View Full Version : My home-made digitizing probe


aolshove
10-31-2007, 10:14 AM
I build this dimensionally identical (internally) to Shaun Wainford's probe plans (http://www.machsupport.com/Downloads/SW_Digitising_Probe.pdf).

I cut the top/housing/ring out of a 2" Delrin rod, and the hub out of HDPE using my home-built "7th Sojourn" (http://www.kleinbauer.com/7thsojourn.htm) dremel-based router.

The pictures show a dart tip that I was using for testing but the digitized dime at the bottom of this post was acquired with a sharpened 2" long 6-32 screw tapped into the hub of the probe. As you can see from the images, the housing of the probe was modified to fit into the dremel mount. It's not as sexy as the probes you guys make out of aluminum but it works for me.

Dime was digitized at .005" between points using Mach3 Wizard and then rendered in Rhino4.

vacpress
11-02-2007, 12:11 PM
Neat!

I really want to make one of these eventually. So, inside that thing you made the circuit board and 3-pins per shaun's design?

R

aolshove
11-02-2007, 12:27 PM
Neat!

I really want to make one of these eventually. So, inside that thing you made the circuit board and 3-pins per shaun's design?

R

Yep! Internally it's identical to Shaun's design. 2mm circuit board. The only difference is that I used imperial 2-56 screws instead of metric, 3/16" balls instead of 5mm, and 1/8" brass rod instead of 3mm "silver steel".

Khalid
11-02-2007, 12:41 PM
can u pls give us an idea about the circuit?

aolshove
11-02-2007, 12:57 PM
can u pls give us an idea about the circuit?

I'm not exactly sure what you're asking. The board itself is laid out quite plainly in Shaun's PDF linked in the first post. The circuit itself is simply 3 NC switches connected in series. Open any one and you open the circuit. Each ball is in contact with the copper of the circuit board causing them to become part of the circuit. The brass rods embedded in the central hub contact two balls together thus making a closed switch. If any one rod is removed from its resting place between two balls, the circuit opens.

Khalid
11-02-2007, 10:38 PM
Ohhh...thanks for explaining.. i thought there is PCB inside ...
once again thanks for clarification..i m gona cut the material soon...

vacpress
11-03-2007, 12:54 AM
so... you etched thecircuit board in f.c. or similar? you built the housing on a lathe and used a grinder to make pins? i am probably not alone in saying that shaun's plans are complete, but difficult to interpret.

all the same - i want one! the 3d scanner at camtronics is also tempting.. i wish there was a user forum of people who could let me know if it is worth the $~200 for a laser and the software...

aolshove
11-03-2007, 10:29 AM
so... you etched thecircuit board in f.c. or similar? you built the housing on a lathe and used a grinder to make pins? i am probably not alone in saying that shaun's plans are complete, but difficult to interpret.


Just to clarify, all the parts are milled using my CNC "7th Sojourn" router as indicated in the first post. The housing, etc are all milled on my 7th. I don't own a lathe. The rods are just 1/8" brass rod you can buy at most hardware stores cut to length with a hacksaw. The circuit board (PCB) really has no solder-in parts other than soldering two wires to it. The PCB itself was milled rather than etched following standard 2D toolpath methods using a 1/16 2-flute carbide flat mill.

I found the plan's to be rather non-instructive as well but decided to forge on and just concentrate on making the various parts and see how they fit/worked together after cutting them. Half the fun was seeing how many of the parts I could cut using just the Newfangled Wizard in Mach3 and I cut all of them except the PCB which I had to layout in Autocad and toolpath the usual way.

ger21
11-03-2007, 11:28 AM
all the same - i want one! the 3d scanner at camtronics is also tempting.. i wish there was a user forum of people who could let me know if it is worth the $~200 for a laser and the software...

Art of Mach3 fame is currently working on a video/laser scanner for Mach3. Here's a quote from the Mach3 support group:

Its all R&D at the moment, but showing some very promising results. Basically, I took a logitech fusion webcam and mounted it to a piece of wood, then I added a laser line device at an arbitrary angle (software reports its 21.6 degrees ), and I an writing an application, (stand alone) that will talk to mach3 in the background as a remote control application. It asks you to put in a block of wood as a calibrator, then enter its size. The software then commands mach3 to raise , lower and move to side to side to calculate the FOV of the camera, correction angles of the various FOV's , scan angle etc.. and then will begin to scan the table, table size is the only limit. Its planned to have it look for complex areas, and zoom the camera down with the Z axis to get closer and zoom away on larger
areas. It takes up to 640 data points per movement, stepping about 1mm each time or whatever you set, it takes a snapshot of those 640 points very quickly, so it should do about 500,000 to 2.5 million points per hour. When scanning it will purposely oversample each point, and I intend to add a gausian filtering to the point objects to get as much accuracy as possible, I know there are a few recent theories on how to get a close approximation of a point through multiply imaged points..

How accurate? Dunno yet. Im hoping for .2mm to start ,as I say its R&D , and Ive been puzzling over the issue for some time, there are many solutions out there using line laser to scan, but I havent seen any under 3d control for getting more accuracy by FOV variation and such to get a more complete picture of the object.. so far, Im seeing some variability due to various things like laser speckle, reflective aberation..etc.. but there are known ways to reduce those quite a bit. It likely wont be a point cloud anywhere near as accurate as a probe cloud, but my aim is for woodworking and getting objects like a clenched fist and such into wood. When under 3d control, a laser line device can do undercut sensing and such so its a cool thing to play with. I add to it as I can, and have it currently self calibrating and reporting stats that are better than I seem to be able to measure by hand, ( like the angle of the beam..), so, keep your fingers crossed, might actually work out. :)

Thanks,
Art

Switcher
11-03-2007, 11:40 AM
Why wait for a 3D scanner?

Free 3D scanner:
http://www.david-laserscanner.com/

Forum:
http://www.david-laserscanner.com/forum/

Gallery (scan your own head) :) :
http://www.david-laserscanner.com/?section=Gallery&page=2


.