Re: Scanner/Digitizer skip function/probe details?
Posted by
Andrew Werby
on 2001-01-05 10:58:56 UTC
Ballendo wrote:
I would love to see the list share thoughts on a digitising probe.
Both by explaining what's out there and how it works (MaxNC, Larken,
Renishaw, Roland picxa, etc.) and then by working to create a probe
we all could build!?!
Hope this helps.
Ballendo
[I've used the MaxNC and the Roland systems, and I'd say the latter would
be the one to emulate, if one wished to build a system which runs in a mill
bigger than Roland provides. The piezo-electric probe is much more
sensitive, and the digitizing routine is more sophisticated. (While the
MaxNC carves tracks in hard plaster parts, the Picza can digitize soft clay
objects without leaving marks.) I asked if I could get some replacement
probes for doing this sort of thing, but the Roland folks were extremely
uninterested in providing them. So the best I can suggest is getting a
Picza and detaching the probe, while studying the software so that it could
be adapted to a larger envelope.
The MaxNC works by dragging the probe along the surface until it is
deflected quite a bit, then it raises up a little and tries again. If it
doesn't feel contact, it will drop until it does. It reports a series of
positions as short line segments, stating each point value twice; once as
the end of a segment and once as the beginning of a new one. It tracks over
the surface in a meandering pattern, like a raster toolpath, and if you
open the file in a graphics program, it will look like one long polyline.
Although this can be run as-is as a milling toolpath, it takes a lot of
work with a graphics program to create a surface mesh from it.
The Picza software allows you to first set the scanning area graphically-
you size a rectangle which corresponds to the area to be scanned. Then you
test it by invoking "begin area test", which sets the needle down in eight
points bounding the rectangle, so you can see how it relates to the part.
When this is correct, you set the top and bottom of the scanning area; the
top is set by sending the probe down to the highest spot on the part (I
generally put a piece of cardboard on it to give a margin for error); the
bottom is set numerically. When you've told it how fine you want the scan
by setting the x and Y scan pitches numerically (the finest is .001, which
captures a lot of detail, but creates a huge file) then you press the scan
button and the scan function starts. The first thing it does is reconnoiter
the part by making a few passes with a fairly wide spacing, so it has an
idea of where the high points are. Then it settles down to tracing parallel
paths in the x-direction, dragging the point along much like the MaxNC, but
with much less pressure, proceeding with a slight clicking noise. At the
end of each pass, it ranges about, feeling for the topmost point it will
have to encounter before settling down to digitize the next line. When it
is finished (which can take quite a few hours) it cogitates a bit, then
constructs a mesh from the data it has collected, which it shades so you
can see the surface. This can be exported as a DXF or STL file which most
3d graphics programs will recognize as a surface and open.
Andrew Werby
http://www.computersculpture.com
.
Andrew Werby - United Artworks
Sculpture, Jewelry, and Other Art Stuff
http://unitedartworks.com
I would love to see the list share thoughts on a digitising probe.
Both by explaining what's out there and how it works (MaxNC, Larken,
Renishaw, Roland picxa, etc.) and then by working to create a probe
we all could build!?!
Hope this helps.
Ballendo
[I've used the MaxNC and the Roland systems, and I'd say the latter would
be the one to emulate, if one wished to build a system which runs in a mill
bigger than Roland provides. The piezo-electric probe is much more
sensitive, and the digitizing routine is more sophisticated. (While the
MaxNC carves tracks in hard plaster parts, the Picza can digitize soft clay
objects without leaving marks.) I asked if I could get some replacement
probes for doing this sort of thing, but the Roland folks were extremely
uninterested in providing them. So the best I can suggest is getting a
Picza and detaching the probe, while studying the software so that it could
be adapted to a larger envelope.
The MaxNC works by dragging the probe along the surface until it is
deflected quite a bit, then it raises up a little and tries again. If it
doesn't feel contact, it will drop until it does. It reports a series of
positions as short line segments, stating each point value twice; once as
the end of a segment and once as the beginning of a new one. It tracks over
the surface in a meandering pattern, like a raster toolpath, and if you
open the file in a graphics program, it will look like one long polyline.
Although this can be run as-is as a milling toolpath, it takes a lot of
work with a graphics program to create a surface mesh from it.
The Picza software allows you to first set the scanning area graphically-
you size a rectangle which corresponds to the area to be scanned. Then you
test it by invoking "begin area test", which sets the needle down in eight
points bounding the rectangle, so you can see how it relates to the part.
When this is correct, you set the top and bottom of the scanning area; the
top is set by sending the probe down to the highest spot on the part (I
generally put a piece of cardboard on it to give a margin for error); the
bottom is set numerically. When you've told it how fine you want the scan
by setting the x and Y scan pitches numerically (the finest is .001, which
captures a lot of detail, but creates a huge file) then you press the scan
button and the scan function starts. The first thing it does is reconnoiter
the part by making a few passes with a fairly wide spacing, so it has an
idea of where the high points are. Then it settles down to tracing parallel
paths in the x-direction, dragging the point along much like the MaxNC, but
with much less pressure, proceeding with a slight clicking noise. At the
end of each pass, it ranges about, feeling for the topmost point it will
have to encounter before settling down to digitize the next line. When it
is finished (which can take quite a few hours) it cogitates a bit, then
constructs a mesh from the data it has collected, which it shades so you
can see the surface. This can be exported as a DXF or STL file which most
3d graphics programs will recognize as a surface and open.
Andrew Werby
http://www.computersculpture.com
.
Andrew Werby - United Artworks
Sculpture, Jewelry, and Other Art Stuff
http://unitedartworks.com
Discussion Thread
ballendo@y...
2001-01-03 19:09:24 UTC
Re: Scanner/Digitizer skip function/probe details?
Alan Marconett KM6VV
2001-01-03 20:14:38 UTC
Re: Scanner/Digitizer skip function/probe details?
Matt Shaver
2001-01-03 22:55:56 UTC
Re: [CAD_CAM_EDM_DRO] Re: Scanner/Digitizer skip function/probe details?
dave engvall
2001-01-04 09:04:46 UTC
Re: [CAD_CAM_EDM_DRO] Re: Scanner/Digitizer skip function/probe details?
Alan Marconett KM6VV
2001-01-04 12:56:07 UTC
Re: Scanner/Digitizer skip function/probe details?
Lee Studley
2001-01-04 16:30:19 UTC
Re: Scanner/Digitizer skip function/probe details?
dave engvall
2001-01-04 17:12:31 UTC
Re: [CAD_CAM_EDM_DRO] Re: Scanner/Digitizer skip function/probe details?
Alan Marconett KM6VV
2001-01-04 21:59:42 UTC
Re: Scanner/Digitizer skip function/probe details?
diazden
2001-01-05 00:09:08 UTC
Re: [CAD_CAM_EDM_DRO] Re: Scanner/Digitizer skip function/probe details?
Smoke
2001-01-05 08:38:22 UTC
Re: [CAD_CAM_EDM_DRO] Re: Scanner/Digitizer skip function/probe details?
ballendo@y...
2001-01-05 09:04:08 UTC
Re: Scanner/Digitizer skip function/probe details?
ballendo@y...
2001-01-05 10:13:05 UTC
RE:Re: re: Scanner/Digitizer skip function/probe details?
Andrew Werby
2001-01-05 10:58:56 UTC
Re: Scanner/Digitizer skip function/probe details?
Smoke
2001-01-05 11:57:03 UTC
Re: [CAD_CAM_EDM_DRO] Re: Scanner/Digitizer skip function/probe details?
Alan Marconett KM6VV
2001-01-05 13:04:11 UTC
Re: Scanner/Digitizer skip function/probe details?
ballendo@y...
2001-01-05 14:09:36 UTC
Re: Scanner/Digitizer skip function/probe details?
Alan Marconett KM6VV
2001-01-05 14:33:51 UTC
Re: [CAD_CAM_EDM_DRO] Re: Scanner/Digitizer skip function/probe details?
Smoke
2001-01-05 15:25:56 UTC
Re: [CAD_CAM_EDM_DRO] Re: Scanner/Digitizer skip function/probe details?
ballendo@y...
2001-01-05 16:36:27 UTC
re:Re: Scanner/Digitizer skip function/probe details?
ballendo@y...
2001-01-05 16:50:12 UTC
Re: Re: Scanner/Digitizer skip function/probe details?