Re: RE: Re: Stewart Platforms...

>    From: "Askew, Jason" <jaskew@...>
> Subject: RE: Re: Stewart Platforms...
>
> Thanks, Ray!

<s>

> It seems the EMC is quite the piece of software.  I am downloading the
> CD ISO right now.  I hacked together a 200mhz pc to load it on.
>  Hopefully I'll get it up and running tonight.  I'd seen Paul's setup
> online before.  Very impressive...  does Paul frequent this list?

The effort that he has put into the BDI is impressive. He did read this
list for a while. I don't believe that he's reading much of it lately.
More tuned to emc@...

> I've finally got my mind wrapped around the math involved with forward
> kinematics...  esp. why there can be so many solutions.  

Whoa! That's way beyond this old dog. Yes there are many possible
solution to the forward math. The EMC tends to pick the one closest to
the current position. There are a number of good comments in that part
of the code in emc/src/emcmot. Look for the *kins files.

> So, if I
> understand you correctly, EMC can handle this?  If I were to finish a
> 3x3 system tomorrow, EMC could drive it?

Yes, I believe that is true but you will need to install the EMC on a
rather fast PC, and set up the appropriate geometry, so that it can keep
position calculations ahead of the motion loop. Robot arm calculations
require a lot more work that those required for Stewart type platforms.
Think it'll be ready for the NAMES show?

I don't know of anyone currently running stacked rotaries. Most of the
proof of this ability was done with a Puma 560 (6x6?) that, the last time
I saw it, was sitting in a back corner storage area of the metal shop at
NIST. It and the autonomous Humvee were looking like they needed a
little loving care.

Hope this helps.

Ray