Rough cost for a CNC router

[m_c]

I'm currently making a product that involves making custom beech handles, as I need a size that I just can't find ready made.

At the moment, I buy planed all round blanks, screw a template to them, rough cut them on a bandsaw, trim them to the template on a router table, remove the template, switch to an ovolo/roundover bit to round them in the table, manually thin down the end to fit in a bit steel box section, and then sand them to smooth out the inevitable ripples from the router table bouncing as it profile cuts a full 32mm depth.
I reckon I could get the time down to 30 minutes per handle if everything goes to plan, and I make another jig/sled for the final sizing.

But I've been wondering if it would just be easier to build a suitable CNC router table. Screw some blanks down, hit a button, and walk away. Flip the blanks over, and do the same again.

Work area size, at a minimum all I'd need is 4ft by less than 6 inches, but it would obviously be wiser to make it bigger. I'm probably thinking about 4 x 2 to avoid it taking up too much space and to minimise component sizes/reinforcing, but could probably stretch to 4x4.

I'm thinking throw together some 30x30 or 40x40 box section.
Bolt some rails on top (would profile be good enough, or would full linear be worth the extra cost?)
Build a gantry from some Ali plate/profile.
Add ballscrews (or would R&P be simpler, to avoid having to shield screws?)
Add a water cooled spindle.
Add some steppers and a controller.


Can anybody give a rough cost of what something like that will cost in parts?
I'm trying to make a decision before I spend too much time on this idea, as I've never really paid much attention to the woodworking machine builds on here previously, as I've never had a need for one :-/

[ngwagwa]

I make wooden fishing lures using a template and a router and I can profile and round over both sides in a couple of minutes and don't think it would be faster by CNC. There's a video here showing how I make them, may give you some ideas?
https://m.youtube.com/watch?v=3SsZSEAbkEk&feature=youtu.be

[JAZZCNC]

I make wooden fishing lures using a template and a router and I can profile and round over both sides in a couple of minutes and don't think it would be faster by CNC. There's a video here showing how I make them, may give you some ideas?
https://m.youtube.com/watch?v=3SsZSEAbkEk&feature=youtu.be

Oh, you've got a happy shock coming when you get your machine working. I could have 6 in the time it took you to do that 1 in the video...Lol

[m_c]

I make wooden fishing lures using a template and a router and I can profile and round over both sides in a couple of minutes and don't think it would be faster by CNC. There's a video here showing how I make them, may give you some ideas?
https://m.youtube.com/watch?v=3SsZSEAbkEk&feature=youtu.be

Interesting video.
If I could get the total time down to 10-15minutes per handle, I'd be happy, but it's just time consuming.
Screw the template on, and doing the trim on the bandsaw I estimated takes 5 minutes minimum.
Then profiling to the template is realistically 5 minutes, as my router table can't take that deep a cut as it's not rigid enough, so takes a bit finessing to avoid too many gouges.
Removing the template then rounding over takes another 5 minutes.
That's 15 minutes already, not including tool changes.
I machined 3 handles today, and it took me nearly an hour, including tool changes and tidying up afterwards (even with a good extractor, the router table throughs a lot on the floor).

Then even if I bought another router and made a jig/sled for thinning the ends, I'd estimate that'll be another 5 minutes.
So that's about 25 minutes per handle, and doesn't include time for sanding the handle to get rid of the ripples from the router table resonating/flexing, which is about 5-10 minutes per handle.

So even with everything as streamlined as possible, that's 30-35 minutes per handle.

If I switch to a CNC table, I guess labour input is likely to be around 5 minutes per handle, including swapping tools/blanks, if I can do 3 or 4 in one setup.
That's a 20 minute saving per handle. Plus there is the option I could probably switch to cut timber and have the router finish the surface, saving money on planed timber.

However, I need to workout if the labour saving will be worth the cost of building the machine.

[JAZZCNC]

Moray if you can get away with 4 x 2 then do so because it makes the build so much simpler and cheaper because you can get away with 1 ball screw on the long axis.

Stay away from R&P it's a pain in the arse unless you use helical rack which is expensive. Even then it doesn't come close to the smoothness and efficiency of ball screws and it's much more complicated to set up and keep adjusted properly so backlash is something like acceptable. It needs a gearbox or belt n pulley ratio set up to do it correctly so get decent resolution. You don't get any of this with ball screws as you know from Mills lathes etc, just fit and forget except for odd bit of grease.

I wouldn't build a frame from anything less than 50x50x5 box and brace it up well. 5mm gives you something to tap into for the rails and lowers the resonance better than thin-wall 3mm stuff.
Stay away from the Round type rail and go with profiled rails. Much better far less hassle. Last 10x longer

Gantry use profile and My L shape design it's strong simple and proven on 100+ machines now.

Cost is difficult because as you know it's the electrics and ball screws, rails etc which eat up the money and how far do you want to go with quality and where are you willing to buy from.?
The best place by far is China but there's a lot of rubbish out there, the cheap screw/rail kits you see on eBay, etc are nearly always low quality and mismatched, like 5mm pitch for a router which you really need a minimum of 10mm. Same with the electrical kits, steppers with high inductance, and low voltage PSU.

That said if you keep it simple on the electrics and go with say Linux CNC and Parallel port with a simple BOB then it's possible to build at sensible money.

[m_c]

4 x 2 I think is the smallest I would be willing to go, as it'll make it a more versatile machine.
I was toying with the idea of making the gantry on the long axis, to leave room for future expansion, but in all honesty, I doubt I'll do much other work in wood.

Cost I want to try and keep as low as reasonable, but would prefer to source in the UK for speed, but even if I decide to go ahead with this, it'll probably take me a couple weeks to get the frame built anyway (I've got a queue of other jobs I need clear out the shed first!), so that isn't set in stone.


One thing I can't find much mention off, is what do you do about protecting the ballscrews?
Or is it fine to leave them exposed, with regular lubrication?
(my lathe had been used for wood when I got it, so I'm well aware of how sawdust can penetrate components!)


Electrics is the bit I know the most about, it's just all the practicalities/mechanics of a woodworking machine I don't.

[Kitwn]

Have you thought of farming out the blank cutting to someone who wants to earn a bit of pocket money with their existing machine? I'd offer to help but the shipping might be prohibitive.

[JAZZCNC]

One thing I can't find much mention off, is what do you do about protecting the ballscrews?
Or is it fine to leave them exposed, with regular lubrication?

They handle sawdust ok up to a point but as you know it does get in there eventually. Anything you can do to keep them out of the firing line helps which is why I put them under the table out of the way rather than stuck on the sides. The "L" shaped gantry puts the screw at the rear again out of the way.
It's a simple proven design that doesn't require fancy guards or bellows etc and does a good job of protecting screws.

[m_c]

Kitwn, I have considered it, but I like a good excuse for building another machine!

Thanks Dean. I'll just make sure to keep things well oiled to try and keep the dust flushed out.



What are the good build threads to look at for inspiration?
So far I've found that 45x90 is the recommended profile for the gantry, then what looks like 15mm plate for the major components.
What kind of spacing are people aiming for on the linear bearings?

I'm currently trying to decide whether to have a single underslung ballscrew for the long axis, or go for twin.
Single means the entire length of the bed will be unsupported (probably not that much of an issue?)
Twin means I can add triangulation, but I'm thinking a couple extra lengths of box over the length would probably give as much support as I'd need.

My current thoughts would be, aim for 8" clearance between gantry and the frame. That will give room for a work surface, and the carriage assembly, leaving about 6" useable clearance, and aim for Z-axis travel of 6-7".
Aim for 24" cross travel, so probably gantry rails will be about 32" to give plenty spread on the linear blocks.
Then aim for 48" of gantry travel, so again rails would need to be about 54" long.

[Boyan Silyavski]

Oh, you've got a happy shock coming when you get your machine working. I could have 6 in the time it took you to do that 1 in the video...Lol

I was going to say that. From that chunk of wood and 2 dowel pins i can 3d machine like 20 at once for half a day, ready to sand with 120 grit. People just don't get what a tremendous upgrade is a simple vacuum fixture or ATC on a CNC, not to speak of a whole CNC.


I am a big fan of "Push the START button and leave the area" and come back after the job is done. Robot comes from Cyrillic and means SLAVE in my language. So is the origin of the word Worker , but that's another matter.


If the machine is not wider than 50cm 1 ball screw will be ok. Probably even wider as 60-70cm will be ok, if you don't use it for heavy duty jobs like aluminum. The great decision will be about the Z travel. Do you make it 150mm for normal wood work or 300 for special jobs, now that you are there and will not cost so much.

Price will be very dependent on the linear bearings and if you can weld save from aluminum structure.


I see a machine like this so:

Base frame table made from steel box. On top of box bolted the square rails 20 size and shimmed, so save 200 from epoxy. 1 ball screw, servo driven and rotating screw. Mounted directly on bottom of table. So you see, steel box table must be strong as has to be one piece and the screw to pass beneath and rotating nut to attach to both sides,
but thats cheap, steel is cheap. a couple of 100x100x3mm for the bed. forming a ladder for simplicity. wood or MDF on top.


It could be done cheap but many decisions has to be made. I think that for 2k could be done. Having in mind that for a proper 90x60machine there are 3k in parts or more.



PS. 2 extra bearing blocks that are next to nothing so the bed has 3 long bearing blocks each side positioned correctly / not in the middle, rather reinforcing each side end will help for a no problem 60-70cm width. I have never seen something like that but just came to my mind and i am sure will work fabulously. saving cost of driversm motors, rotating nuts and so

[m_c]

This project is now on hold for the time being.

I had a slight mishap with my lathe, and the turret is now dismantled across the floor in more pieces that what it should be able to be dismantled into, so that is the current priority. If I can't get the required parts to repair it, I've got about a month to come up with a replacement :-/

[m_c]

So the resume button has tentatively been pressed for this project, as I'm getting fed up hand machining bits of wood.

Some initial sketches have been done, so I can get an idea of component layout.
My current plans are wide enough to machine 4ft with the gantry on the long axis for better access.
I'm still undecided about length. 2ft would be enough, but having measured it's potential location, I think I could go to 4ft without any issue, which would increase versatility.

Gantry will be based around Jazz's tried and tested double profile L shape. I'm thinking 80x160 profile for this.
Where are people buying aluminium profile?
I'm aware of KJN, but my googling isn't turning up many other UK sources.

Main frame will be steel, and I'm thinking of using an epoxy poor to create the level surface for mounting rails.
I was going to go for either 50x50 or 60x60 5mm wall box for the bed frame, with probably something a bit bigger for sides where the linear rails will mount, but I'm currently wondering if stacking two of those would work. I think I need to work out how much bed clearance I'd like, before deciding on this bit.

[Cube3]

Applied Automation for another aluminium profile supplier.

Sent from my COL-L29 using Tapatalk

[Kitwn]

Here in my new home town of Devonport, Tasmania the local steel supplier will either sell cut pieces by the metre or will sell you whole 6m length and make as many cuts as you want for free, a signifficantly cheaper option. If the same applies in the UK it may be worth designing your frame and the choice of material sizes used around making best use of full lengths of one or two sizes only.

Looking forward to seeing how it goes.

kit

[m_c]

Since I have finally sold the machine that was sitting where this machine will go, and I'd like to get the base and epoxy done while we have reasonable weather, I've been making a point of adding at least one component to F360 every night, starting with the spindle and working from there.

This is the current basic layout, so I can get a better idea of how things will be positioned.
30048

I have managed to misplace the notebook I had my previous sketches in, so have had to start from scratch.
The spindle is modelled around a 5.5kw ATC spindle. The machine may start life with just a basic 2.2kw, but I'd like to get something with an ATC at some point.

Profile is Kanya 80x160.
Rails are Hiwim 20mm.

Does anybody have a reliable source for 3D models of ballscrews and support bearings, that doesn't involve endless searching for something that might not be the correct size?
For ballscrews I'm currently thinking 1620 or 2020, depending on what fits in best.

[Doddy]

Does anybody have a reliable source for 3D models of ballscrews and support bearings, that doesn't involve endless searching for something that might not be the correct size?
For ballscrews I'm currently thinking 1620 or 2020, depending on what fits in best.

daft question, but you've tried the McMaster lookups from within F360?

[m_c]

daft question, but you've tried the McMaster lookups from within F360?

Nope.
I have seen that in the past, but never actually bothered to try it.
Thanks for the heads up, as it meant I managed to add a 2020 ball screw, nut, and a couple support bearings. Probably not the exact components I'll end up using, but near enough for layout purposes.

So here's tonight's progress-
30054

Not much, but I've had a couple other jobs to do tonight, but it's something.

[m_c]

After what has been a pretty rubbish few months for personal reasons, and an absolute manic 6 weeks with only 2 days off due to everybody running events that have been delayed due to a certain global issue, I've found some motivation to progress this over the past few evenings.

This is the current layout iteration, after various sketches-
30640

Or if anybody would like to view the 3D model. this link should work - https://a360.co/3wyt3zD

I'm still not entirely convinced this will be the final layout.
X-axis ballscrews will be in the way of putting legs on the corners, but I can probably shorten them.
Frame is currently all 60x60x5 box section, but I may use lighter section for the cross beams (I've only added two for now, but more will be added)

I need to get the gantry end plates drawn up, and adjust a few dimensions so things will clear/fit.