Re: Newbie has some questions

[Ray <rehenry-at-up.net>]

On Saturday 27 October 2001 06:59, Dean wrote:
<s>
> Probing:
>
> It sounds like the probing feature is kinda new.  What I would like to
> have is some kind of dialog that asks me for a range to probe through
> (X,Y,Z),  a resolution to probe at, and the dimensions of an unworked
> piece that will ultimately be machined.  I let it run, come back and I
> am left with a usable fragment of G-code.
>
> I suspect this probing feature does not yet do all of this.  This is OK,
> I am a pretty good Linux C & C++ programmer.   I feel confident that I
> can write a piece of code that can do this.  Where do I start?  What
> does the probing feature do at this point in time?
>
> I intend to use an led/photo transitor pair for the probe switch.  I
> will stradle these across a small teloscopic pair of brass tubes with
> proper holes to interupt the light.  I will attach a piece of silicon
> rubber to the end of this that is the shape of the bit.  Sound good?

Dean

Fantastic.  We need a few good programmers!

The probe features were migrated into the emc code from a parallel NIST 
project with CMM.  Will Shackleford wrote the probe access script to test 
and demonstrate it there.  It simply takes a linear move and sees if the 
probe switch trips.  If it does, the motion is interrupted and the 
location along the path reported.  With a bit of modest hacking that 
script could be made to probe a grid pattern like you describe.   

I would think that you could begin in manual and teach the corners of the 
pattern, enter a repeat value and let it go.  It might also be possible to 
speed up probing by teaching some sort of envelope within which the 
pattern lies.

The details of the probe sound interesting.  The difficulty comes in the 
interpretation of the probe points on contoured surfaces.  This can be 
aided some by computing lines between successive points and offsetting the 
touch point by the radius of the bit.  Making a g-code program from a 
cloud of these kind of points should not be that difficult.  There might 
be programs or code available for parts of this.

I'm ready to help!  I'm sure that Till would enjoy putting a probe on the 
platform of one of his hexapods.  Watching PriModell 2 do five axis 
probing would be interesting indeed. 

( see http://www.isw.uni-stuttgart.de/personen/t_franit/primodell/ )

A demonstration like that would be a real incentive to visit Stuttgart 
again.

Ray