[turpija]

It's pretty simple and straight-forward design with as less complications as possible ...

So let's take a look at pictures as it will save me few thousand words





So, the green shafts are hardened and polished steel, pretty ordinary shaft, the lowest ones (X axis) are 25mm diameter, the middle (Y axis) 20mm, and Z axis will probably be either 20 or 16mm. The length of long green ones, would be around 90cm, and i hope there wouldn't be problem with deformation or twisting or vibration ... should it be bigger/stiffer ?

The blue things are obviously lead screws, still can't decide would the ordinary shaft cut as screw do the job ? and if I found nice and cheap ballscrews, I'll switch to those immidietly

Grey parts will probably all be alluminum.

Still thinking weather to use two motors for X axis, or one and toothed belt ...
I was thinking about 1.6Nm torque for motors, as they are quite affordable.

Targeted work area would be 70x70x12cm

Design is of course not finished as I try to find as much part needed for project locally so I adjust it every here and there

If anyone has some suggestion why something should be different then how I imagine it, please don't be afraid to curse my concept

[toastydeath]

I don't like to see ground shaft used as a guide surface unless it is supported along the length. Unsupported shafting means your structural loop is absolutely huge. It will deform badly, at least from what I consider a reasonable set of expectations for a machine tool. Vibration is more complicated issue, but you are going to set yourself up for chatter and other vibration modes. You may not experience them all the time, but you will experience them far faster than you would otherwise.

I would get pins, cross-drill the shaft, and use short cylindrical pillars on bolts to hold up the ways (if you are DEDICATED to using shafting). I would mill a slot in the bushings to accomodate the width of the pillars. This isn't a great idea either, but it will make your machine far more rigid and drastically decrease the structural loop length.

[turmite]

Nice design work and I love the transparent Rhino model. I personally am not a supporter of unsupported rails as I think they will flex. I have seen very solid machines move more than you could imagine when a person just leans on them and this is reading the flex with a dial indicator. I would personally try to figure a way to support the rails. I am intending to build a very similar machine later this year and I plan to use toothed belts as all my drive compontents. Faster, quiter and less expensive than ball screws, and, they are not nearly as sensitive to dust.

Good luck with your build.

Mike

[turpija]

I saw more the few machines with unsupported shafts so I'll leave this one also like that, supporting it would needs for more modifications and additional expense, and after I finish it I'll see if it needs support, and if it does, I'll add it.

And i was thinking about use one motor for X axis and belt drive the other screw side ... as most of kits (electronics) are constructed to work with 3 axis machine, this might be less complicated way to go.

[Rhodan]

I saw more the few machines with unsupported shafts so I'll leave this one also like that, supporting it would needs for more modifications and additional expense, and after I finish it I'll see if it needs support, and if it does, I'll add it.

And i was thinking about use one motor for X axis and belt drive the other screw side ... as most of kits (electronics) are constructed to work with 3 axis machine, this might be less complicated way to go.

You'll be sorry!

I bought a K2 2514 (25"x14" working area) and its great except for the unsupported shafting. On the Z it doesn't seem to be a problem (very short 3/4 inch thick) by the Y is a major issue. Cutting MDF with a 1-1/2" V-bit and plunging 1/8th inch deep at 10 IPM I get a divot where the router has tilted away from the Y. Hell, the dust collector skirt (bristles) was enough to tilt the router the moment the brushes touch the table. I replaced that with clear shower curtain plastic and even that would deflect the router unless I cut all the corners so it wouldn't resist the downward pressure.

I just recently re-made the Y with some Thomson 1/2" supported rails and open bearings (not really the proper use for them but I had them handy). Now I can plung 1/4" deep at 30 IPM without divots providing I do so at the ends of the X axis. The X is also un-supported so at the center of travel I'm limited to about 15 IPM for a 1/4" plunge. Most of the deflection from the X is verticle but get some force in there and the Z axis tilts backwards as the X rails bend.

What prompted me to rebuild the Y axis was the horrid vibration, chatter, and deflection when trying to cut aluminium. Didn't matter how well I bolted down the work, anything more than a .002 cut at 10 IPM would start the vibrations and chattering. Broke a 1/4 upspiral attempting to cut 0.005!

With the Y replaced I can now cut 0.030" with a 1/8th upspiral at 20 IPM before things start going south. I'm in the process of figuring out how to get supported shafts or rails on that X axis now.

Of course, if your application is something like cutting foam or circuit boards then this really isn't a problem.

P.S. Plan on two mounting brackets for the spindle. If you get any kind of speed going its going to want to tilt the router. With two brackets a couple inches apart this problem disappears.

P.P.S In another thread one person is lamenting the fact that he didn't make his X bearing carriage long enough and had terrible racking problems with the gantry on a double screw setup. I don't know what the magic ratio is but my single screw driven gantry has 3/4" round bearings on the X that are further apart than yours... Total length is roughly 6 inches with the two carriages about 21" apart. Maybe a 1:3 ratio? So if your width between bearings is 100 CM maybe making the end-to-end length of the bearings around 30-35 CM would be safer?

[Mcgyver]

I saw more the few machines with unsupported shafts so I'll leave this one also like that.

you're done a beautifully rendering, but not addressed the forces and what acceptable movement of the structure. this is basic to design..... no matter how what you use, supported or unsupported rods or a 2 ton cast iron box way, it will flex. what counts is making sure that it has less flex than your design objectives. how much flex are you willing to put up with, and have you calculated how much this design will flex? That will tell you whether it is a good design or not. Intuitively though like others have said, I'd guess that long unsupported span like that will be like a guitar string. check out wigitmasters machines for ideas on using support shafting while still using low cost components

[spalm]

Sorry man, but I agree, you need to support those rails.

Steve

[turpija]

Appreciate all suggestions, especially Rhodan, as he has similar machine !

I'm not sure how much this machine would flex, but original use is for engraving mainly plastic, and cutting it to dimension (1.6mm thick), and engraving brass, and possibly aluminum. I was unsure about using unsupported shaft for 80-90cm of length, and I now see it was not the smartest idea but for this purpose I think it should work out.
As I don't know how to calculate exact deflections of shafts, how much actually can flex ? at middle of shaft ?

2 Rhodan: how thick are shafts on your machine 1/2", 3/4" ?

And one important thing about (un)supporting shafts ... 50% bigger price for bearings for supported shaft and additional machining required, as I don't have tools to do machining myself I will have to pay for it in local machining shop. And budget is low, after all this is my first machine (hopefully not last), and it won't be perfect no matter how hard I try.

[High Seas]

With engraving and plastic as your target product and materials, your loads should be low. You didn't mention how tight you want your resolution.
An option - if you can source it - would be something like 8020, with slides. Not as "tight" as some materials - but certainly fully supported and gives you the frame structure at the same time. Just something else to add to your considerations! Oh Yeah - you'd have to make some more real cool drawings to show that option!!!
:cheers: and, Welcome aboard! Jim

[Rhodan]

My shafts were all 3/4" and let me tell you, it was (and still is) quite easy to deflect the router.

Right now I have the open bearings and supported rails on the Y (holding the Z on the gantry) and unsupported on the X (holding the gantry) and I can still tilt the router with just a bit of hand pressure.

I have some IKO rails that I'm going to swap with the thomsons then I'll use four supported shafts and open bearings to replace the X unsupported shafts. I'd rather have two 31" supported rails but instead I have four 18" so I'll have to line two up end-to-end on each side. Getting them perfectly straight I think is going to be a pain...

The machine won't be useless without supported rails. With some care and keeping speeds fairly low (40IPM max for routing and 5-10 for plunges) you won't have too much problem with woods or plastics, but you can basically forget aluminium. Not sure about brass as I haven't tried it.

Oh, you can see a video of my machine carving a sign on youtube with the unsupported rails. It'll give you an idea of how slow you need to go to avoid too much deflection. It still had little divots where I plunged but they were small enough that only I noticed them.

http://www.youtube.com/watch?v=M9fG8trDLxQ

http://www.youtube.com/watch?v=xtmUbquYuuI

This sign took around an hour to cut. After replacing the Y rails I did the same thing in 1/4 the time and even I couldn't see the divots.

[turpija]

Thanks ALOT Rhodan, those movies are great !

It definitely shows machine, and which looks very usable and I didn't think of much higher speeds on my machine. 1 hour for that sign is quite fair amount of time, and if I could get those results with my machine I would be more than happy with, considering budget I have for it.

2 High Seas: It's not easy to find parts I require for this machine, and when I do find it, it cost alot. About resolution, I didn't think of that, it depends of screws I find, but with those up to 1/8 microstep drivers it will be more than enough. And real cool drawings, hm, I did quite a few with various options, until I get straight and try to make it as simple as possible, yet functional.
But real "cool drawing" would be picture of finished machine, that's for sure

I think I'll go for test build out of MDF first to check for possible flaws in design, and if it works out, aluminum is next ...
All I need is to source all required parts, and off course get the funds all together

[Rhino3Master]

Hi guys, newbie here. It's my first post on the boards.
Nice Rhino enders, turpija!
Here is what I would do if the router was mine. It's a 3-6mm thick aluminum plate with standard triangulation, welded (or glued, or bolted) to the tube rail. Adds lots of strength and eliminates the flex. And you won't need any tube bearings (I don't know their exact name). Just use cheap ordinary bearings - they will do excellent job.
http://i217.photobucket.com/albums/c...n/aae05ba0.png
http://i217.photobucket.com/albums/c...n/e5c050a2.png
http://i217.photobucket.com/albums/c...n/c2dff2cf.png
http://i217.photobucket.com/albums/c...n/9201400b.png
http://i217.photobucket.com/albums/c...n/86334050.png

Mirror that construction if you need even more stability:
http://i217.photobucket.com/albums/c...n/0c13dd52.png

Thanks,

Bobi


PS: Hi Mike, I'm glad to see you here :cheers:

[turmite]

Hi Bobi,

Been here since it opened or within a couple of weeks. Great site and people.

Mike

[turpija]

Been some time brainstorming, and come up with quite simple and possibly effective way to support rails, picture ...



Yellow things are ordinary screws, was thinking about 6-8mm diameter, ordinary stuff ...

And to compensate cost X shaft will be 20mm, instead 25, cause bearings (open type) are cheaper in that option, and with supporting them I don't think it should be a problem ...

[solodex2151]

I think you might find that it will be extraordinary difficult to drill those holes perfectly straight with one another in a round object like the pipe, even if you did have a drill press and a jig. It sounds like you are going with commercial bearings anyway. Why not pick this bearing system and save yourself the hassle...

Also, on another not it looks like you have two blocks that are wedge against each other to attach to the gantry support. Why not space the blocks apart a little to get a wider base?

[turpija]

I think you might find that it will be extraordinary difficult to drill those holes perfectly straight with one another in a round object like the pipe, even if you did have a drill press and a jig. It sounds like you are going with commercial bearings anyway. Why not pick this bearing system and save yourself the hassle...

The backup solution if holes won't go straight in shaft is to mill the surface of shaft so it would be straight then drill holes there.

That bearing system on link would save a hassle but not wallet, and i don't think i could get it here, and if i do, it would certainly be far more expensive than on that link.

Also, on another not it looks like you have two blocks that are wedge against each other to attach to the gantry support. Why not space the blocks apart a little to get a wider base?

Thats because of material i can get, aluminum 100x100 mm, so if i widen them more it would me more machining involved -> more money, more chance of mistake, and incorrections.

[Rhodan]

Glad to see you're making progress!

I saw an engineering paper in one of the other threads (sorry, can't find it now), that said for parallel bearings on a CNC machine you need 1.6:1 ratio for stability. So if the rails are 16" apart you need a length of 10" on your beariings.

If driven from the center the leadscrew acts like a guide and you can raise the ratio a bit (didn't say how much). But the higher the ratio the more force your leadscrew will have to handle to keep the gantry from racking.

With twin leadscrews I imagine the ratio can be much higher. The higher the ratio though the more critical it becomes to get the two leadscrews exactly in sync.

The other consideration for the bearing spacing is the ability of the gantry to handle plunges. The less "wheelbase" you have the more torque the long axis rails have to handle.

[Rhino3Master]

Hi Bobi,

Been here since it opened or within a couple of weeks. Great site and people.

Mike

Hi again, Mike. I'm amazed how small is the World. I just searched for CNC machines on Google, and during reading that topic I accidentally saw your nickname here

According the topic name - I think that the leadscrew's height level must be located on the approximate height of the potential parts (i.e. in the midddle of the parts' height). Thus, the forces will be as low as possible, and the whole moving part (the motor and the cutter of the CNC machine) will be in its maximal stability.

Bobi

[turpija]

According the topic name - I think that the leadscrew's height level must be located on the approximate height of the potential parts (i.e. in the midddle of the parts' height). Thus, the forces will be as low as possible, and the whole moving part (the motor and the cutter of the CNC machine) will be in its maximal stability.

Makes sense, but i can't change much design considering leadscrew position, so it will stay like this ... and i think it should be enough, after all it's not some tough machine, it is suppose to be engraver at first place, and router at second

[Rhino3Master]

Placing the leadscrew on the same level as the rails will eliminate the need of stability ratio, caused from the different level between them.

http://img218.imageshack.us/img218/216/cnc7th9.png