[Oldmanandhistoy]

You mean 90° counter clockwise? Then I'd have to mount it to the table top (or flush to it) instead of the edge. No welding, no steel, aluminum only. No room for more tools.

Yes counter clock wise.

If you were to use steel and have some one track weld L plates for you then you could increase the gap between the vertical bearings. This would then allow you to increase the height of the rail so it would have a larger contact area with the side of the table.

Also which is probably very obvious to you; you could use the same system for your Y axis fixing rails to a torsion box beam.

Imho you would end up with a very ridged system.

John

[jsage]

Ger21,

I was playing around with the idea yesterday. The dimensions on this drawing is off not optimal but presents two options. The first top and bottom inset make it not too different from a standard rail.

The t slot was another option with a double bearing. Perhaps spring dampened or tensioned.

I was originally thinking your rail was inside not outside. Perhaps inside would be easier.


I name it the double option sideways pi

I added a better visualization still missing some things.

regards

[Oldmanandhistoy]

Ger21,

I was playing around with the idea yesterday. The dimensions on this drawing is off not optimal but presents two options. The first top and bottom inset make it not too different from a standard rail.

The t slot was another option with a double bearing. Perhaps spring dampened or tensioned.

I was originally thinking your rail was inside not outside. Perhaps inside would be easier.


I name it the double option sideways pi

I added a better visualization still missing some things.

regards

Interesting option but imho I still think Ger21’s G arrangement is the way to go with my suggestions added in.

John

[jsage]

John,

Yep, The 90 degree flip makes it conventional. Very hard to recreate the wheel.

[jsage]

Wait a second... If I flip mine 90 then I like it better Beauty is in they eye of the beholder or drafter in this case. ; ) Just kidding.

[ger21]

My original goals here were to also be easy to build with basic woodworking tools, and also be relatively inexpensive. While the original plan meets those goals, adding more bearings and angle starts to get a bit tricky. I'm not sure I'd want to deviate from my original design, but the limit of that design would be in the 48"-60" range.

For a fully supported table, I really like what Ted is doing here. http://www.cnczone.com/forums/showthread.php?t=33124

The V rollers and eccentric bushings would run about $175 for the X and Y axis. I'm pretty sure I'd spend at least $100 on a skate bearing design. I'm going to order some of these V rollers from www.cadcamcadcam.com and have a good look at them.

[jsage]

I think the problem you run into is despite the rigidity of the screws you are still stretched out with all the force on the screws. I think what your saying john if I stand inside the router and reach to one side put my thumb on top of the side support and curl my hand around to the bottom of the outside you optimize your distribution of forces.

[jsage]

GER21,

Can't disagree with you there. http://www.cnczone.com/forums/showth...?postid=252016

If i had turned it 90 degrees I would come up with same answer ; )

I think you could do more complex profiles but you would half to split it and glue it with epoxy then plane your flat sides.

[Oldmanandhistoy]

My original goals here were to also be easy to build with basic woodworking tools, and also be relatively inexpensive. While the original plan meets those goals, adding more bearings and angle starts to get a bit tricky. I'm not sure I'd want to deviate from my original design, but the limit of that design would be in the 48"-60" range.

For a fully supported table, I really like what Ted is doing here. http://www.cnczone.com/forums/showthread.php?t=33124

The V rollers and eccentric bushings would run about $175 for the X and Y axis. I'm pretty sure I'd spend at least $100 on a skate bearing design. I'm going to order some of these V rollers from www.cadcamcadcam.com and have a good look at them.

I have been and had a look at Tom’s design and like it a lot. With your original design remit in mind I would agree than the first design you proposed is the place to stop. I have the habit of looking at a design and then trying to figure out how to get the absolute best form it. This does tend to over complicate a clean simple solution which is not always of benefit and becomes over engineered.

My thoughts on round pipe and skate bearing are that for simplicity of design and build it is an excellent option but imho using three bearings set at 0, 90 and 180 degrees is a must for maximum benefit with that system. I have used the 0, 90, 180 system myself and am at present building another machine which will incorporate it in the X axis only. In my experience the round pipe very quickly becomes no longer just a round pipe because the bearings form flat surfaces as they compress the steel under them. This then again imho makes the system at least equal to a V bearing design and possible even better because of the amount of contact area between bearings and rail. So it comes down to cost, build time and complexity of the design only.

In summary, again imho:

If you have the time but a low budget go with skate bearings and ger21’s bearing set up in the first post of this thread.

If the extra cost is not an issue and you want the machine up and running as soon as possible but you do not have a bottomless pocket. Then the V bearing would be the way to go.

There are other options but if you are reading this thread then they will not concern you as they are expensive.

Jerry I like your original design and I am sure it would make a very nice system but personally I would use your bearing set up but instead of your rail I would use round pipe for simplicity of design.

John

[sdantonio]

Iteresting design Gerry, definitely will have to keep it in mind for the next machine upgrade.

[Oldmanandhistoy]

A question for you ger21 just for my educational benefit and any one else who reads this thread.

What in your opinion are the advantages of using your rail system compared to the more traditional round pipe system?

I value your opinion as some one whom has devoted a lot of your time to the art of home built CNC machines.

John

[rippersoft]

If you take 2 skate bbearings and place them side by side on the same screw with a spacer in the middle, you have a poor man's v-bearing.

Adjust the spacer for the amount of width required to "hug" the rail.

Works real good on angle iron, unistrut (u-channel), and may even work on some v-rails.

RipperSoftware

[ger21]

What in your opinion are the advantages of using your rail system compared to the more traditional round pipe system?

With the easy way of using bearings mounted to angle on round pipe, the loads on the bearings are at a 45° angle. On my router, I use threaded rod to pull the bearings tightly to the pipe. But with the 45° load, the bearings qickly start to bind, so you can't really put much preload on the system. And this preloading adds a lot of friction.

And also, when you do put a lot of preload on the bearings, the bearings will quicly wear flat spots onto the pipes due to the small contact point.

Both of these issues showed up as soon as I mounted my gantry. And neither exist with this design. The bearings carry radial loads only, which they're good at, and the contact area is much greater on the flat rail opposed to round ones.

I'm trying to develop a very solid machine that any woodworker with a decently equipped shop can build, without needing a CNC to cut the parts out. I hope to have my router running this fall, then I'll stat building a prototype of the new design. I'd like to have a full, complete set of drawings before I start, so I'm starting now. I have a lot of projects I need to complete before I get there, but I can work on drawings a little each night.

[Oldmanandhistoy]

What would be your opinion on running three bearings at 0, 90 and 180 degrees on round rail as I discussed in my post #29 ?

John

With the easy way of using bearings mounted to angle on round pipe, the loads on the bearings are at a 45° angle. On my router, I use threaded rod to pull the bearings tightly to the pipe. But with the 45° load, the bearings qickly start to bind, so you can't really put much preload on the system. And this preloading adds a lot of friction.

And also, when you do put a lot of preload on the bearings, the bearings will quicly wear flat spots onto the pipes due to the small contact point.

Both of these issues showed up as soon as I mounted my gantry. And neither exist with this design. The bearings carry radial loads only, which they're good at, and the contact area is much greater on the flat rail opposed to round ones.

I'm trying to develop a very solid machine that any woodworker with a decently equipped shop can build, without needing a CNC to cut the parts out. I hope to have my router running this fall, then I'll stat building a prototype of the new design. I'd like to have a full, complete set of drawings before I start, so I'm starting now. I have a lot of projects I need to complete before I get there, but I can work on drawings a little each night.

[Oldmanandhistoy]

Two advantages I see in using your rail design over pipe is that the flat is already there for the bearing so will give good contact immediately. This would eliminate the time necessary for flats to be made on pipe and consequently save time readjusting preload. It would also give greater accuracy as you would not get an uneven surface as you might with round rail once the flat is there.

John

[Oldmanandhistoy]

I just thought of another advantage; you could use hardened flat bar which would greatly increase longevity of the system.

Ok your system is the winner on this occasion.

John

[ger21]

Round pipes can also be hard to mount straight, unless you CNC cut the brackets.

[Oldmanandhistoy]

Now I have to give you top marks for your original design and imho your system would be better than the V bearing approach.

But you loose all the marks because you posted it too late for me as I have just today finished my X axis which uses a double pipe system and it’s too late to change it now.

John

[sdantonio]

One of the things with the alternative that would have to be taken into consideration. If you compare the new proposal with an existing pipe machine... say Joe's for example. Joe has 8 threaded rods used as adjusters running pretty much just to the outside of each corner of each individual tortion box. In his design these were incorporated into the ribs that hold the pipes. In Gerry's alternative we would have to think about how to keep those features since they do help in straightening out any warps that develope in the system over time.

My other question would be, how much stiffness does each of the pipes conrtibute to the system as compared to the flat steel tacks, and how detrimental would it be (if any) to loosing that stiffness. Especially for those of us with fairly sizable tables.

On the other hand, several people have reported wear on the pipes and the need to rotate or even change them out. In Gerry's new proposal, changing out the steel flatstock would be a breeze.

I do like John's idea of hardened flat stock. I tried pricing out some 440c pipe and almost had a stroke whan I saw the price.

[sdantonio]

Round pipes can also be hard to mount straight, unless you CNC cut the brackets.

I can attest to that