Re: tool changer logic- mathematical challenge

Hey,

This is an existing tool turret and optical sensor design on an
Emco cnc lathe. I'm not building it, I'm trying to make it work with
new controls (Mach2 software).

The "C" shape is built in from the factory. It is exactly a 135
degree opening. This creates 2 trip points for each sensor giving a
45 degree resolution with only 4 sensors.

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Jeff Jones" <cnc@d...> wrote:
>
>
> > I think you're missing something :)
> >
> > If the opening is 90+ degrees, then there's a 90- degree span
over which
> > the sensor output doesn't change, so you won't see a 45 degree
change in
> > toolholder position.
> >
> > Consider it this way:
> > ...
>
> Yep, I guess I was missing something. You're wanting to know when
to *stop*
> advancing the tool and back up up to lock it? I assumed you'd run
it forward
> a fixed amount of time, then back to lock and then sense the tool
position to
> make sure you're where you thought you were. Or to pick up the
initial tool
> position.
>
> Does it have to be a "C" shaped thing? How about 5 single holes
> around the ring? Put three at 90 degrees apart and two 45 degrees
from
> where a 4th 90 degree position would be. i.e. starting with 0 at
the top,
> one at 0 degrees, one at 90 degrees and one at 270 degrees and the
> other two 45 on either side of 180 or 135 and 225 degrees.
>
> Now if you're between tool positions no sensors will be tripped.
If you're
> at an even position, three sensors will be tripped with the
two "odd holes"
> 45 degrees from tripping the 4th sensor, uniquely identifying the
position .
> If you're at an odd position, you'll have two sensors tripped with
the three
> 90 degree holes 45 degrees on every side of the other two sensors,
uniquely
> identifying those positions. If you only see one tripped,
something is
> horribly broken.
>
> Jeff