Cheap Linear Bearing Solution

[GarageWoodworks]

I'm still struggling with my X,Y tables for my H-mortiser. I've tried full extension drawer guides and they had way too much vertical slop. I also looked at under-mount slides today at Woodcraft and they too had too much slop.

I am thinking about using 4 steel rods (2 per axis) and using guide bushings. Four bushings total inserted into two separate equal lengths of wood. The wood would support the table (two per axis). If I lube the crap out of the steel rod do you think this would work? Hope this makes sense.

[Robatoy]

Give these guys a look. A lot of their stuff is very reasonable.

[-MIKE-]

I haven't really been following, but the rod/guide description sound a lot like what's in the arm of my RAS. Those concave guide bearings that would fit snug on the proper size steel rod are all over the net and ebay for cheap.

google RADIAL ARM SAW CARRIAGE BEARINGS

[J. Clarke]

You might find it helpful to take a look at the "linear motion" section of vxb bearings: . They've got a wide range of options for not too horrible prices.

[Robatoy]

I have tried similar approaches and found it less than ideal. Mind you, you are not looking for a lot of travel so you might be able to bang the stuff around so it will work for a bit...but linear bearings are a breed on their own. The bushings have the wrong aspect ratio to the rods, even when using two. The slightest bit of skew and you jam.... and you just need one of them to do that by a couple of thou.

I quite like Mike's idea.

[Robert Bonomi]

You might want to look at the commercial offerings from Rollon Corp.

I haven't purchased anything from them, but have had discussions with sales and engineering abut some of their _big_ stuff (multi-ton loads). I found them _most_ helpful, even when dealing with an application that was _way_ outside the scope of what they normally deal with. And this was research _only_ for an admitted one-off project.

Their "mono-rail" has accuracy of +/- 0.1mm in the 'normal' precision units. +0.000/-0.04mm in the really high-precision ones. lengths for single sections, up to roughly 4 meters.

lots of other alternatives, too, if you need something telescoping. Catalog data includes the data to figure out how much play there will be in your application. Assuming you can build to the precis ion of their gear. :)

[MGH]

About six years ago I was dealing with the same problem you are having in making the X, Y tables. My solution was to use KV 1129 center under mount drawer slides because I was given a couple cases of these slides. Using a set of slides for each table gave unacceptable vertical and lateral play. This slide has a carrier holding the ball bearing in position as the two parts move past each other. Seeing I had so many slides I added a second carrier to each of the drawer slide, cut the drawer slide shorter and limited the slide travel to several inches. The result is a short travel slide supported by ball bearing throughout most of its travel with no lateral and vertical play.

Saw dust has not been a problem yet.

Works for me.

MGH

[Larry Jaques]

How were they mounted? Were the X guides in the closed position when the table was up against the Z gantry? Were the Y guides just a couple inches open when the table was centered? Most joints won't be but 2 inches in either direction, so you should have the ball bearing glide stiffness working -for- you. Weight of the table and part should keep them pretty much together, too. (Support that stray end of a long table leg, bed sideboard, or whatever, though.)

If you use old-technology, go with sintered bronze bushings and grease or oil them, not the rods.

If you use linear motion bearings, you won't have slop and they'll have seals around them. They're not cheap, though.

CNC routers usually use 5/8", but for the shorter tables, you could probably get by with 7/16 or 1/2".

-- You can't wait for inspiration. You have to go after it with a club. --Jack London

[Brian Grella]

Got any more? :^P

[Paul Franklin]

Instead of guide bushings, consider linear plain bearings. Ones made from Teflon are not very expensive (3/4" for $5-10). They are long relative to shaft diameter so they avoid the jamming problem mostly.

They don't need lube and can handle high loads as long as you are moving them relatively slowly and can tolerate more friction than you would get with linear ball bearings (which can be quite pricey).

For a bit more money you can get self aligning versions which are a lot more forgiving when mounting....

As this is a key part of your cool project, you don't want something you will be constantly cursing..

HTH,

Paul Franklin

[Larry Jaques]

for a 500mm, but the bearing carriage is 15 times that!

Where's the reasonable stuff? I couldn't find it.

-- You can't wait for inspiration. You have to go after it with a club. --Jack London

[GarageWoodworks]

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 --Jack London

I found a local company that sells sintered plain bearings at $4.50 a pop. Very reasonable IMHO.

rod is $13.50 for 18". i just need to think of a design that uses the least number of plain bearings. 4/axis might be pushing it.?.

[Larry Blanchard]

That actually sounds like a pretty good solution if there isn't too much free play.

You can get a T-bar that fits in a T-slot. If the bar was full length I wonder how that would work?

[Larry Jaques]

Lets see, 4 for each axis, plus a pair of rods for each, comes to $135 plus shipping, which will be heavy.

I still don't see why the ball bearing glides wouldn't work decently.

-- You can't wait for inspiration. You have to go after it with a club. --Jack London

[whit3rd]

Long-extension drawer guides will keep the motion parallel, but aren't intended for precision positioning (there will be a millimeter or three 'slop'). A pre-tensioned linear bearing is expensive (even the circular plastic ferrule gizmos have to be precision-cut, which is not an easy task in plastic).

Can you add something, like a featherboard, to provide a preload force on the drawer glide? A hydraulic tensioner would be ideal, like maybe a spring-assist shock absorber. You'll want two axes, so two such will be required, along with suitable slide plates or rollers to engage the moving part. Some drawer glides engage balls in V-slots, you can use two of these and apply the load pressure between the two glides. The stationary 'plate', then, becomes a ZZZ-frame, with parallelogram type linkage so the glides stay parallel, and the shock to keep the load force will be anti-diagonal to the linkage.

[Swingman]

Therein above, lies the reason the JDS Multi-Router, is an expensive, well engineered piece of precision equipment ... and worth the price for a small production run shop.

[whit3rd]

Yeah, but... a home cruftsman can change a plywood plate to a ZZZ-frame with a Forstner bit and a few quick kerfs... and I bought the shocks for something similar years back, they're still in the basement...

[Swingman]

Yeah, but ... ITMT, I'm using an MR, and yours is still in the basement. ;)

[Robatoy]

IMNSHO, the MR is nothing but a grown-up, more precise and robust Domino. Before I got my router, the Domino would give me a Semi and the MR the Full Chub.

g,d&r

[Swingman]

Eatcherheartout ...

Pssst ... Got a cnc for trade? :)