large DiY 3-axis. need guidance

[Loz]

Hi everyone, I'm wanting to build a large machine, 2m x 1m which for those who use imperial is roughly 80" x 40". this machine will be used with a variety of materials, alluminium, copper, perspex, MDF, foam and a few others. I'll be making a large variety of items which will vary allot in size, from car parts, to pc parts, circuit boards and allot more.

Initially I've been looking at the designs on buildyourcnc.com. but one of the guys there thinks I might have issues with the linear bearings on a larger machine. Here's some pics of my first design. the whole machine will be made of steel.

pic 1 is of the bed, I should mention that it will be a gantry machine. but I'm sure you'll all work that out after looking at the pics. this design uses 40x40mm box for the frame of the bed, 25mm tube for the rails which there are 2 of on each side and some custom made bolts which I can make on a laithe to attach the rails to the bed. the bolts are 100mm in from either end, and 450mm appart. depending on what changes need to be made I'm thinking of expanding this a bit so that I get a 2x1m cutting area, not 2x1m overall size.


end view of rails


This is the first rough design for the linear bearings. the bearings are 120 deg from each other.


end pic


let me know your thoughts and don't hold back. I don't care if you insult my designs as long as at the end of the day I end up with a better machine.

[stef_d]

i wouldnt use ball bearings and rods on such a large machine.
they are already difficult to setup correctly on a smaller scale. also, your carriages are very pretty. but impossible to tune and adjust.
how can you garantee to 0.01mm that your 6 bearings will contact the rod ?
for this size machine, i would save up and purchase proper linear guides.

another point: with your welded frame, any twist or misalignement is really going to have a major impact on the smooth runing of your machine.
I cannot see how you could adjust it if you were only a 10th of a millimeter out on your rod alignments.

i am no expert, but i would advise you to start again !
(sorry!)
I hope for you that someone will reply that there is in fact no issue with your design !

[jsheerin]

I would only go with bearings like that if nicer ones are not in your budget and agree that you need some adjustment mechanism built in to them. Also, I'd make them more beefy, less carved away into fancy shapes.

For your frame, dividing it into triangles is a good way to brace it. However I'd recommend making the triangles smaller to give your horizontal beams some support at places besides their ends. Think about what a bridge looks like - lots of small triangles.

[Loz]

thanks guys,

I'm now in the process of re-designing the bearings, I'm thinking of using a U shaped bearing carrier for each individual bearing that slots into the outer shell that can be adjusted using a grub screw from the outside of the shell through to the carrier that will push the bearing against the rail. I should be able to get about 5mm travel out of each bearing. I'm also thinking about changing it so the rails are on top, rather than on the sides so only 1 either side rather than 2. I'll put some designs up later to give you a better idea.

with the frame, it's impossible to get anything perfectly square and flat when welding, the expansion/contraction from the heat will make it move a little, hopefully my welding technique can reduce the ammount then there are other ways of overcoming the slightest twist etc after the welding is done.

thank's for your comments on the bracing for the frame, that was actually one of the things I really wasn't sure about. if I was to do a few more triangles do you think I could use 20x20 instead of 40x40? to reduce the weight and that way I can drop all the bracing down 10mm from the top so that I can put adjustable supports at different points throughout the bed to overcome the above mentioned warping. enough from me for now. revised designs soon.

one other question though, if I wanted to have a 2m x 1m useable area, how wide apart would the rails need to be? it doesn't need to be exact, just close to 2x1m.

[Jesse B]

Originally Posted by Loz one other question though, if I wanted to have a 2m x 1m useable area, how wide apart would the rails need to be? it doesn't need to be exact, just close to 2x1m.

For the Y-Axis, you'll just need to subtract the width of the Z-Axis assembly. If it's, say, 15cm wide, you'd want to add that much width to the Y-Axis (1.15m). As for the X-Axis, you'll just need the width of the gantry. Say it's 25cm, you'd want 2.25m. How much bigger the machine needs to be depends on the sizes of all the other axes.

I strongly encourage you to use proper support rails and pillow blocks. They can be had for quite cheap. There is a store on eBay (linearmotionbearings2008) who sells them for quite cheap. I just feel it'd be easier and more accurate having professionally designed and built linear guides. Just keep your eyes open for deals.

[hemsworthlad]

Hi Loz,

I agree with Stef, save up and buy rails.!!
Your just going to cause your self lots of grief with this design and all though your bearing carriages look Cool they'll be a pain in the arse to make with very little adjustment if any at all.?
With such a large gantry spanning over such along length you will struggle like hell with alignment and sticking issues, even with precision linear rails over this length it's critical to align them properly else stickage can happen and with your idea it will be 10X harder and you will have 2 per side to contened with.
Not only will you have to deal with making sure they are perfectly parallel and level to the opposite side you with have to make sure they are perfectly parallel to each other, which will make setting up your gantry a proper Bit.ch
And thats not even taking into account adjusting them so there perfectly flat/level.
The base frame and rails are the foundation for the whole of the rest of the machine and if any of it is out of wak then every thing else you attach to it will be, so it's critical that you have either quality rails with precise alignment or if you make your own you have a very very good easy system of adjustment.? . . . Because you'll be doing a lot of tweaking and adjusting which you'll be doing often for a good while untill the machine as settled down and worn off all the little high n low spots.

IMO if you want a good accurate machine forget this idea and if you do still go this root then i can tell you for sure forget trying to cut Aluminium or copper of any reasonable thickness to any sort of acceptable accurecy.

Sorry to be so negative but you did say dont hold back and i am speaking from experience here not just a theory i have.
When you go large so do the problems which makes the tolerances even more critical if you want good accurecy and repeatabilty.

My machine as all the right stuff, Precision rails, ballscrews etc and i can tell you it took me days and weeks of minute adjustments and shimming just to get it to the point where i could even think off starting to consider cutting aluminium, it would eat any wood for breakfast straight of the bat but when you start on the hard stuff it's a all differant ball game.? . . . If you have a weak spot it will show instantly on Aluminium.!!

Good luck.
Dean.

[metalworkz]

Hello Loz,

I had started my build with a DIY bearing design and actually have several of the bearing slides made, but then realized the price for the round rails was going to blow the little bit I saved away. I found it much better to watch eBay and found some great Thomson supported rails and carriages for my use for not much more than what I was looking at for the round rails un-supported alone. The bargains are not always there but watch or run a search for what you want regularly and you might see some great deals go bye. If you do proceed with the DIY bearings you should be able to make the idea work by adding adjustment slots for the bearing centers etc., but your bearing slide design could use some gussets or other method for stiffening the unit. Good luck with your build, and keep us posted with your progress
Regards,

[jsheerin]

If you can afford them, the THK linear rails will support a lot more force than even the round rails (there are other cheaper types as well like Hiwin). THK has some good info on this in their catalog on their webpage. These are definitely pricier, but deals on ebay can be had.

[Loz]

thanks for your honest opinions, so the general consensus is that this design will be too difficult to accurately tune. ultimately this will be a functional tool that helps me with my other projects/hobbies so will need to be able to work with a variety of materials accurately. in saying that, should I drop this design, and either start from scratch or go with a more well known design?.

one that I've been pointed towards is the MechMate. has anyone got any experience with this machine? what I was thinking of doing was putting the design into solidworks, then putting it out here for the pro's to make some suggestions be it minor or major on how to improve the machine.

[hemsworthlad]

Originally Posted by Loz ultimately this will be a functional tool that helps me with my other projects/hobbies so will need to be able to work with a variety of materials accurately. in saying that, should I drop this design, and either start from scratch or go with a more well known design?.

Loz, When you say accurate how accurate do you want it to be.? and in what materials.?
The Gantry style router/mill is never going to be super accurate compared to a fixed gantry and to have it really accurate for the cutting the likes of Aluminium and copper it takes a lot of careful thought and the use of quality components.

You are always going to have to find a happy medium for a multipupose machine like you are wanting. To cut wood and plastics you need more speed than accuracy(to reduce burning etc) and for metals you need rigidity and accuracy and in all of these you need repeatabilty.
The solution on how to achieve this for a multiprpose machine is a difficult thing to get just right and even more so for a larger machine.
The problems are, for speed you need eithier fast screws or R&P and in your intended size your probably looking at R&P, problem with this is the resolution drops their for so does your really fine accuracy.
The reverse is true if you want really fine accuracy and precision, it's difficult to have speed because you need slow fine pitch screws also the extra rigidity and weight to overcome requires larger motors etc making it an expensive job.

Dont get me wrong I'm not saying you cant have both, what i am saying is it's not as easy as it first appears and to do it right you will need to put lots of time in and research the best ways to get out what you want.
You may find to get the best working machine you will have to accept that you cant or wont be able to do everything you want on one machine.?

Just remember even the best designed machine in the world will suffer if you use components that are not upto the job and vise versa the best components in the world wont make a apeth of differance if the foundations are not set right.?

IMO and my advice would be you Look long and hard at what it is you want to do the most off and work towards designing or building a machine best suited to this task.

Hope this helps and i cant comment on Mech mate because i,ve never seen 1 in action or the flesh but a lot of people like them and use them successfully day in day out for there business so they cant be all that bad.
That said i think they are mainly used in wood shops.? . . . But dont quote me.!!
Like all things you can look at any machine and see the good and bad in every one, even if you've built it your self. . . . I'm constantly looking at mine and thinking IF ONLY. .?? . Or . . .I WONDER..?

Study every option untill your 101% sure thats what you want then go for it.!!

Good luck.!!
Dean

[Loz]

Thanks Dean,

I'm not sure what the normal tolerances are with a cnc, understand all I really know is what it does, and I understand mechanics & electronics from other hobbies which I can transfer to this. so in saying that the sort of tolerance I'm looking for is +-.1 to .2mm. I don't know if that will be completely unreasonable or extremely easy to achieve.

I'm not sure what style of cnc I want to go with yet, the gantry style was the first I saw and I liked the look of it so that's where my first ideas came from. is a fixed gantry style machine one where the bed moves along the x axis and the gantry takes care of the y and z axis? the only thing that would worry me about this is the extra space requirements. I'd need to have 2m extra clearance at the end of the bed to be able to fully extend the x axis. but I suppose if I position it right in my garrage I can just open the garrage door every time I use it and have it stick out the garrage lol.

if you tell me my expectations are going to be dificult to achieve with this size machine, then I might go down the path of building a smaller one first, then a larger if needed later.

[hemsworthlad]

Originally Posted by Loz Thanks Dean, I'm not sure what the normal tolerances are with a cnc, understand all I really know is what it does, and I understand mechanics & electronics from other hobbies which I can transfer to this. so in saying that the sort of tolerance I'm looking for is +-.1 to .2mm. I don't know if that will be completely unreasonable or extremely easy to achieve. I'm not sure what style of cnc I want to go with yet, the gantry style was the first I saw and I liked the look of it so that's where my first ideas came from. is a fixed gantry style machine one where the bed moves along the x axis and the gantry takes care of the y and z axis? the only thing that would worry me about this is the extra space requirements. I'd need to have 2m extra clearance at the end of the bed to be able to fully extend the x axis. but I suppose if I position it right in my garrage I can just open the garrage door every time I use it and have it stick out the garrage lol. if you tell me my expectations are going to be dificult to achieve with this size machine, then I might go down the path of building a smaller one first, then a larger if needed later.

Hi Loz,

Yes 0.1 to 0.2mm is very easy to achieve and to be honest wouldnt be much of a machine if it didnt, esp for machineing alloys. . . look to be in the 0.01-0.05 region at least.
To give you some idea the theoretical resolution of my 5mm pitch ballscrews at 10x stepping is 0.0025mm but i doubt a get any where near that and i havent got anthing that can measure to that accuracy to check.?

Your right about the fixed gantry taking to much space and this why i didnt go that root as well, this is my build but is a lot further on than this post and i need to update. http://www.cnczone.com/forums/showth...t=92404&page=4

It's difficult to answer whether a smaller machine will be easier to start with than a large machine, suppose a lot depends on you and your abilitys.
The principles are pritty much the same large or small.
You still need to workout the the figures to determine motor sizes etc so you'll just be working with larger numbers that all.?

If you have any doubts about your abilty to work on a larger scale then start at a size that your more comfortable with.?

Good luck.