seconds. Don't tell him it was me :black_eyed:

19-03-2017, 12:44 PM
Hi Robin,
Don't worry your secret is safe with me :very_drunk:

I think i understand the intention of what you are suggesting and like the idea, but not exactly sure on the details.
Could you explain a little further?


Robin Hewitt
19-03-2017, 01:22 PM
Right now you are simply stacking layer on layer and moving the router further and further away from the gantry, ignoring the fact that overhangs are not your friend. This is the peril of designing a 3d object on a 2d sheet.
The Y and Z linear blocks could be sunk in to the plate until the rails hit the plate. Less overhang. If they didn't line up with each other then extra metal could be bolted on behind to compensate for the loss. The clever bit would be removing still more metal below the blocks so you approach the limit which is where the rails rub against each other.
I am all in favour of pushing the limits of the design rather than slavishly following what everyone else has done in the past. I am not building a router so some might say I should keep my fat mouth shut, feel free to tell me where to stick this stuff.
Your base is overhanging the X axis bearing blocks presumably to hold the X axis screw. All you have to support is a straight push pull, you seem to have a lot of excess metal thereabouts :very_drunk:

19-03-2017, 01:44 PM
Ah I think I see now.
I will have to think about that. I can see the benefits, but it complicates construction quite a bit and i'm very limited on skill, experience & tools! Will ponder if i'm brave enough

Good point on the base, i will take a look at that and see what i can do.

I appreciate your input, & most certainly won't tell you where to shove anything :chuncky:
I also fully agree with your attitude of thinking "outside the box" or "racing your own whippet" as we say up here :beer:


19-03-2017, 01:54 PM
A half way to Robin's suggestion to reduce the overhang would be to place the Z screw outside the blocks so the risers (in red on "A") can be removed.

This would reduce the overhang by 30mm, but widen the Y plate by 75mm to accommodate the screw. Although some weight could be saved by having a cut out as in drawing C1
However the risers also serve to stiffen both Y & Z plates, so i'm not sure.

I would be interested to know anyones thoughts on this (or any other part)


Note: The orange X & Y plates are 10mm steel guage plate it would be easy to up them to 12 if i removed the risers, only adding 4mm to overhang
Drawing B & C are the same just different angles.

19-03-2017, 02:32 PM
Sinking the rail blocks into the plates i think i could reduce overhang by about 16mm (8mm + 8mm)

I don't think i can reduce it further as the rails are fixed to the 2 outside orange plates and move with respect to the center plate.
If i sank the blocks any deeper the rails would interfere with the centre orange plate.
They move over the whole length of the centre plate so, if i went any deeper, i would have to put a groove in the whole length of centre plate weakening it.

Basically i'm saying the orange plates must be at least a rail height apart, so i could only sink them about 8mm each.
Or have i misunderstood?


Robin Hewitt
19-03-2017, 10:49 PM

If you can't do it, you can't do it. Picture shows it is possible. The white arrow points to a 1mm clearance between the X axis linear rail in purple and the 8mm linear Z axis shaft in yellow :topsy_turvy:

20-03-2017, 01:12 AM
Thanks Robin,
However, am i right in thinking that's only possible when using unsupported round rails on z?

I think I'm probably interpreting your drawing wrong, as there's a lot of stuff there and I'm not sure how it all fits together in 3d, do you have a side on view?


Robin Hewitt
20-03-2017, 02:54 PM
What you could do is slip pairs of linear blocks on to two pairs of linear rails, lay two rails on top of 2 rails at right angles then slide the blocks around a bit and see where the metal of the Y and Z slides can fit in around it.

Does that make sense? :beer:

20-03-2017, 03:05 PM
ok, will give it a go, thanks.

20-03-2017, 03:28 PM
I think i see now. Playing with two pairs of rails it took about 60 secs.
I will take some photos & do a couple of drawings later this evening. (i'm supposed to be working now... shhh)
Cheers :thumsup::thumsup:

20-03-2017, 07:13 PM
Well i think Robin's idea is pretty nifty, but i don't think it's going to work for me, as the rails i have are not long enough!
A good example of not buying until the design is finished!

The only way i can get to sensibly work is as per the drawings below.
The Z table must fit between the Y blocks which means the y blocks are spread quite wide.
In this case restricting my Y movement to about 250mm which is too small (I'm aiming for 350mm)

I do quite like it though, it makes for a wider Z assembly than usual, but makes it very compact in other respects . (always some tradeoffs!)


20-03-2017, 07:43 PM
oops, forgot the ballscrew!