Re: lead screw or rack and pinion ?

Dave,

I too have been designing a table for a 4 x 8 router (orginally
moving table.... but have since changed to moving gantry due to floor
space considerations). After considering many options I'm going to
go with a welded steel frame, using various structural steel shapes
mostly C channel and square tubing. Of course, that will still leave
a lot of work to prepare the mounting surfaces for the 25mm linear
rails, but I plan to grind/scrape using a mechanist level (I'll
probobly cheat and end using some shim material as well <g>). The
main thing I'm analysing as I'm designing the gantry and side beams
is the "stiffness" vs. weight of the steel members. Too heavy a
gantry and it greatly adds cost to the servo/drive system, but of
course the more rigid the better. It's no doubt better to analysis
the whole frame using finite analysis tools, but I'm lacking those so
I'm just calculating deflections for various beam / force combos
using a simple spread sheet. which I'll upload. Really more of
a "what if" game than a real engineering analysis. You can get the
moment of inertia figures for common shapes from here:

http://ourworld.compuserve.com/homepages/MJVanVoorhis/techdata.htm

One thing that became very clear after playing around with the
numbers is that aluminum extrusions are not very stiff, unless they
get VERY large. I'm not really sure what should be considered
an "acceptable" deflection as the gantry and head change position or
as cutting forces on the tool change, but I decided rather
arbitrarily that I would try to keep all deflections under .001 In.
I doubt this would be good enough for a metal machining table, but
since I'm mostly looking to work with wood and plastic, I figure it
will be good enough for my very limited budget.

If you want to get some idea of what a minimum solution might look
like, check out http://www.shopbottools.com I'm sure you're planning
on something better, but I do like the "horizontal" gantry approach.
I'm still debating the pros/cons for the gantry layout, most
commercial machines use a "vertical" approach of one rail over the
other, but I like the wide tool bay offered by a twin beam horizotal
gantry. Also depends how much Z travel you want. I'm planning on 12
inches, evem though most of what I want to do will be flat panel
work, because it really doesn't cost much extra at this stage.

With regard to drive choice.... I still haven't decided yet. I've
ruled out rotating lead or ballscrews because of the lenght of X axis
and the corresponding size of a drive system needed, but I am
considering a rotating nut design. Alternately, rack and pinion
looks pretty good as well, especially if you ratio down the motor RPM
using a gearhead or a toothed belt arrangement. I do want to use DC
brushless servos , which develop more power at higher RPM, hence not
a good idea to drive the pinion directly. But if you are using
steppers... well the shopbot folks seem to have made it work.

Just my 3 cents worth.... this is my first large format router as
well... so I'm not nearly as experienced as others on this list.
But I hope it's helpful all the same.

-Craig Edwards

--- In CAD_CAM_EDM_DRO@y..., "turbulatordude" <davemucha@j...> wrote:
> Hi All,
>
> I am persuing the design of a 8ft x 4 ft router table, moving
gantry
> style and am looking at the best way to move the table.
>
> seems the consensus so far is to power both sides to prevent
racking
> of the moving gantry.
>
> the question is wheather it is better to use lead screws, driven by
> timing belts, lead screws driven by gears and solid shafting, or
rack
> an pinion driven by timing belts or gears and solid shafting.
>
> of course alternate methods are welcomed. I'm in the design stages
> and it looks like I am going with Hi Win linear rails and bearings.
> (Mclem@ ??? on the list offers a good deal.
>
> I was thinking of using NEMA34 steppers to drive each axis, geared
> down as speed is not too important (yet) and it looks like the
table
> is simple to build. I checked out welded tables and the twisting
is
> too much to handle. seems like you have to learn a multitude to
> subjects to get a good result in this hobby. (Mariss makes driving
> the steppers just too easy)
>
> Dave