My first go at making a CNC router

[Norgmonster]

Hi All,

Its great to have such a community to be able to talk to. I'm 31 and i have never joint a forum before.

I hope i can contribute as much as i will ask of you :wink:.

First of all (TOTAL NEWBIE QUESTION), were can i get preexisting CAD files of components for my cnc router plan? Stuff like ball screws, rack rails, drive rails and so on?

Best,

Norgmonster

[Clive S]

You can get them for the hwin site and this place https://grabcad.com/ Have you decided on a cad program Fusion360 I think is free for hobby use and it also has CAM built in. There are plenty of tutorials about on the web.

[Norgmonster]

Thanks for the info. It will be a great starting point.

I've needed a project like this to finally make me learn a descent CAD program, and AutoCAD has been on my desktop gathering computer program dust for far to long :-)

I've never heard of Fusion360. Will it run on a Mac? If it also runs CAM then it might be worth a look.

Whatever gets my plans out to you guys faster the better :-).

Best for now,

NorgMonster

[lucan07]

Fusion360 will run on Apple Mac® OS® X Mavericks (10.9.5), OS® X Yosemite (10.10.5), OS® X El Capitan (10.11)

I personally found it by far the easiest to use and learn from the many I installed.

[Norgmonster]

Good to know. I will give it a try first thing in the morning.:sleeping:

[Norgmonster]

Most of you guys may know all about it but I have found a great resource of 3d CAD files on a site called www.b2b.partcommunity.com (http://www.b2b.partcommunity.com). The files are also available in a number of formats so they can be read by almost any program. https://grabcad.com/ only seams to support Solid Works.

If anyone is interested then let me know what they think.

My CNC router is moving along and i hope to have something to show you all tomorrow evening.

Good night all,

Norgmonster

[Norgmonster]

Hi All,

I've been working all week so no time to post much.

I'm learning Fusion and getting my design nicely visualized. In doing so i have come up with some questions:

For my X, Y and Z axis I'm thinking to use Hiwin rails and blocks. What size blocks would be sufficient / not overkill for a R&P driven, 4x8 bed? I'm thinking 20 but would 25 be better? Also, as the axis gets smaller can the rails get smaller as well? For example, 25 rails on the x, 20 on the y and 15 on the z.

Also, is it worth using a belt between the stepper motor and ball screw on the z axis to prevent backlash. If the stepper motor is directly driving the ball screw then will an anti backlash nut be sufficient seeing that the travel is only around 20 - 30cm?

[Neale]

20mm rails seem normal on that size of machine. Overkill in terms of strength for the Z axis, but when you start looking at the Z design in detail you are probably going to end up with two parallel plates, one fixed and one moving. 15mm rails don't really give you enough room between the plates for a ballnut. Use 20mm and you will need to cut some clearance slots and pockets, but it's a much easier fit. Belt or direct drive won't make any difference to backlash in a ballnut (assuming that you use a decent quality coupling between motor and ballscrew) - it's only the way that the ballscrew is driven and backlash comes down to the quality (=price!) of the components you use which affects the fit between nut and screw.

[mekanik]

The use of a belt between your ballscrew and motor can help prevent resonance in motor, so would include them on X & Y drive.
Mike

[Norgmonster]

Thanks Mike.

I'm only using a ball screw on the z axis. I'll be use a belt with the pinion stepper motor mounting plate on the X and Y.

[Norgmonster]

When it comes to the Linear Guideways (http://www.hiwin.de/en/Products/Linear_Guideways/4263), what is the differences in performances between a "standard" block and a "flat" block? Could I use standards for the x axis that will be taking the most load and flats for the y?

[Norgmonster]

This is my first stage design for my router. I will add more photos from different sides soon but if you have any comments or questions then let me know.

1874518746187471874818749

[JAZZCNC]

That's ok for the broad strokes of the design but it's the little details that will bite and you don't show any.! You also need to ask your self if copying someone else's Bad complicated design is worth the trouble.?

There are easier simpler designs that will work just as well and be stronger. For instance frame with high sides with simple beam gantry sat directly on rails is much stiffer than high gantry sides and far less complicated.

Word of warning if this is your first machine.? It's common mistake to think that building large machine is just as easy smaller machine. While in essence it appears just matter of scale, which it is to some degree, problems lie in that the obsticles to over come, like rail alignment etc scale upto match so if you've no expereince then expect some serious hair pulling.!!

Unless you really need large machine I advise to cut your teeth on something smaller.

[Norgmonster]

Thanks for the reply Jazz,

As the title suggests this is my first go :-)

My original plan was to make a 2' x 4' bed with the gantry running across the 4' length. That way when the time comes i could just extend the 2' sides to 8' without remaking the hole frame. As you pointed out, this can work in theory but comes with issues to consider.

When you say high sides do you mean on the x axis?

In terms of frame materials, is the 80x80 aluminum profile i used sufficient? I originally considers 80x120.

Should i be using rack and pinion with my plans to extend the frame in the future?

Let me know your thoughts and I'll go back to the drawing board.:-)

[routercnc]

Sides
The high gantry sides introduce a weakness - delete these, raise the longest X axis rails (and supporting aluminium profile beam that it is sitting on) so that they are level with the lowest point on the gantry cross beams. Then re-draw a small gantry end design that links the gantry beams to the X axis rail bearings. This is much stiffer overall as the raised part is now integrated to the bed and can be supported as much as you like.
This would also mean you could delete the EXTRA lower pair of hiwin rails as there is virtually no moment on the gantry ends anymore. This would save alot of money!

Gantry beams
Putting rails on the inside faces of the gantry cross beams makes it difficult to get them aligned/in the same plane. You can't use epoxy leveling for example, and those aluminium beams are not flat. This means you will need to use shims etc. which would be fiddly to get all 4 aligned. Machining them flat to start with might help, but they need to be accurate to prevent binding. These precision guideways do not like much mis-alignment.

Z axis
You have not shown details of the 'box Z design'. Be aware you need to allow for head tramming adjustment, Z rail alignment, and Z rail pre-load /adjustment in the way it all goes together. CAD is 100% perfect, but real world is not.

As per Deans comment if I was starting out I would not go for something quite so complicated.

[Norgmonster]

Hi All,

Thanks for the feedback on my last design. I had a feeling that it would not hold up but it was an opportunity to use Fusion 360 for the first time.:-)

I have taken on your comments and hope that this next one is a step closer.

For what ever reason, a lot of the components in the design could not be rendered to look like they should be so i hope you can tell what it is you are looking at.

The size is approximately 2' x 4' with the idea to extend it to 8' x 4' in the future. The z axis trail will be around 20cm.

I have not done a Z axis yet or properly defined how i plan to mount the stepper motors for the R&P system so beyond that let me know your thoughts.

1877818779187801878118782

[Norgmonster]

I shore this Rack and Pinion mounting setup and was wondering what any of you guys thought of it?

https://www.youtube.com/watch?v=-m_aoWduEXI

Is it worth replicating for my router design?

Any feedback on my design would be great as well (see photos one posting above).

[JAZZCNC]

Yes the principle of spring loading into rack is ok and how it should be done. But wouldn't copy is weak design for router (he's using Plasma so not so much an issue on the cut). Watch how the mounting flexs @ 0.22s when he puts is hand on it.
This flex will show in the machine. Maybe not in accuracy but will in resonance and this resonance transfers thru the machine and into the final cut. Is too bad it will cause excess wear in the pinion/rack.

[Norgmonster]

The flex is down to the plate being quite thin (looks to be around 5-10mm). I have also see other spring mounted systems that use a belt between the stepper motor pinion and another pinion that makes contact to the rack. I assume this reduces kickback?

Could you also have a look at my new designs and let me know what you think (smaller bed, lower gantry and less rails)?

[JAZZCNC]

Ok we are cross purposes here. The Mechanism of being spring loaded is what I thought you was wanting advise on.
The fact it's direct drive off the stepper shaft is because it's being used for plasma machine which doesn't require the same resolution and torque that router requires.

This is why you have seen other systems that use belts n pulleys. It's got nothing to do with Kickback. The belts/pulleys are applying ratio to increase resolution and torque.

[JAZZCNC]

Could you also have a look at my new designs and let me know what you think (smaller bed, lower gantry and less rails)?

It's Ok for general layout and better than last but lots of very important details missing or not thought about.! . . Like. How do you fasten the bearings to the Gantry profile.? . . . Z axis .?? Well no detail there at all.!!

Now given the Short length and to be honest even the Gantry isn't massive wide then suggest you use Ballscrews and save lot of hassle. You'll get much better accuracy and far better reliabilty. Won't cost much more money either.

Personally I'd drop the profile that sits on the bearings and increase the height using taller profile or 2 pieces. Then sit the Gantry profile directly on the bearings fasten to a plate.
I'd also have stiffer gantry. 2 pieces arranged in L config is very stiff. Like this

18787 18788

[Norgmonster]

Got it :-)

I should be looking to make something like this then:

187891879018791

[Norgmonster]

It's Ok for general layout and better than last but lots of very important details missing or not thought about.! . . Like. How do you fasten the bearings to the Gantry profile.? . . . Z axis .?? Well no detail there at all.!!

Now given the Short length and to be honest even the Gantry isn't massive wide then suggest you use Ballscrews and save lot of hassle. You'll get much better accuracy and far better reliabilty. Won't cost much more money either.

Personally I'd drop the profile that sits on the bearings and increase the height using taller profile or 2 pieces. Then sit the Gantry profile directly on the bearings fasten to a plate.
I'd also have stiffer gantry. 2 pieces arranged in L config is very stiff. Like this

18787 18788


Thanks for the feedback on the new design.

I was thinking to fasten the bearings to the gantry profile using bolts and t nuts that slot into the profile groves.

The x axis will be around 100cm and the gantry (y axis) will be around 150cm. Is this considered small?

No z axis because i wanted to address what i had so fare.

I'm using 80x120 on my x axis. To increase the height, the only larger profile is 80x160 in that range. This would add 40mm. Should i go with two 80x80, one 80x160 or somewhere between?

The L shaped gantry in your photo looks to be using two lengths of 80x120 making the gantry 200mm high. Is this right?

My original plan was to make a router that could be upgraded in size (4'x8') when the time came. That's why i was going for R&P and not ball screws. Is this a practical approach considering that the alternative would mean starting from scratch?

Ill make some alterations and post them soon. :-)

Thanks again,

Norgmonster

[JAZZCNC]

The x axis will be around 100cm and the gantry (y axis) will be around 150cm. Is this considered small?

Not small but still within acceptable range of ballscrews without taking special measures.


The L shaped gantry in your photo looks to be using two lengths of 80x120 making the gantry 200mm high. Is this right?

Yes ITEM 80x120


My original plan was to make a router that could be upgraded in size (4'x8') when the time came. That's why i was going for R&P and not ball screws. Is this a practical approach considering that the alternative would mean starting from scratch?

In theory all sounds great but in practice theres that much work and setup goes into machine to start with that you'll probably not want the hassle of converting.
Personaly I'd build with the intent of building another machine when want larger. You'll then be able to put all you've learnt and design changes into the new improved larger machine. Use the First to build second then Sell it complete and working. If do this right you'll make money rather than costing money converting.