pushing the "homemade faceting gear" further

Prospecting Australia

Help Support Prospecting Australia:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Sodabowski

Thomas Rodriguez
Joined
Mar 24, 2014
Messages
189
Reaction score
115
Location
Southern France
G'day all,

So we're a few here to have an engineering background and working on faceting hardware and software. Nena started one here https://www.prospectingaustralia.com/forum/viewtopic.php?pid=366729

and since I don't want to pollute his posts too much (and we're taking different approaches to this business) here's a new topic about what a physics dude can come up with.

Let's start with a 3D model of the machine I'm working on (old post somehwere else in this website):
https://www.prospectingaustralia.com/forum/viewtopic.php?pid=123485#p123485

The plan is building a digitally-controlled faceting machine with a cost not too far away from the existing machines. The goal is under $ 2000 for the hardware, with no laps (it will take the standard 8" master laps and laps anyway).

As I progressed along this lengthy project, I eventually realized that I needed a way to enter the faceting designs into the computer and be able to translate them into cutting instructions for the machine. Soooo, I added a new branch to the project, which is a 3D faceting software la GemCAD. So here I am working on a polygonal modeler that simulates the controls of a faceting machine and outputs a set of cutting instructions and 3D views of the design.

Then, before you know, the perfectionism fairy chimes in and whacks me with her wand... and suddenly I find myself devising a 3D scanning device/method for gemstones. Rotating platform, stone stuck to a needle with whatever glue will do the trick, backlight and camera to shoot the profile of the stone at varying angles.

This is where I am so far : the 3D scanner soft is almost finished but far enough for a demo.
Here's the result of a hard test done with a tiny 0.7 carat pink saph (negative cryztallized faces!), that I stuck to a toothpick with a lil' dab o'wax, backlit heavily (glowing edges!) and shot every 10 degrees with my iPhone, with a very bad focus (hey, hard test!). I erased the toothpick afterwards in Photoshop, aligned and cropped the images, but once the dedicated scanning platform will be done there will be no intermediate processing.
Here's one of the 18 pictures:
1508787702_img_3941.jpg.jpg


A screenshot of the homebrew 3D scanner software, with one of the side views imported and the contours vectorized (two profiles actually, left and right):
1508788621_image_2.jpg


And the resulting 3D mesh of said stone, imported in Cinema4D (since the 3D faceting modeler is not finished yet to display anything):
1508788672_image_1.jpg


The number of points per contour is adjustable and I set it on a rather low points count here, which already gives a pretty decent outline of the stone.

Why do I care? well, having the 3D outline mesh of a stone to be cut, I can import it into the faceting diagram modeler and choose/optimize the design for the particular piece of rough that it will be cut into. Yield does matter, right?

Before you ask, yes, I have plans for the future for this software and the hardware that it will be meant to run. This is the early prototyping phase.

Cheers, and thanks for your interest.
 
Hi Thomas,

Looks nice! Also need to program such. I have 2 3d scanners from Optimet laying around, they are normaly used for scanning teeth for a dentist. Didnt try those but these are old but useble. Only i need to figure out the drivers how they work so need to get a winxp pc and try the test software i have got from Optimet mayby i can hack the bus driver and figure out how i can measure under the micron. Those have a parallel port connection so it shouldnt be that hard to hack the communication :D But still!

I like what you did but why not your own 3d software?! I've written something like gemcad but on a website :D also can import cad files if i want to or export G-code. I also did a huge amount of 3d work on a standard diamond cut and did find the calculations for setting up a diamond. That was a good lerning project but not an easy one. If i get the scanner working i can use it to measure out the stone and the lap but the thirst thing i want is to get the basics working. Then ill go on with the scanners and the auto lap/stone changer.

By the way i program in #C and Perl with MySQL as Database the output graphical with SVG (2d) and OpenGL for the 3d. Also C code for the chip programming.

Greets,

Christian
 
Well i AM doing my own 3D software! but since it's not finished I did the 3D mesh shot with Cinema4D (imported the points generated by my 3D scan soft). C is awesome but really too far fetched for what I aim in terms of coding ease and time, I've done quite my boatload of C though. I'm also thinking MySQL to store the models, since Revolution/LiveCode is so easy to interface with any database server. BTW there's also a web version of LiveCode but I don't plan on using that feature for the time being.
#C is mandatory for proper MacOS coding but I didn't take the time to learn that one, I don't really have the use for that kind of power ;) If I need a UNIX tool I can use C which is far enough. The big thing with Revolution/LiveCode is the "develop once, deploy everywhere" workflow, plus it's a visual IDE. I really don't think we would need to handle tens of thousands of polygons for a gem faceting diagram (unless you would tackle a copy of the Cullinan with way more facets) so what the heck ;)

Cheers!
 
Hehehe, ya im doing everything on linux debian. So that my ipad can handle the machine :D lol. uh my wifes ipad lol :p i put everything into the mysql also photographs that are taking during cutting. My brother in law is still figuring out how i can speed up my program... did some nice coding but it can done much faster then i have written. But for now im using the things i have.

Did you check the scanners fom Optimet?! they are awesome to use!!! Hope its going to work so that i also can measure the end product :D but thats all future!

Yeah we are both nuts :D lol but it is also a need project!
 
I think that this could be a really excellent idea for certain stones. Some, like sapphires, really need to be oriented by eye to maximize the color, since these stones are often heavily zoned. But once you got them oriented, this program would give you the best yield for that orientation.

I say go for it! It's way beyond my capabilities and understanding, but it could have tremendous potential, especially for commercial cutters.

Debbie K
 
Thanks for the nice words Debbie, I'm working towards that goal indeed! At the moment I'm examining different technical means of aligning the rough after it has been scanned and a best-yield design prepared and orientated.

On the exact matter of sapphires I devised a dopping accessory that I submitted to UltraTec several years ago, but after some prototyping they came to the conclusion it would be unprofitable for them to bring to the market, so I'm slowly gearing up to produce and market it myself.

This is how far I've gone to date on the 3D modeller :

1509144177_image_3.jpg


With a very basic gem design only to test the interface. Next will be the controls to actually "cut" into the rough and build the design as in GemCAD. It already has the importing features for the reconstructed 3D mesh from the scanning software.
 
Thanks mate, not really sure how high (or low) your pay rate is but mind you I won't be asking thousands for that software ;)
 
Hello Thomas,

Did also a test with the measurement! :D

Img_4060.jpg

Img_4059.jpg


Still need to write some software for the scanner... and use a fluid for scanning the stones. But the test is there :D hope i get it all working!

Greets,

Christian
 
Hey Christian! Excellent start! Could you send me a sample file from that scanner? I'll also need to be able to interface to that kind of stuff, but what I most need is the file format to be able to import data, I found nothing on the web so I reckon I'll have to resort to the hacking method ;)

Cheers!

Thomas
 
Hi Thomas,

Dont have a sample yet :D now i need to hack the parallel port to send and read the data from the scanner itself. I want to use the scanner on a arduino/teensy and write all data first to an SD card. Then i need to order the data and write it into a totall measurement. Ive looked on ebay what these kind of scanners cost but they are expensive +/- 2000 euro's second hand. But the signal is verry stable! About 1 micron!!!

Greets,

Christian
 
'Sup gents,

So it's almost been three years that I've been working my arms out on this project... which obviously had to be pushed over the edge once I attacked the raytracer and realized it would be JASPER (Just Another Stupid Pixel raytracER :playful: ). I've expanded GemSuite Scanner and Modeler to make the full project handle actual mineralogical data, which means * drum roll * it will handle anisotropy and all the good stuff that goes with it. Which meant I also had to design a line of dedicated gem scanners. Well for those of you that are in my circle on facebook you already knew that. I'm still working on the beamtracer, which includes a LOT of real-life physics waaay beyon the usual raytracers. To my knowledge, nobody ever did that before in the faceting community.

All the project progress can be found on my dedicated blog : http://blog.sodabowski.com
I can't seem to be able to upload any image at the moment. Oh, well.

Cheers!
 
Top