CNC design for consideration

Hi all,

I just wanted maybe a more experienced builder, to have a look at my design and see if there is anything obvious that might hold my design back before I commit. I posted on cnczone and took all comments on a previous design on board and have now come up with this one. Figured it would be more beneficial to be a member of a UK forum. I am already a CNC owner too so not a complete noob, although it would be naive of me to say I'm not still learning.

The goal was to create it as cheap as possible but try and keep rigidity. I've used 12mm multi-start trapezoidal lead screws on x, y and 10mm on z. Fully supported 16mm rails on x and y and 16mm round rail on z. My main concern is the lead screws on y and the 16mm rail. I used that size because it is easier to implement with the size of aluminium profile I'm using. The machine is 1500mm x 1000mm at the rails, ignore the length of z as im going to shorten it anyway. I designed it with a view to implement a fifth axis at a later date. I will be cutting MDF and PLY only.

Any advise would be greatly appreciated. Thanks

http://i280.photobucket.com/albums/k...09at124441.png

Hi Jon,

No problem on the double post's, I have taken care of this for you. Not sure if the link to 1500mm x 1000mm is supposed to be there so I have left it.

Good luck with the build, as noted I would consider going with 20mm rails and if you can Ball screws on each axis. 16mm will work as you know but I would too rather see 20mm rails over a 1500mm length, you could always go to 80x80 profile for the Y axis and this would give you the 40mm centre’s the rail would require for mounting, however 80x80 is not cheap I know so...

Preparation for a 4th axis at this stage would be a good idea, people have been known to bolt a 4th axis to the y table as and when they need it.

Regards,
.Me

Great, I've been looking at ballscrews on zapp and the difference in cost is not really that great. I could easily go with 16mm ballscrews and that just leaves the 16mm-20mm rails on y to sort which shouldn't pose too big a problem. My only question is, can the ballscrews be mounted directly into a flanged bearing with 16mm opening and a shaft coupling at the other end without machining?

Quote:

---

Originally Posted by jonbabbz

Great, I've been looking at ballscrews on zapp and the difference in cost is not really that great. I could easily go with 16mm ballscrews and that just leaves the 16mm-20mm rails on y to sort which shouldn't pose too big a problem. My only question is, can the ballscrews be mounted directly into a flanged bearing with 16mm opening and a shaft coupling at the other end without machining?

---

most of us buy 1605 screws from china, "linearmotionbearings2008" (ebay) he will machine the ends for you... its worth giving him a shout for a quote

im sure theres a way of mounting without machining, not sure it would be worth the effort though to save £20 ??

That's great thankyou. Agreed for £20 it would be worth getting them machined. I was just referring to zapp who charge £75 for machining. That would add a considerable amount to the build cost.

Yea that guy on ebay is ok to deal with, i got ballscrews off him a few moons ago and they turned up in less then 4 weeks. Didn't know he dose maching now as well, as mark said it would be better to have the ends turned down to what you need rather then go about it the other way.

Here is a few pics to give you some ideas of what you could do.

Quote:

---

Originally Posted by jonbabbz

That's great thankyou. Agreed for £20 it would be worth getting them machined. I was just referring to zapp who charge £75 for machining. That would add a considerable amount to the build cost.

---

iv no idea what chi (linearmotionbearings2008) charges for machining... i do remember him being nice n cheap and good comms :)... i dont remember delivery taking ages either

it would be silly not to give him a shout and see whats what :)

I normally get a quick order from Chi, 7 to 10 days max but the last half dozen orders have generated a VAT plus handling charge from Fedex, usually about £15, not excessive but worth bearing in mind.

You do have to contact him, tell him just what you want and he'll put that exact order up on Ebay and send you the link. If by some chance someone else buys it before you just tell him and he'll list a duplicate order. The guy is 110% reliable.

Thanks for all your help guys, really appreciate it. Asked for a quote, so just waiting now.

Ok so Chai has replied with a quote. For the 4 rails, 8 blocks, and three machined ball screw assemblies 529usd shipped?!? That seems awfully cheap to me. Does that sound right? I'm about to rip his arm off.

Quote:

---

Originally Posted by jonbabbz

Ok so Chai has replied with a quote. For the 4 rails, 8 blocks, and three machined ball screw assemblies 529usd shipped?!? That seems awfully cheap to me. Does that sound right? I'm about to rip his arm off.

---

told you it would be silly not giving him a shout :)

only three screws ?? if you are not using two screws on your x axis id have a good think about the implications before you press buy
that extra screw will more than earn its keep!

your design shows four screws? have you already got one for your Z axis?

That sounds OK to me, as I said before he's 110% genuine.

Quote:

---

Originally Posted by blackburn mark

only three screws ?? if you are not using two screws

---

Yes! Definitely get 2 for the X axis...
Price sounds about right for Chai. I always pay him directly via a paypal invoice as I think he knocks a bit off due to no eBay fees.

Quote:

---

Originally Posted by John S

I normally get a quick order from Chi, 7 to 10 days max but the last half dozen orders have generated a VAT plus handling charge from Fedex, usually about £15, not excessive but worth bearing in mind.

---

Customs recently changed the rules - now everything above £18 value stated on the form gets charged. I don't really mind paying the VAT but £10 handling fee is just extortionate. Last time I just tried paying the VAT (£4.08) and not the handling charge (£10)... so far so good!
Expect to pay another £20 or so since it's a reasonable size order.

Quote:

---

Originally Posted by Jonathan

I don't really mind paying the VAT but £10 handling fee is just extortionate. Last time I just tried paying the VAT (£4.08) and not the handling charge (£10)... so far so good!

---

If you don't want to pay the £10 then nip down to Heathrow and collect them and pay the VAT there.
Let us know how long it takes you and how much petrol / train fare is

Quote:

---

Originally Posted by John S

If you don't want to pay the £10 then nip down to Heathrow and collect them and pay the VAT there.
Let us know how long it takes you and how much petrol / train fare is

---

I've already paid for shipping to my door, not Heathrow, so the £10 appears to be to send one letter and collect the money. For the most part it will be an automated system so I don't see how it can cost them £10 to do that - they're clearly trying to make a bit of money. The handling charges used to be much less, and before that they were illegal - I'd happily pay £3-4. Even sending the letter via Royal mail doesn't cost that much!

Sorry, didn't realise it was illegal for a company to try to make a bit of money.
You could always collect from Chi

So its all ordered then! Very nice guy. He sent me the standard drawings for the machined ends and I had a peruse to see if they would fit my machine and wouldn't. So I simply emailed him my requirements and he has not charged me anything extra! I will definitely do business with him again. The screws I ordered are just for x and y. i need to redesign the z axis because I'm not sure I want to go with round rail. I think there will be too much deflection at the cutter so ill just buy another smaller ballscrew when I'm ready for it. I want to change the bed too. Here's what I ordered screw wise. I'll create a build log when I get all my parts together. Can't wait :)

http://i280.photobucket.com/albums/k...chinedends.jpg

Quote:

---

Originally Posted by jonbabbz

Hi all,

I just wanted maybe a more experienced builder, to have a look at my design and see if there is anything obvious that might hold my design back before I commit. I posted on cnczone and took all comments on a previous design on board and have now come up with this one. Figured it would be more beneficial to be a member of a UK forum. I am already a CNC owner too so not a complete noob, although it would be naive of me to say I'm not still learning.

The goal was to create it as cheap as possible but try and keep rigidity. I've used 12mm multi-start trapezoidal lead screws on x, y and 10mm on z. Fully supported 16mm rails on x and y and 16mm round rail on z. My main concern is the lead screws on y and the 16mm rail. I used that size because it is easier to implement with the size of aluminium profile I'm using. The machine is 1500mm x 1000mm at the rails, ignore the length of z as im going to shorten it anyway. I designed it with a view to implement a fifth axis at a later date. I will be cutting MDF and PLY only.

Any advise would be greatly appreciated. Thanks

http://i280.photobucket.com/albums/k...09at124441.png

---

Hi The Y axis looks like it is set on one side; if it is, this will jam as the axis drive needs to be central or with two stepper drives simultaneously operated one on each side of the machine. I would also consider widening the Y axis bearings so the machine has less Y wobble. As you are only cutting MDF and Plywood accuracy is not an isse here that design is good enough. If you can squeese a bit more out of your budget, go for the linier bearings as in the photo.
Attachment 5909 These are far more accurate and rigid and will increase rigidity and production speed. Keeping this design square in the X & Y plane is going to be dificult.

Quote:

---

Originally Posted by John S

You could always collect from Chi

---

Anyone got an address for Chai? PM please. (EDIT: Thanks, got it now)

Hi guys, been really busy lately so updates will be few and far between. Im collecting my parts together at the moment but just wanted to show you a revised y and z axis on my machine and see what you think. Ive moved the rails above and below and gone for supported on z. Ive tried to make z as light and thin as possible so I can get more movement on y. As for the previous design LotusPack, it is driven from both sides you just cant see it because of the angle of the drawing that's all. I considered going for profile rails but I don't want to spend that amount on a first build. Maybe the next one ;)
http://i280.photobucket.com/albums/k...16at153204.png

Quote:

---

Originally Posted by jonbabbz

Hi guys, been really busy lately so updates will be few and far between. Im collecting my parts together at the moment but just wanted to show you a revised y and z axis on my machine and see what you think. Ive moved the rails above and below and gone for supported on z. Ive tried to make z as light and thin as possible so I can get more movement on y. As for the previous design LotusPack, it is driven from both sides you just cant see it because of the angle of the drawing that's all. I considered going for profile rails but I don't want to spend that amount on a first build. Maybe the next one ;)
http://i280.photobucket.com/albums/k...16at153204.png

---

Hi CNC design firend, this looks really good, have a few points to consider on your design:
1) The Y Axis is very close to the X axis (A), going on the diagram there seems to be little space under the X axis for your work piece, (Work Area my sketch) if the Y axis bearing is level with the work table.
2) The Z axis feed drive plate is working outside the bearing range (C), 4 bearings below the Z leadscrew
3) Can't quite see how the Y axix bearings will work (D) assume there are two bars 1 top and 1 bottom
4) It would be better to increase the X axis bearing distance (B) at each end, the wider the stronger, although you will lose some travel distance.
Attachment 5993

The tool stress area is best kept inside the bearing array, if possible.

hope it goes well...

Quote:

---

Originally Posted by LotusPack

1) The Y Axis is very close to the X axis (A), going on the diagram there seems to be little space under the X axis for your work piece, (Work Area my sketch) if the Y axis bearing is level with the work table.
2) The Z axis feed drive plate is working outside the bearing range (C), 4 bearings below the Z leadscrew
3) Can't quite see how the Y axix bearings will work (D) assume there are two bars 1 top and 1 bottom
4) It would be better to increase the X axis bearing distance (B) at each end, the wider the stronger, although you will lose some travel distance.

---

1) The bed is well below the X-rails and presumably its height will be adjustable. Having that distance as small as possible is excellent to maintain rigidity.
2) That shouldn't make a difference.
4) Yes it would be better to space the X-bearings out more. What is their current spacing?

1. Ah yes, I see, that sounds good.
2. Ok cool but the further away the Z plate drive point is from the bearings the more compound the tolerance will be on the drive forces. You are right it should not affect the travel accuracy.

The Z axis travel looks quite short compared to the X&Y if you are designing for large surfaces and need to jump up and down the job work plane will this have enough for your need? I originally had 250mm Z distance and have since increased to 400mm for those extra long tools and thicker jobs. My Z axis runs from the collet face at 20mm above the table to 420mm +Z my machine has a fixed 200mm throat under the gantry. This now allows for quite a long tool for example drilling through 200mm you obviously need at least 410mm travel. As the -Z movement starts to put the pressure on when drilling the whole gantry structure starts to go under stress, which is why I advocate centering the spindle in middle of the Y axis bearings, if it is off centre the gantry start to tilt back against all those bearing tolerances and pulls the drill out of square and jams up the drill.

I have a distance of around 400mm between the bearings on a 1.3m Y axis total length so I get an affective length of approx 900mm which is perfect for the standard linear scale I have fitted.

This is a good design as most fixed table machines have the "Throat" restriction under the gantry. How are you going to adjust the table up and down and keep things square? I would think a secondary Z knee adjustment for the work piece height, suggest a manual mechanism that can be locked after part fitting otherwise although flexible just like a knee milling machine, will bring in further rigidity issues.

The table is going to be fixed but I have left room for sacrificial packing on the workspace to build it up if need be. The point about widening the bearing spacings was on my to do list and has now been done. The z axis question im not quite sure I understand what you mean. Here are a few more pics to show exactly how it is in its entirety. I have yet to put a lot of holes and strengthening brackets on but I can do that as I go along with the build. I will be moving onto building the z axis next week :)

Z AXIS DOWN
http://i280.photobucket.com/albums/k...16at224341.png
Z AXIS UP
http://i280.photobucket.com/albums/k...16at224552.png
Y Bearings and rails
http://i280.photobucket.com/albums/k...16at224437.png
Side View
http://i280.photobucket.com/albums/k...16at224401.png

You've done well to get the Z-axis that narrow, but the problem is the Y-axis bearing spacing is now too small. Forces parallel to Y will tend to twist the Z-axis, so you need the bearings spaced out further to resist this. Clearly you'll have to either sacrifice travel or make the frame wider, neither of which is ideal but that is surely better than the whole machine being compromised for the sake of a bit more travel on Y?

So what spacing would you suggest Jonathan? It's a valid point but I would have thought any forces trying to twist it on the y axis would be counteracted by the bearing diagonally opposite as it is. Obviously bigger spacing would make it more durable, but I can't widen the actual machine any more as I already have my screws and rails. I guess I'll just have to sacrifice travel on y.

Quote:

---

Originally Posted by jonbabbz

So what spacing would you suggest Jonathan? It's a valid point but I would have thought any forces trying to twist it on the y axis would be counteracted by the bearing diagonally opposite as it is. Obviously bigger spacing would make it more durable, but I can't widen the actual machine any more as I already have my screws and rails. I guess I'll just have to sacrifice travel on y.

---

i would at least try to fit them flush with those side plates... the tool deflection forces are not your main worry, what would worry me is resonance
if the resonance from cutting finds the sweat spot on your Z the forces will be masive compared to the cutting force and will result in a crap finish
it may only take a little extra stiffness in a design to prevent your typical router resonance getting a grip.... so in my opinion, every tiny little bit of stiffness you can squeeze out of the bugger will pay


imagine you push on a lamp post... your not going to cause it any mischeif... then you push and pull the lamp post with the same force in a opportune rhythm you can get close to wrecking it... with this in mind the problem becomes a little more stark

in an ideal world i guess you would space your bearings at least as wide as you have them tall and if not, as wide as you can

Hi Jon,
These are all point of ease of use really.
Now I see the whole picture, you have addressed quite a few limitations on standard design for rigidity, etc. But there is always a compromise on access and travel distances. I can also see there is room for linier gauges if you want them later.

Because of the limit or gap between the bed and under the gantry, I am assuming that the Z axis travel to the bed is 200mm for my example, 2nd picture shows the tool down. Now looking at the first picture, tool up, the travel is only going to be as high as the Z ball screw length. So my suggestion is to double the bed to gantry gap size and use that as the ball screw length (400mm). This will allow you to use long tools that need to machine through the 200mm. So the Z ball screw needs to be 400mm to cater for the machine dimensions.

One other point on the use of the machine, being as you will have to work over the Y axis rails for job setting it could be quite difficult to set a job “inside this area”. Also as the machine has sides if you have a larger piece of work there needs to be an opportunity to hang the work over the edge. The X axis should be ok for long jobs and you have given enough to push the work through the X axis and machine the next section providing you leave the gap at the back. I have just recently machined a kitchen cooker hob aperture in a very long worktop, much bigger that the machine length. The Y is always a restriction so I made sure I could cope with at least 700mm so I can cut outside the standard 600mm units on the machine Y length. Your design seems to handle all this ok.

The last picture “Side View” shows the elevation on the bearing widths quite well; there you can see the spindle in line with the front Y bearing, this is where you could increase the end piece length to move that bearing to the same distance as the one at the back from the spindle centre and give yourself a equal pressure on these bearing in the Z axis when tool pressure is applied. I know that will reduce the Y axis working distance. You have said that this point was already on you to do list and now done. :-)

How are you going to set, adjust and maintain the X, Y and Z squareness?
The nice thing about two Y axis ball screws is you could independently adjust to fine tune the X and Y squareness. The Y Z plane can be adjusted ok using the fixing into the end pieces but need to consider how to adjust the X Z plane as this will be dependent on the construction. It may be worth thinking about this because when finished getting it all square can be a nightmare of rebuilding with shims, etc. This of course depends how accurate you want is all to be. Of course you could bolt it all together and it may come out spot on for you, in my experience a Gantry Mill is always out of square and needs some mechanism for fine adjustment.

Other minor points
Another spindle (Motor) brace at the top of the motor
Axis stepper cable management particularly for the X behind the gantry

It is looking good, nice CAD what do you use?
Regards

It does make sense to space them out with regards to resonance. This is the reason I wanted a second opinion, thank you. The z travel is only 90mm now. I was going to go for a longer travel but decided to keep it simple for a first build. Squareness shouldn't really be an issue as the t slot nuts can be moved/adjusted then tightened so that doesn't really concern me. As I said before, high resolution finishes are not my main goal(I'll be building furniture mainly), but a certain degree is essential. For cable management I'll simply be using a piece of angle aluminium on the back of y with a train on it. There are a lot of items I'll be using that are not on the drawing just yet. I think the biggest mistake was ordering parts before I was completely finished with the design so it kind of limits my adjustments now. Oh well, you live you learn. Nothing that can't be sorted. The cad software is Rhino 5

Ok so I think I'm finally happy with it now. Ive done a complete redesign of the z axis and widened it enough so that I can still get the amount of travel in y I need. Ive also managed to get 130mm travel in z doing it this way so it actually works out better. I also wont need to put my existing z ballscrew on the lathe to shorten it :)
http://i280.photobucket.com/albums/k...17at223112.png
http://i280.photobucket.com/albums/k...17at223723.png

Hi Jon,
I think in the main you have been brilliant with your design changes and the discussion are hopefully positive. There are many little areas of improvement that can be and have applied at design stage and it is nice to see that you are willing to gain other opinion even if not so relevant to your need. Most who are in your position are just looking for recognition and take any suggestion as a criticism. Well done and hope this turns out to be a successful machine. When you get onto the electronics and controls, if you haven’t already, we can all hopefully help to get the best most economical in the market.
Regards
Paul

I have been involved in engineering for all my working life but am the first to admit that I don't know everything. This to me is an interesting ans probably most important part of the build and if I can get it right first time then great. As the old saying goes, 2 heads are better than one and you guys have already taught me so many essential requirements to designing a successful machine. Thanks guys, now onto the easy part.....the build ;)

Hi Jon,

Looking good now. Like the compact Z axis, reminds me somewhat of an idea I was playing with last year . . .
Attachment 6013

Hi guys, rather than start a new thread I may as well just use this one for my build. I've made a significant amount of progress and didn't really want to keep posting every time I tighten a bolt or tap a thread. So here she is so far. I've decided to make the bed and x axis out of 60mm square box steel instead of Ali. Mainly to keep costs down and strength. I also fancy having a bash at welding as I've never tried it before :)
http://i280.photobucket.com/albums/k...709_205746.jpg