Re: [CAD_CAM_EDM_DRO] Re: Digitizers, Renishaw

"Roman Black" <fastvid@...> wrote:

> How about a vertical hanging probe, with
> like a disc at the top. The disc sits on/in
> a ring. The exact mech design I leave to
> the experts!

How about instead a "jack", something like
Alan Marconett KM6VV <KM6VV@...>
suggested:

>The tips of the star are indeed conductive, but I'm thinking the three
>"arm" contacts need to be insulated from each other. 3 pole, single
>throw. Connect the switches in series.

>I'm with you on the delrin body, and ball bearings, and depressions for
>the balls. the shape of the "star" may indeed be angled. I'm thinking
>of the "break water" cement pyramids, 4 legs, three of which are always
>down, one up. In our case turn it all upside down, the single arm now
>is thin and long, the other three angle up, and contact the three pairs
>of balls.

The "jack" has 3 upward balls (electrical contacts) and one
downward ball (touch probe contact). The thing normally hangs
down by gravity, suspended on three "cup" electrical contacts.
The jacks downward leg is replaced by a standard ruby sensing stylus.

However, the cups are not wired in series, but instead the jack is
supplied with power (say TTL 5Volts) by a flexible wire. When
the touch probe contacts the object, at least one of the supports (cup)
must loose contact (because it is lifted up). (Determining which
support lost power might aid in determining any slope of the object,
but that is not a big feature).

There could be quite a large "safety zone', maybe 1/4" or more, because
there is no upward limit to the movement of the jack (you have to make
sure the housing doesn't crash though), which leads to an advantage,
because the downward movement, over generally unknown distances,
can be fast (assuming you can stop fast vertical travel within (say) 1/4"
of first noticing the electrical break).

Then slowly the probe is raised, and you THEN determine the position
of the head when gravity allows the tripod to settle back into its 3 supports
(and energizes the 3 cups). The advantage is that you are going up slowly only
for 1/4" max, whereas the downward movement can be at relatively high speed
over long distances.

Also, I suggest two passes (somewhat akin to a roughing pass and a finishing
pass), first with a BIG ball (ie ball bigger than the diameter of the housing,
to get a general idea of the shape, especially the verticals), and then a final
finishing pass with a sufficiently small ball for the required detail. The
software of course takes into account the shape given by the roughing pass
to generate a suitably optimized trajectory for the detail pass.

Doug Fortune