Re: [CAD_CAM_EDM_DRO] Re: RotoBot hexapod

>
>
>>Does anyone have any ideas on how to implement this approach, in a
>>home built machine? I mean, I could probably go to a machine shop
>>
>>
>I do think that this and the hexaglide/triaglide solutions are more
>realistic for the amateur. I have to say that although the
>rotational method looks elegant I can't think of a way of doing it on
>a real budget. You really want a rack around the circumference and a
>motor with pinion on each of the rotating arms. Fairly normal
>bearings in the middle and some sort of roller system to stop each of
>the arms from flexing upwards or downwards.
>

I guess it would depend on the scale of the machine...external gears are
commonly available in fairly large sizes. I'm thinking in terms of a
benchtop machine with a fairly small work envelope. Someone else
suggested an excellent possibility -- an IC engine flywheel. It's
heavy, accurate, well made, and easily available. I've not looked yet,
but I'd be willing to bet that McMaster-Carr will also have something
usable. Not sure if they'd qualify as "budget" though.

There was a discussion, about a week ago, of how to make an antibacklash
pinion for a rack&pinion drive system. Not sure if it was on this list,
or the CNCZONE.COM site I've mentioned before. Anyway, it looks as if
that's fairly straightforward, too -- two thin pinion gears, and a
spring between them that holds them under some tension.

Electrical connections could be done with rotating couplings -- bushings
or bearings, centrally located, would make easy connection points. Even
something like PCBoard could be pressed into service, for this aspect of
it. Since part of its design is a rotational component, the wiring will
have to be considered ahead of time.

Any ideas for making cheaper ball-socket joints? Maybe a u-joint from a
socket set (too much backlash -- I already looked) or something equally
obvious? Unless I'm completely mistaken, it appears that the common
balljoint at the platform end of the struts is coplanar -- it functions
as nothing more than a hinge between the struts, and not a 360 degree
pivot. The connction to the platform could also be just a hinge. I'm
not sure how well that would end up working, but it would give the
needed flexibility...

I'm not "intimately familiar" with hexapod movements...what range of
movement is necessary, for each joint? From their drawings, it appears
that the "worst case scenario" would be reached at two points -- struts
touching at the base, and struts 180 degrees apart. It looks like a
total range of motion of 90 degrees or so...45 degrees to either
side...is all that would be required.

And, a question. Thinking about it, it would seem that a hexabot's
resolution would be variable, depending on leg geometry. Is this the
case, and is it a problem?

>If it counts for anything, I'm up for it.
>
>If the machine is there, the software will follow.
>
>

Yeah, but how quickly? Who wants to have a cutting edge machine tool,
and then have to write the control routines, from scratch? (Especially
if there already exists an alternative) ;-)

-- Chuck Knight