Re: X-Y Table, Rotary Table for CNC INFO

Ray,

Thank you. As I mentioned before, I'm only pursuing this as a thought
experiment, since I believe the economics won't work out. But
someone else may not agree. So here's some more...

(added on re-read. If you ARE interested in this as a real item,
please see the PS below this message.)

In the simple single scissor I have in my head, there would be a
SIN/COS relationship to leadscrew nut movement versus vertical
movement.

The single X of the scissor is arranged this way: (of course there
are two X's, and the nut would act on the lower leg of both, but for
the example, we'll just look at one of the X's)

pivot points at left bottom, crossing point, and left top. Nut is
attached to right bottom, with motor hanging out past the left
bottom. (The leadscrew remains horizontal at all positions.) The
right top will slide "captured" in a slot. A ball bearing carrier
with 2 or 3 bearings riding on either sidewall of the slot will allow
this final pivot point to be in line with the left top (which is
needed to preserve accuracy), while eliminating this otherwise
possible source of backlash. Similarly, the right bottom legs of the
x will need to travel without vertical backlash in a straight line
(the axis of the leadscrew).

Now here's what we have. At exactly 45 degrees (the X legs are all at
right angles to each other), one unit of movement of the leadscrew
wil result in an equal vertical movement. Less than this, it will
take "more" leadscrew, and above this point it will take "Less"
leadscrew movement. (following the SIN/COS angular relationship) So
let's keep the error "single ended" by restricting ourselves to less
than 45 degrees... We choose "less than" so that the table can be
kept compact vertically.

Anyway, Hope this makes sense. This is a very "buildable" method for
any who wish to pursue it. I would be glad to help any who want
to "go there" with comments or ideas offlist.

Ballendo

P.s. I actually have a pretty complete design for this in my head.
I'd like to hear from anyone who thinks this might be an item worth
having: How much table force/load would you want? What "accuracy"
would you need to have to consider REALLY buying one of these? And
finally, At what price would you REALLY buy one, given that your
specs were met? (Speedwise, I'm thinking that 100 IPM would be easily
done.)

--- In CAD_CAM_EDM_DRO@yahoogroups.com, Ray Henry <rehenry@u...>
wrote:

>
> Ballendo
>
> The z motion compared to the rotation of the screw is really riding

along

> an arc or a cam. The leadscrew comp produces a linear compensation
> between points. This should be okay for most of the arc but like

you say

> it would be big as you approached the top.
>
> Till Franitza built a little hexapod using a a crank shaft and

connecting

> rod to get linear motion. This would be very similar to the motion

of

> the scissor table. If we get serious about this, I can see if we

can get

> him to explain it to us.
>
> Ray
>
>
>
> On Saturday 18 January 2003 04:56 am, Ballendo wrote:
> > Message: 2
> >    Date: Sat, 18 Jan 2003 04:18:32 -0000
> >    From: "ballendo <ballendo@y...>" <ballendo@y...>
> > Subject: Re: X-Y Table, Rotary Table for CNC INFO
> >
> > Ray,
> >
> > Good point.
> > But it would be repeatable, and therefore possible to correct in
> > software. Possibly even with EMC's leadscrew correction, though

the

> > error is "squished at one end" of the travel.
> >
> > Ballendo
> >
> > --- In CAD_CAM_EDM_DRO@yahoogroups.com, Ray Henry <rehenry@u...>
> >
> > wrote:
> > > 
> > > >    From: "ballendo <ballendo@y...>" <ballendo@y...>
> > > > Subject: Re: X-Y Table, Rotary Table for CNC INFO
> > > <s>
> > > > Think scissor lift table. A couple of accurately placed

pivots;

> >
> > ONE
> >
> > > > screw, lots of force. Problem: Stability in the direction

perp. to

> > > > the scissor. Fix that with a vertical guidepost (or two. Note

that

> > > > these DON'T need to be heavy duty linear motion component

quality;

> > > > they are just keeping the structure stable in one axis

(resisting

> >
> > only
> >
> > > > side loads).
> > > 
> > > With a scissor you will need to be a bit careful where you place
> >
> > the
> >
> > > screw.  The usual place is in the middle between pivot arms and
> > > perpendicular to the motion.  Many car jacks are made this

way.  

> >
> > With
> >
> > > something like this, z motion would NOT be linearly related to
> >
> > rotation
> >
> > > of the screw.
> > > 
> > > Ray