Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Posted by
Jon Elson
on 2001-08-12 21:30:41 UTC
rab@... wrote:
acceleration/deceleration ramps. In practice, all controls have a limit on
how many blocks per second they can process. As long as
you are not requiring the CNC control to process more than that
many blocks/second, it should do it without any side effects.
Some controls have a limit on lookahead, ie. they only look ahead
50 blocks, or some fixed number. This is dangerous, because a
CNC control can get into the fix of moving at high speed through
a series of small moves and suddenly a stop or reversal of the
path comes up at the 'horizon' of its lookahead, and it is already
too late to begin decelerating. This can cause a crash similar to
the infamous terrain-following radar that sometimes couldn't spot
the cliff from far enough away to not hit it.
EMC has 'infinite' lookahead, so that it should never get into this
position, or at least, it never starts a move before it knows what the
end of the sequence of moves will look like.
Jon
> Art,In theory, the segment length should have no effect on
>
> How would you go about determining the optimal segment lenght
> settings when exporting a toolpath from a CAD/CAM program so that the
> machine ramps smoothly ?
acceleration/deceleration ramps. In practice, all controls have a limit on
how many blocks per second they can process. As long as
you are not requiring the CNC control to process more than that
many blocks/second, it should do it without any side effects.
Some controls have a limit on lookahead, ie. they only look ahead
50 blocks, or some fixed number. This is dangerous, because a
CNC control can get into the fix of moving at high speed through
a series of small moves and suddenly a stop or reversal of the
path comes up at the 'horizon' of its lookahead, and it is already
too late to begin decelerating. This can cause a crash similar to
the infamous terrain-following radar that sometimes couldn't spot
the cliff from far enough away to not hit it.
EMC has 'infinite' lookahead, so that it should never get into this
position, or at least, it never starts a move before it knows what the
end of the sequence of moves will look like.
Jon
Discussion Thread
Art Fenerty
2001-08-10 17:02:45 UTC
UCMI Interface
stevesng@n...
2001-08-11 07:29:49 UTC
Re: UCMI Interface
Art Fenerty
2001-08-11 12:46:31 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Matt Shaver
2001-08-11 15:09:42 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Art Fenerty
2001-08-11 16:15:38 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Jon Elson
2001-08-11 23:20:59 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Art Fenerty
2001-08-12 05:44:35 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Jon Elson
2001-08-12 13:25:27 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Art Fenerty
2001-08-12 13:43:37 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Art Fenerty
2001-08-12 16:04:58 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
rab@r...
2001-08-12 16:18:56 UTC
Re: UCMI Interface
Art Fenerty
2001-08-12 16:25:39 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Jon Elson
2001-08-12 21:30:41 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Jon Elson
2001-08-12 22:22:10 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
William Scalione
2001-08-12 22:30:27 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Art Fenerty
2001-08-13 03:10:44 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Jon Elson
2001-08-13 09:50:41 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Art Fenerty
2001-08-13 10:01:30 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Jon Elson
2001-08-13 10:27:10 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
William Scalione
2001-08-13 10:38:07 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Jon Elson
2001-08-13 22:45:12 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface
Jon Elson
2001-08-13 22:55:54 UTC
Re: [CAD_CAM_EDM_DRO] Re: UCMI Interface