[PaulH]

I have seen some CNC routers that have the linear bearings under the table like the JGRO design, and other that have them above the table, like the one shown at http://solsylva.com/

What are the pros and cons of these two designs? Is one more desirable than the other for some reason?

[jeffs555]

I like the overhead rail designs like that one. The overhead rail designs(and also the moving gantry designs) only have to be a little larger than the cutting area whereas the moving table designs have to be at least twice as long as the actual cutting area.
The main disadvantage of the overhead rail designs is access to the cutting area for putting the work material onto the table. The moving table will move out from under the fixed gantry so you have easy access to the work area. With the overhead rail designs, you have to reach down inbetween the rails to place the work material. For smaller machines where items being machined are not heavy or bulky this is not much of a problem, and also for a small machine it wouldn't take a very large hinged cover to make it completely enclosed to keep the noise and dust in.

[PaulH]

I would never build a moving table design for exactly the reasons you say. I was going to build the JGRO design, or a variation thereof until I saw the Solsylva design.

Are there any mechanical advantages to one design over the other?

[svenakela]

Shorter "leg length" causes less vibrations in the gantry.
Shorter "leg length" causes smaller fault tolerances if there's a play in the linear guide.

But on the other hand you need a more complex drive unit.

--S

[PaulH]

Cool! Now that's the kind of feedback I was looking for. Thanks, Svenakela!

[jdownie]

The driven mass is closer to the drive axis on the "top slide" design. This means that it is easier to achieve rigidity from the ballscrews to the router. There are , however, more planes to align, and more material is used.

Have a look at this one. Then tell me you don't want one. Go on. Tell me.

http://www.rainnea.com/cnc_5axis.htm

[PaulH]

*drooooooooooooooooooolz*

Yeah, I'll take one of those! Thanks for the link. I had seen this before but didn't remember it in relation to this discussion.

[envcons]

Another bonus with the overheads is that its easy to put things under the work surface, like a vacuum system for holding sheet metal.

[sol]

Okay, so I am biased, but from my experience the two huge problems with homemade routers are the flexing of the materials made worse by the lousy mechanical advantages.

The best way to minimize flex is to reduce the length of the components and to limit the number of connecting joints. With the dual leadscrew over-table system the gantry ends are short so there is less material to flex, and there are fewer connections between the lead-nut and the load at the spindle.

Regarding mechanical advantage, it is desirable to keep the force, fulcrum and load as close to the same vector as possible. With the dual leadscrew over-table rail machine this is the case, the lead-nuts, rails and spindle are near the same plane.

With an under table leadscrew and rail system, the force has to wind from the leadscrew to the rail, up to the gantry and finally to the spindle. This long lever arm invites amplification of backlash and rocking on the rails which also allows chatter that can be quite severe.

If all of the materials used were rock solid, the longer plates and convoluted force paths would not be such a problem, but when the tables are made of MDF and skate bearings these mechanical disadvantages become magnified. So for homemade routers I am a strong proponent of the over table system, obviously.

[sol]

In the picture is worth a thousand words department . . .
The red components are the primary force and load elements in the X axis; the closer these are clustered together the less leverage there is for flex and slop to develop.
A rehash with a couple more images is here.
http://www.solsylva.com/cnc/mechanics.html