Michael,

On Mon, Apr 11, 2011 at 12:06 AM, Michael Haberler <[email protected]> wrote:
> Norbert Schechner and myself are trying to support a machine with what 
> basically is a rack toolchanger (Heckler&Koch BA20, see pictures 2&3 here: 
> http://innovac.de/kartei.php?code=3741) using procedure-based toolchange. So 
> this has a spindle pocket with the current tool, BUT change is NOT swapping 
> spindle and current pocket.
>
> - so it's non-random
> - we want to use the pocket field which is different from the tool number
> - we would like to have more tools in the table than pockets and support 
> operations like 'remove tool 1 from changer', 'empty pocket 2' and 'insert 
> tool 3 into changer at pocket 5'
>
>
> what puzzles me is the tool table organisation in 
> src/emc/rs274ngc/tool_parse.cc:
>
> - the way the table is read in for nonrandom machines essentially makes the 
> pocket information go away - tools are stored sequentially in the table and 
> pocket == toolnumber?
> - what is the rationale behind the fms[] array in tool_parse.cc?
>
> what I could do is use RANDOM_TOOLCHANGER = 1 and introduce yet another ini 
> option which adapts the change operation in iocontrol but that's kludgy
>
> I'd appreciate advise how to make pocket information usable for nonrandom 
> changers
>
> -Michael
>

A bit off topic, but ive done a few rack and shuttle tool changers, none emc

CollisionProtectionMode
 one important thing i used was choosing 'collision protection' on or off
 on these controls (Heidenhain), this was a mode used during
not-cutting ( think spindle off or possibly rapid )

 while 'collision protection' was on, a 12V few mA current was placed
between spindle and work.
 this requires some insulation, maybe the rack maybe the worktable or spindle.

 during the tool change , all motion except when very close to rack,
was done in 'collision protect on' mode
 the sequence of motions were carefully chosen to move safely and
assume as little as possible about the tools

 what is described in tables is not necessarily  what is in rack.
 the rack may have no tool or a longer tool or a wider tool or a
misplaced cocked tool
 or you may be trying to put a tool into a rack slot where a tool already is

 by use of 'collision protection' mode, i could abort many toolchanges safely

 it would be like putting emc into a probe mode during all non cutting motion
 while the tool receiver was griping the tool and releasing the tool,
the mode would be off.

TOOL TABLE DATA
wrt your idea of using the tool tables for more info
i never changed pockets, i married tools to pockets forever ( T<n>
occupies Pod<n> )

the Heidenhain system also had tool tables with data
these data were intend for specific uses and were accessible thru
system variables ( they use Qnnnn vs #5xxx in emc2)
this meant i could ignore what they were intended for and use them as i liked

the arbitrary use of these data is handy

i implemented crude tool life ( number of uses ), flushing modes ( jet
or flood) and other items
because it was just a list of data available to me

also the titles of the columns for the tool data was just a text file,
so i could make it look nice for the user

the system had a finite number of float variables so i had to limit
the number of tool positions

btw: the idea of using a fixed X and a varying Y did NOT work well in the field
i have a few dozen of these in use now and i used an X and a Y
position for each tool stud
i used a common Z position.
i found the variance non-linear and of magnitudes near 0.5mm,
i set the machine positions into memory while moving up & down onto
the studs by hand
i indexed thry coord pairs using the tnum as an offset from the base 'address'
( >>like<<   rackx=#5398+(tnum*2)  racky=#5398+(tnum*2)+1 )

and from your pix,
use a Zhi posn (to grab you gotta start above and end up lower, moving
away sideways)
and a Zlo posn (to putaway you need to start low, move sideways, slip
onto fingers, release and go up & clear of stud)

i got a model coming along now, will post

hth

tomp

------------------------------------------------------------------------------
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers

Reply via email to