View Full Version : Design for 1800x800x80 wood router

15-08-2014, 07:10 PM
Having got a rough quote for the CNC routing I will need on a repeat basis (hopefully) I want to work up a design to build my own machine when demand is established.

First thing is to get the basic layout sorted, and I want to get an idea of what components I will be using and how much they will cost.

This is for 2.5D routing of almost exclusively 18mm melamine faced chipboard.

I've started with a simple ladder frame, from 80x80x3 steel tube. Then I've used that for the gantry too. Laser cutting of steel and welding I can get done locally.

I've gone for a small 80mm travel, just because I don't foresee the need for any more.

I've put SBR16 rails on just because I saw them in a thread today.

I'm interested to hear what ballscrews and motors I should be using. This router isn't going to be used non-stop so linear speed and power is only needed for cutting performance rather than to churn the job out fast. 50m of full depth profile cutting is the most it will do in a day, so even a cutting speed of 1m/min would be sufficient, but a bit boring.

I've drawn a couple of X rail layout for comment.

sketches are from powerpoint I'm embarrassed to say.

01-03-2015, 07:03 PM
No interest last time, must have been such a brilliant design it didn't warrant comment :)

I've now made the decision to get a bigger machine, so it's time to explore the various possibilities and either import, buy, build, or get someone to build me something.

I'd like a machine that is stiffer than my cheap 6040 for cutting plastic, and also big enough to cut 1650x590 wood panels. I'm thinking of making do with a 6090, but this thread is for exploring a machine sized for the job (and some spare width for larger designs that the 6090 won't be ablo to do), ie 1800x800.

So here is a general layout.

The things I need help with are making the decision on construction materials and methods. Welded square steel tube, or aluminium sections are the two main choices.

Also single or twin ballscrews on the long axis? I think I prefer the twin design and the base will be easier and more compact, but I need to work out the cost delta. The single ball screw makes it easy to undermount the rail, if that's a good thing?

I can get a steel frame welded quite easily, but then how to align the rails on it?

With aluminium section, I guess you rely on it being straight? Then the question is how to square up the frame. My friend suggested poor a concrete bed and bolt it down to that to get it flat (equivalent to epoxy method).

I will start to gather some prices for parts, hopefully people will then comment on how to do it cheaper.

And, before anyone asks; No, I don't want to make it 1200 wide :)


01-03-2015, 07:26 PM
The concrete base idea would be something like this.

It looks like a lot of work and materials now, but if it would work, it does require much skill.


01-03-2015, 07:46 PM
Has anyone done this?

Weld up a steel ladder frame, bolt strips of acetal to every member, and take it to a friendly wood cnc place and get them to skim it flat. Also drill some locator hole where you will tap and thread later.

If you had a flat surface, would that be enough to ensure a straight linear rail? ie can the rail be reliably attached to the aluminium profile and be straight, using countersink washers and screws or something like that?


01-03-2015, 11:14 PM
Nope not done anything like that....

IMO the best thing to build a frame out of would be Sq section steel. There's no advantage to making a steel and concrete base, or steel and acetal frame so they're accurate and then bolting an aluminium frame to it that's probably not as accurate. Make the whole frame from steel, build in lots of adjustability and shim/file/fill the top rail (or go the self leveling epoxy route) so it's flat and fit the rails (which if supported round will not be level/flat) to that? Round rails are best mounted upright as the bearings aren't as strong on their side. At 1800 or 2000mm long you're into 20 to 25mm ballscrews to stop whipping at high rpm and Nema 34 motors...or rotating ballnut on 16mm ballscrews. Have a look at Kingcreaky's builds for inspiration (I'm sure there are others like Jonathan's first build to look at too but his spring to mind first as large wood routers, especially the last one!!)

If you want to do any concreting, name a nice flat base for the frame to be bolted to.

02-03-2015, 01:04 AM
Ah I forgot to change the sketch, I'd use profile rail now given all the comments I've read on this forum.

I've just looked at some more posts, looks like building the frame and then bolting on 2 main beams with shims is the way to go.

When you shim a linear rail, is that putting a washer under the rail at each screw position? And do you file it to thickness or just use shim washers?

02-03-2015, 06:08 AM
You shim the main beams not the linear rail, either using proper shim material or this forums favourite method...metal from an aluminium can (Guinness being a favourite) and then using a straight edge and feeler gauges find the high spots and file/scrape them and fill any low spots. Then the linear rails can be mounted straight on the beams.

02-03-2015, 12:21 PM
Ah thanks, shimming the linear rail didn't seem quite right.

By shimming the main beam, I assume that means shimming an intermediate rail on the main beam.

My acetal rails plan would do the same thing, as long as the larger cnc has a flat bed. I'm not sure how my cnc place would feel about helping me make my cnc machine though! It's an obvious money loser for them.

I'm thinking now that a welded steel frame is making life hard for me. This is for wood mostly so an aluminium extrusion design might be more suitable. I quite like this design. It looks easy to move/install elsewhere which also makes it more re-sellable than a great big welded frame.

The question I need answering is: Can you just bolt a hiwin rail to an aluminium extrusion, and the result will be straight enough for wood panel cutting ?

02-03-2015, 12:35 PM
You mentioned a larger screw being needed, I guess that depends on feed rate as well.

The other structural option is single screw with a long beam.

Can someone help out with rough estimate of the cost comparison. I need the cost of a twin ballscrew set-up versus a single screw for 1800mm travel. That includes the screw assembly, and the motors and motor drives. I realise there's a lot of calcs in that, and it will take me some time to track them all down on the forum.

Here's the single screw layout, I'm not sure I even like it. If the beam is going to need to be deeper then that's already counteracting the single screw cost saving.

If someone was to say forget single screw, you need 2 screws, I wouldn't be upset.


02-03-2015, 01:44 PM
Have a look through some of the build threads and all your questions will (OK not quite all) be answered.

Here's a good starting point http://www.mycncuk.com/threads/8242-8x4-Wall-Mounted-Upright-ply-cutter

Then there's Jonathan's first build: http://www.mycncuk.com/threads/2288-1-7*0-74*0-4m-Mill-Router-building

Another couple of Matt's builds...


Read the above build logs as a minimum, download yourself the free version of Sketchup and have a play drawing up frames with dimensions on etc and then post them up for critique. I'm not saying that the way they've built their routers is the correct but it's a proven route so can't be bad....

02-03-2015, 10:38 PM
Forget single screw you need 2 screws! But don't buy anything until you have the design totally down. It looks like this design has a little way to go. Have you drawn it in 3d?

02-03-2015, 11:03 PM
Funny you should say that, I was just tending towards one screw.

I've read some more threads, especially the kingcreaky ones where he seems to just weld up a steel frame a whack the rails on it. OK, he's probably an expert welder. But still, I might be worrying too much about getting the rails straight and flat. Which means the earlier steel frame is, well, back in the frame.

I can't use any 3D drawing packages yet, but that's fine, you have to decide the general configuration before you draw the details.

I'm still interested in the best way to do make this, but it might be academic by tomorow. I'm likely going to order a 1800x900 machine from Eric at stepmores.

02-03-2015, 11:49 PM
I don't know stepmores I'm afraid. To get the most out of the machine you are going to have to get into 3d sooner or later, I would strongly recommend sooner! You don't need to be an expert welder and getting the rails flat can be achieved by epoxy levelling as evidenced by loads of build logs on here... Steel frame is the way to go.

05-03-2015, 03:04 AM
OK, in order to procede let's go for the twin ballscrew design.
Aluminium extrusion structure or welded steel doesn't affect the drive system.
We can assume an aluminium gantry.

What drive system do I want?

2 ballscrews, 2020 to give about 10000mm/min ?
Driven by Nema 34 was suggested.
Pulleys to reduce machine length, and spin motors faster? (What ratio?)