Re:Re: What to do with a robot!

Hi Ray,

How's progress on EMC's handling of the rotary axis issue ?
I've recently been working on an algorithm for calculating the
distance that a tool moves during 4-axis operation, but it requires
an input of the distance between the centre of rotation of the A-axis
and the tool's zero point position. For example, this value would be
the radius of part if the tool rests on its circumference when its
zeroed. Once the distance is known, the feedrate can be adjusted on a
per move basis to better suit the rotary movement. Initial tests in
Mach1 seem pretty good and I'd think that it would only require a
small change to work with EMC or any other package.
I'll need to try and extend this principle to 5-axis operation but
that'll need the tool lenght or more specifically, the distance from
the toolpoint / zeropoint to the centre of rotation of the 5th axis.

Regards,
Rab

ps. I had a bit of fun working with Tux, I decided that he's a well
balanced peguin for home-brew CNC work, with a precision micrometer
in one hand and a very large hammer in the other :)

--- In CAD_CAM_EDM_DRO@y..., Ray Henry <rehenry@u...> wrote:
>
> Hi Rab
>
> I would think that they would not be affected by the rotary axis
issue
> but r-p-w or a-b-c were not working the last time I was near the
cable
> hexapod. Till did build a little hex with a footprint the size of
a
> piece of paper. It ran from all rotary axis definitions. The
advantage
> I think, is that motor (joint) and xyz (world) systems are
separated by
> the kinematics of a Stewart Platform or it's relatives.
>
> Again I don't believe that anyone has tried it but the tool length
from
> the tbl file should be available. Just how it would act with a 15+
> degree tilt to the table is well beyond my kin.
>
> Ray
>
> BTW - I dearly treasure my EMC mousepad. Your graphic invention is
> awesome!
>
>
> >    From: "rainnea" <rainnea@b...>
> > Subject: Re:Re: What to do with a robot!
> >
> > Thanks Ray,
> >
> > That's quite an interesting machine that Till has built, I'll
need to
> > take a closer look at the EMC sources. I take it that the
robot /  
> > hexapod kinematic modules will be uneffected by the angular axis
> > issues of the standard XYZA module as they must calculate the axes
> > moves with full knowledge of the axis inter-relationships. Do you
> > know if there is a tool-length input in these module's GUI's ?
> >
> > Rab