Re: [CAD_CAM_EDM_DRO] G02 - G03
    Posted by
    
      Jon Elson
    
  
  
    on 2000-08-21 12:03:50 UTC
  
  Art Fenerty wrote:
CAM or postprocessor uses one over the other. The radius of the circle
is clearly stated with the R word, while the endpoint is clearly stated
with
the IJK words. There are some errors that accumulate as the circle
algorithm works its way around the path, and so the IJK may get you a
little closer to the exact point desired, while the R will get more
precisely
to the desired radius. I used to use IJK until there was a problem with
the
early EMC program. With IJK, you can calculate radius from the start
and end points, and if they differ by more than some tolerance, flag an
error. The original tolerance was set so tight that roundoff errors in
xx.xxxx coordinates were sufficient to trip that error. It needed the
start and end radius to match to more than 6 decimal places.
Really, there is just a tiny bit of additional syntax decoding to take
both. Once you've got the numbers read in, the arc processing
is the same.
Jon
> Hi:You should try to accept both. There are some reasons why a particular
>
> I am trying to update MasterControlCNC to do G02
> and G03 circular interpolation and although I now have working code (
> thanks
> to Doug Fortune) I don't quite understand which is the most common
> form of
> interp.
> My post processor on the software I use the most does
> not generate circ. only linear. Am I better off using R registers or
> IJK
> interp? or both? Is IJK now pretty much obsolete?
CAM or postprocessor uses one over the other. The radius of the circle
is clearly stated with the R word, while the endpoint is clearly stated
with
the IJK words. There are some errors that accumulate as the circle
algorithm works its way around the path, and so the IJK may get you a
little closer to the exact point desired, while the R will get more
precisely
to the desired radius. I used to use IJK until there was a problem with
the
early EMC program. With IJK, you can calculate radius from the start
and end points, and if they differ by more than some tolerance, flag an
error. The original tolerance was set so tight that roundoff errors in
xx.xxxx coordinates were sufficient to trip that error. It needed the
start and end radius to match to more than 6 decimal places.
Really, there is just a tiny bit of additional syntax decoding to take
both. Once you've got the numbers read in, the arc processing
is the same.
Jon
Discussion Thread
  
    Tim Goldstein
  
2000-08-21 11:07:18 UTC
  RE: [CAD_CAM_EDM_DRO] G02 - G03
  
    Art Fenerty
  
2000-08-21 11:13:11 UTC
  Re: [CAD_CAM_EDM_DRO] G02 - G03
  
    Art Fenerty
  
2000-08-21 11:15:40 UTC
  Re: [CAD_CAM_EDM_DRO] G02 - G03
  
    Jon Anderson
  
2000-08-21 11:35:05 UTC
  Re: [CAD_CAM_EDM_DRO] G02 - G03
  
    Jon Elson
  
2000-08-21 12:03:50 UTC
  Re: [CAD_CAM_EDM_DRO] G02 - G03
  
    Darrell
  
2000-08-21 16:16:19 UTC
  Re: [CAD_CAM_EDM_DRO] G02 - G03
  
    Darrell
  
2000-08-21 16:17:58 UTC
  Re: [CAD_CAM_EDM_DRO] G02 - G03
  
    Art Fenerty
  
2000-08-21 17:17:07 UTC
  Re: [CAD_CAM_EDM_DRO] G02 - G03
  
    diazden
  
2000-08-22 17:41:45 UTC
  Re: [CAD_CAM_EDM_DRO] G02 - G03