Re: Digitizing arm - CMM

On Thu, 13 May 1999 21:38:41 -0700, Tom Kulaga <tkulaga@...>
wrote:

>Oops, it's a 5-axis. I didn't know that the wrist had 2 DOF. This
>makes it more difficult, at least from a hardware point of view.

Weeeell, mebe, mebe not (at least, mebe not much more). Lemme know if
you still feel that way after looking at both the MicroScribe and Faro
patents.

>Can that be looked up at IBM's patent page? And does anyone have the
>URL close at hand? I can't find it at the moment.

Sheesh, I forgot to include it, as some aren't aware of what a great
resource that is:
www.patents.ibm.com

>The Rhino page says that the software directly interfaces with the
>digitizing arm. That tells me how they do the surfacing stuff, but
>I'll need to get info from Rhino as to just how the arm needs to
>communicate with the program.

and...

>Anyone have a good
>friend at Rhino, or do you know if they'd share the required info
>with mere hobbyists? I'll drop them an email if I have to, but if
>we've got someone with an "in", we'd be best to use it.

You musta missed my later post, until after you had typed this. Exactly
what I had suggested, in your last sentence above.

>We need to stipulate that Gar's advice ain't the law, it's not the
>advice of proper legal counsel, and you follow it at your own risk.
>
>Thought I'd add that last line after I saw that Gar didn't cover his
>butt in his last post. (That's one ya owe me there, pal ;)

Uh, yeah, OK, here ya go, "I'm not a lawyer, nor do I play one on TV".
How's that?

>A side thought I had about this involves the initial setup of the arm
>and how it might be tricky. I was thinking that if we could build a
>fixture that held the arm in a known position, you could put the arm
>into the fixture when you started the program, and know exactly what
>the joint angles were. Of course, this is dependant on what Rhino
>needs from the arm, so I might be way off here.

The issue of a "reference" arm configuration and calibration procedures
is mentioned in the MicroScribe patent, quite nicely.

Gar