Re: digitizer probe and software
Posted by
Fred Smith
on 2001-11-05 18:45:42 UTC
--- In CAD_CAM_EDM_DRO@y..., Alan Marconett KM6VV <KM6VV@a...> wrote:
proprietary, but very simple here is an example point:
point(1.0,2.0,.1);
Place as many of these in an ascii file, or copy-Paste Special them
through the clipboard and the points appear in Vector as if you had
laboriously entered each and every coordinate. They also are
selected in order if pasted, ready to be splined or just line
connected.
If you scan a lot of points in a row, Vector can draw a spline
through them. I did quite a bit of work with these, as I was working
on a routine to get Picza scanned DXf files imported into Vector and
Bobcad. the Picza format is 3D faces. These are the same basic form
as the triangles in an STL file, and in fact StlTrans converts
between the two file formats. Bear in mind that the picza uses a
needle and a piezo electric pickup to scan the surface form. It does
not use a 1/4 diameter ball for contact measurement. I wrote a VB
program to turn the Picza data into a series of rasterized lines in
Vector. Very similar to the lines that you are getting from the
MaxNC scans. By playing with the selections, I de-select the end
connections. This has the effect of selecting a series of parallel
but planar curves, not a series of interconnected curves. Vector can
offset any 3D planar curve (it was hopeless in Bobcad 16 and 17 as
they did not have good enough control over chaining and selecting
operations)
At the time in Vector the processing time was extremely long and the
32,000 entity limit of Vector 7.0 on a P166 was also discouraging.( I
was trying to create arc splines to cut the models with fitted 3D
arcs). With 32 bit and unlimited entity counts today, and P4's at
nearly 2.0 GHZ, it is much more practical. I have also played with
using backplotted StlWork files in Vector and arc-spline reducing the
toolpaths. It is not much use on a PC based controller that has a 40
GIG harddrive for a memory limit (it does help if the controller does
not do well on 3D continuous contouring), but on hardwired
controllers, memory is very dear and if you can squeeze an arcfit
program into machine memory, but not a linear splined one, it is a
significant productivity improvement.
to Vector. For a simple example, draw an arc from 90 to 180 degrees,
and orient it to the Y-Z plane. ( Rotate about the Y axis so that it
is standing up). Now draw-other curves-Offset and pick a side(go to
the inside) If you chose .125 and to the inside, you have just
removed as much of the 1/4 diameter probe offset as possible.
If you create an stl file in StlWork you can offset to the side that
the probe cannot reach by using negative stock values. -.125 should
in theory result in the surface that you have scanned, if cut with a
V tipped cutter(no radius). Leaving -.06 stock should compensate for
the difference between your 1/4 ball and a 1/8 ball cutter. I have
not cut one of these kinds of models yet myself, but I believe it
should work as I have described.
By the way, My Vector 9.3 is still running on Win-XP, 1.4 GHZ Athlon.
Best Regards,
Fred Smith
IMService
> Hi Fred,XYZ
>
> The XYZ format output is good news! Can Vector CAD/CAM also IMPORT
> files? Certainly will be easier now to convert Gcode (that I didn'tVector can import individual points through the PLC format. It is
> generate) to STL.
>
proprietary, but very simple here is an example point:
point(1.0,2.0,.1);
Place as many of these in an ascii file, or copy-Paste Special them
through the clipboard and the points appear in Vector as if you had
laboriously entered each and every coordinate. They also are
selected in order if pasted, ready to be splined or just line
connected.
If you scan a lot of points in a row, Vector can draw a spline
through them. I did quite a bit of work with these, as I was working
on a routine to get Picza scanned DXf files imported into Vector and
Bobcad. the Picza format is 3D faces. These are the same basic form
as the triangles in an STL file, and in fact StlTrans converts
between the two file formats. Bear in mind that the picza uses a
needle and a piezo electric pickup to scan the surface form. It does
not use a 1/4 diameter ball for contact measurement. I wrote a VB
program to turn the Picza data into a series of rasterized lines in
Vector. Very similar to the lines that you are getting from the
MaxNC scans. By playing with the selections, I de-select the end
connections. This has the effect of selecting a series of parallel
but planar curves, not a series of interconnected curves. Vector can
offset any 3D planar curve (it was hopeless in Bobcad 16 and 17 as
they did not have good enough control over chaining and selecting
operations)
At the time in Vector the processing time was extremely long and the
32,000 entity limit of Vector 7.0 on a P166 was also discouraging.( I
was trying to create arc splines to cut the models with fitted 3D
arcs). With 32 bit and unlimited entity counts today, and P4's at
nearly 2.0 GHZ, it is much more practical. I have also played with
using backplotted StlWork files in Vector and arc-spline reducing the
toolpaths. It is not much use on a PC based controller that has a 40
GIG harddrive for a memory limit (it does help if the controller does
not do well on 3D continuous contouring), but on hardwired
controllers, memory is very dear and if you can squeeze an arcfit
program into machine memory, but not a linear splined one, it is a
significant productivity improvement.
> "At present time Vector will process the scanned tool path bycutting
> > offsetting for a different tool size and following the same
> > pattern as that which was scanned."understand
>
> Then is there a tool offset parm for imports in 9.3? Did I
> that right? Or is there another way to remove the "tool diameter"from
> scanned surfaces?Not remove necessarily, as much as adjust. This is not a new function
to Vector. For a simple example, draw an arc from 90 to 180 degrees,
and orient it to the Y-Z plane. ( Rotate about the Y axis so that it
is standing up). Now draw-other curves-Offset and pick a side(go to
the inside) If you chose .125 and to the inside, you have just
removed as much of the 1/4 diameter probe offset as possible.
If you create an stl file in StlWork you can offset to the side that
the probe cannot reach by using negative stock values. -.125 should
in theory result in the surface that you have scanned, if cut with a
V tipped cutter(no radius). Leaving -.06 stock should compensate for
the difference between your 1/4 ball and a 1/8 ball cutter. I have
not cut one of these kinds of models yet myself, but I believe it
should work as I have described.
By the way, My Vector 9.3 is still running on Win-XP, 1.4 GHZ Athlon.
Best Regards,
Fred Smith
IMService
Discussion Thread
Smoke
2001-11-04 16:31:42 UTC
Re: digitizer probe and software
Alan Marconett KM6VV
2001-11-05 11:10:08 UTC
Re: digitizer probe and software
Fred Smith
2001-11-05 14:53:24 UTC
Re: digitizer probe and software
Alan Marconett KM6VV
2001-11-05 15:40:22 UTC
Re: digitizer probe and software
Fred Smith
2001-11-05 18:45:42 UTC
Re: digitizer probe and software
Smoke
2001-11-05 21:08:08 UTC
Re: [CAD_CAM_EDM_DRO] Re: digitizer probe and software
Alan Marconett KM6VV
2001-11-05 22:26:32 UTC
Re: digitizer probe and software
Alan Marconett KM6VV
2001-11-05 22:39:39 UTC
Re: digitizer probe and software
Fred Smith
2001-11-06 04:29:50 UTC
Re: digitizer probe and software
Smoke
2001-11-06 08:11:57 UTC
Re: [CAD_CAM_EDM_DRO] Re: digitizer probe and software
Alan Marconett KM6VV
2001-11-06 11:43:28 UTC
Re: digitizer probe and software
David Goodfellow
2001-11-06 11:48:53 UTC
Re: [CAD_CAM_EDM_DRO] Re: digitizer probe and software
Matt Shaver
2001-11-06 12:06:08 UTC
Re: [CAD_CAM_EDM_DRO] Re: digitizer probe and software
Alan Marconett KM6VV
2001-11-06 12:13:11 UTC
Re: digitizer probe and software
Alan Marconett KM6VV
2001-11-06 12:59:48 UTC
Re: digitizer probe and software