Re: Digitizers, "Hanging Jack" topology
Posted by
Doug Fortune
on 2001-01-10 22:56:25 UTC
Alan Marconett KM6VV <KM6VV@...> wrote:
slide down (like the converging Mercedes Benz triangle) when the probe
is unsupported by the object being digitized. It would be such that there
could only be the one final resting place (for mechanical repeatability).
Instead of foam, I would suggest a ball bearing soldered/brazed nestled
between the top three contacts would add a bit of downward force.
Further, a light spring soldered to the ball bearing completes the electrical
contact (to the basic jack) and in addition, provides a torsional resistance
to the whole assembly to make sure the jack doesnt rotate (if it does rotate -
the geometry might change perceptably).
'hanging leg' with threads or a nut (or a longer coupling nut) to accept a stylus.
there would be quite a resistance jump higher if you lost one or two
contacts). But digital is probably best.
digital signal (since you always want it to move slow when it is in contact,
and always want it to move fast when it is not in contact).
the time for the overall digitizing (because you won't be spending time
probing where you know the model isn't), and reduce the chance of
crashing (again, I'm thinking of running the probe body sideways into
a vertical edge of the model, causing a crash, and the probe won't be
triggered in that case).
instead of a single down-hanging probe, you could thread on a "star"
(see http://www.fvfowler.com/styli.html - besides the down-hanging
tip, it has 4 more out horizontally left,right,fwd, back) which you could
use to get into cranny's - but for freeware hobby software that is a ways off.
but otherwise, plain XYZ digitizing should work just as well.
I can see myself wandering off to Toys'R'Us tomorrow at lunchtime
for a small shopping expedition....
best regards,
Doug
> Doug,Yes, there would be three "valleys" upon which the three ball contacts
>
> I think we're on the right track! a "jack" (the small toy) is good
> model to use too, I'm just nearer the ocean! The cups would have to
> cause the legs to always "center" the same way. Also, we want to add a
> little force on top of the jack or my "cement breakwater" (probably
> gotta name, anyone?). A piece of foam would probably do nicely.
slide down (like the converging Mercedes Benz triangle) when the probe
is unsupported by the object being digitized. It would be such that there
could only be the one final resting place (for mechanical repeatability).
Instead of foam, I would suggest a ball bearing soldered/brazed nestled
between the top three contacts would add a bit of downward force.
Further, a light spring soldered to the ball bearing completes the electrical
contact (to the basic jack) and in addition, provides a torsional resistance
to the whole assembly to make sure the jack doesnt rotate (if it does rotate -
the geometry might change perceptably).
> Using a "solid" metal part would simplify machining!yea, just buy the toy (a steel one is already conductive), and replace the
'hanging leg' with threads or a nut (or a longer coupling nut) to accept a stylus.
> But would require three circuits to be "AND'ed" or "OR'ed" togetheryup - or possibly have it analog (with three resistances wired in parallel,
there would be quite a resistance jump higher if you lost one or two
contacts). But digital is probably best.
> And the extra gate could drive my LED.sure - in fact that signal could directly be the "move slow / move fast"
digital signal (since you always want it to move slow when it is in contact,
and always want it to move fast when it is not in contact).
> Are you suggesting the first pass creates a Gcode file to "follow", and,Actually, not to improve our readings per se, but to drastically reduce
> by adjusting it a little "tighter" in (tool offset?), we can thus "ride"
> the part the second time, improving our readings? Could be a plan.
the time for the overall digitizing (because you won't be spending time
probing where you know the model isn't), and reduce the chance of
crashing (again, I'm thinking of running the probe body sideways into
a vertical edge of the model, causing a crash, and the probe won't be
triggered in that case).
> The "skirts" of the object being digitized are what bother me. As inSure, with a smart program that knows the results of the "roughing pass",
> the example of the little car, you can't SEE the wheels! Perhaps after
> the "roughing" pass, we learn to "hug" the sides, and get that info?
instead of a single down-hanging probe, you could thread on a "star"
(see http://www.fvfowler.com/styli.html - besides the down-hanging
tip, it has 4 more out horizontally left,right,fwd, back) which you could
use to get into cranny's - but for freeware hobby software that is a ways off.
> I also suspect that using a rotary table will work good for a lot ofPerhaps if it were inclined (to digitize a turbine blade for example),
> objects.
but otherwise, plain XYZ digitizing should work just as well.
I can see myself wandering off to Toys'R'Us tomorrow at lunchtime
for a small shopping expedition....
best regards,
Doug