Actually Building the Machine!

[Nena]

Hello Guys!

Finaly i can start with building my machine. I will first build it as a CNC milling machine because i cant pay all parts at once. I just spend over the 2500 euro's for some elektronics and the motors. I now have a good way to build everything and i hope it will work. When the machine is done for milling i could make some verry precies parts for the faceting machine. But the sensors that i need to use a so expensive... (+/- 800 euro's each and i need 6 of them) then it is better to build first the other parts. I hope that i can show you the first photo's over a week or two. But the machine will be also the faceting machine!

I thougth long about it but im continuing to build the machine! Had a hard time here, my wife has health problems, sometimes she pass out with no warning so i had to do all thing in and around the house to this day. so i didnt had any time to do anything for myself.

Also a fine new year for you all!!!

Greetings,

Christian

 

[vonG]

Interested to see what you will come up with. Facinating that you are building such a big machine.

Im on the same journey, but my machine is going to be much smaller. More or less the same size as a regular faceting machine.

Have you seen any of the automated machines that already exist?

//Thomas

 

[Nena]

If you think you can build one you should build one! I only saw ONE machine that is build correct for accuracy and it has a possibility to make all kind of faceted stones in all sizes and types. That machine costed over 1 million AUD! Standard CNC software cant handle a facetting machine! Im also building one that can handle all cutting technics. If i want to cut in a Trillion cut i want to do that with my machine and if i want to make a Cabouchon or a facetted Cabouchon i should make that. All the machines i have seen on the web and on youtube are all made for speed!

Did you already build something?! And the software?! Accuracy? From what size to what size stones can you handle?

Christian

 

[vonG]

Hi Nena.

The first problem I see is that most automated machines are build for multiple stones that are being faceted at the same time. Can never get a great polish that way since every stone can have different flaws in the polish and must be dealt with seperately. Then there is the speed as you mentioned, many variables come into the speed part but most machines are built for big cutting houses and there they cut commercial grade stones anyway. In my case I dont care if it takes several hours to finish a stone, as long as it is polished correcly in the end. That brings us to the biggest problem of them all, the final polish. I think there is only 1 machine that can produce good polish and that is the big German machine with 17 axes. The rest produce commercial cuts only.

It all comes down to the programming I guess, because the hardware and extra gadgets can be built in depending on how much money one is willing to put in it.

For now I only have it down on paper and are chasing parts. Have some problems in getting the arbor for the laps, was thinking of getting a used one for the prototype.
Think I have figured out the solutions for some of the problems that the other machines have, but that is to be seen later.
I have done extensive research in this area and have a couple of videos, that isnt online, of different machines and some calculations that is going to help me in my build.

The stone size is going to be from a couple of millimeters up to a couple of centimeters, using regular laps from standard manuell machines.
The software is still to be figured out, I saw that you probably have that down already. That is going to be the hardest part for me.
The accuracy will depend on the drives and stepper motors that is being used, but would be somewhere in the 0.005 to 0.01mm range. The problem is getting the runout on the arbor + lap to be as accurate too, it will wear in time and possibly give some errors in the polish. But have an idea to solve that too maybe.

Have found some dead threads in other forums from people trying to build a machine, but none of them have succeded as far as I know.

Great to see that you are going to build yours.

Thomas

 

[Nena]

Hello Thomas,

Im a programmer. Some params of my machine is the accuracy is +/- 1/1000th of a degree in turning for the x, y and z. 5/100.000 mm for the 3 other axis. The one you call from germany is also where im going :- ) I use around 40.000 euro's on parts. But i had made some good and cool deals for parts. I think the machine will cost me around 5.000 euro's.

Thats why i told you that i only know one machine, and also that can be done better. Mayby not by software but by technics. Still have some great ideas!

My first setup for the x y and z axis.

Christian

 

[blisters]

Looking like a great project you have there, good on you for having a go sounds like you're on top of it.
Jon

 

[Nena]

Just a image how large the last photo is. It is not a small project. It should hold stones of max 2kg. (+/- 4 pounds) But i think that will never hapen but i hope also to do verry small sizes +/- 1mm till 120mm. But i need to see how far i can get! :lol:

 

[vonG]

I guessed that you were some sort of programmer. That is pretty good accuracy and at a very good price too. You must have great connections.

My setup will cost somewhere between 5-10k, depending on how advanced I want it to be. It is much smaller than yours, but accuracy cost anyway. I know of a working automated machine in my size that cost $20k, but it still only produces commercial stones.

The German machine is made by Fraunhofer and Paul Wild, that machine is pretty advanced using lasers to evaluate the rough stone first. Think they have plans for using water jets to cut the stone before dopping too, if they haven't already implemented that. Don't know if the re-dopping of the stone is made automatically, but it wouldn't surprise me. Their goal seems to be to not lose to much material when cutting, that way earn more on the finished stone. More like the diamond business.

The problem with accuracy and repeatability is that it is dependent on the weakest link, namely the lap. You will never get that accurat runout on a lap as you get in the machine itself, the lap is constantly moving and wearing. Particles that get stuck on the lap will also hurt the polish if not rinsed some way.
The base plate where the arbor is attached have to be accurate too. Aluminum in the 30-50mm thickness range or using granite is a must to get a runout that is sufficient. But this depends on the size of the plate itself. Smaller plate don't need more than 25 mm aluminum, but the precision cost anyway. The best runout is going to be in the 0.005mm range I think.

Is it going to be a gantry model? Probably the best way to go when it's so big.
Had some plans for that first too, or more a Cartesian y z axis robot with a and b axis connected to the z axis. No need for the 5th axis for faceting, unless you have more than 1 lap or have plans for concave faceting. But then there will have to be even more axeses anyway. I came up with a different model at the end, will see if that idea holds or not when I get my parts.

 

[Nena]

Yes it will be a gantry type machine 6 axis and 4 laps on it. The laps will be the big issue but it wont stop me, i use robotic parts for the lap axis these have an accuracy from less then 3/1000th of a mm. But my goal is to get About 1/100th of a mm in accuracy in the complete machine. I think that is a nice point to start with.

That german company has $ to spent i do not and im alone and need to do all by myself and if i can do some good faceting i could get there also but i need to start somewhere and i like to do that :- )

Making a feceted stone under the 1/100th will be so hard to do that it does not pay off in the end. The polishing will be semi automatic while i dont have any clue to do this full automatic i will run my machine and check each facet and automatic redo some polishing for the selected facets. If i want good quality faceted Stones it wil take around 12 mins for a stone of the size 8mm but can take upto an hour depending how critical i want to work.

I also have some thougths About automatic dopping :- )

I just want to cut stones and make my wife happy lol and get some $ out off the Stones to pay school for my litle girl. And living from some more money than 50 euro's in the week.

 

[vonG]

Damn, 4 laps. Then you don't need to change out the laps either, depending on what stones you are planning to facet.
With those robotic parts for the platen that's used under laps, the runout will be really good. Helps to have the $ to spend I guess.

I think you will do more than ok with the setup you are planning to have. And semiautomatic is probably the best way to go with the polishing. There are vision systems that probably in some way could make it automatic, but they cost a lot and then there is the algorithm that has to be programmed too.

Speaking of programming, are you making your program from scratch or are you using a base program? You wrote before that standard cnc program won't work, why is that?
This I the part that I probably will have to hire in help to get it to work, best would be to have the ability to grab gemcad files into the program and go from there.

If you can get the re-dopping 100% safe, then I think you have come up with something I haven't. The stone isn't that fun to reattach in the right position if it gets lose. Hope you have, because that would make it more automatic, and would be very exiting to see too.

12 minutes is really fast for a good polished stone in that size.

Making some extra money, yes that would be something. Specially when it is automated. Don't think manuall gemstone faceting have that many years left, the ability to fully automate it cheaply will come sooner or later.

Do you have a deadline for your project or are you just going with it as it comes?

 

[Nena]

Hello Thomas,

I dont have a deadline :- ) but i have all the standard parts to build the basic machine now. Im also considering about the cuttinghead. The cuttinghead need to pickup dopsticks, transfer the stone and placing back the stone. (automatic dopping/transfering). I have a realy good idear for that what is going to work.

About the software, i already made a algorithm for computing/drawing the stone with coordinates. I can use a gemcad file and my program is recalculating all measurements sizes etc and it gives a output for cutting it on 3 axis.

I still need to do the rest but i wanted to do some building. The hardest part is already done now that was the algorithm! I post some images of a output file that is generated by my software only by the given machining setup from gemcad.

Then i can controll my axis with my own software. i can use my own sensors that must be monitored what standard software cant. I can controll more than 6 axis at once as many i want. Can controll heat and freeze. I can do what i want with my software :- ) Making these kind of software will cost you $ 100.000 or more! The algorithm cost me 3 months to rewrite 4 times! Its completely 3d and i can control every point in the drawing. Lol was also the first time i did some 3d coordinate programming lol. But the basis algorithm is done! Now i can create every stone i want with gemcad or without it. Also computer generated.

 

[vonG]

I can see that you have come a long way with the program and it seems pretty advanced, but dont you need the 4th axis for the sweep or back/forth movement on the lap? Maybe you dont need that until you execute the program. I have seen a similar program in an automated machine before, they developed it in a engineering university. Probably took some time for them too.

But why doesnt regular gcode work? Is it because the spindel is the cutting tool and not the gem that gets cut? Shouldnt that be possible to change some way with a open source program?
It should be possible to extract coordinates from DXF files to generate the coordinates for the 3 axes too. The machine can be simulated in catia or autodesk and give the programmers what they need to build the end program.

Im really impressed on how advanced your project is going to be, Im looking forward to see it finished and how good polish you can get with it. If it manages to get competition cut stones, then everybody with a manuell machine can put that on the shelf because they wont need it anymore.
If one person can build that kind of machine, then others can do it too.

 

[Nena]

All other axis are not for de actual cutting. They will have there own functions. Sweeping for example can by done with Z rotating axis. The X and the Y are for the location on the machine. They have all a perpose. Thats why there is not a X, Y and a Z-rotation in the list. Its only for the max cutting settings. I need to program the machine to what i like to do. I dont need a dfx file or from autodesk inventor, my program/algorithm does the calculation, also in 3d. It builds the stone fully in 3d and makes the calculations for the lap to cut. Its my base to work on now

For example G-code is machine controlling. You CANNOT controll a natural stone on its hardness! Thats the problem! That means, you cannot keep on preasureing the stone to your lap without breaking the stone or lap or other parts. I have serveral solutions for this problem! I will test my machine on natural quartz and on natural ruby's. Ruby's are realy a problem with cutting if you are unlucky.

You need to feel the stone! Not force it! Think about that! (But you cant feel the stone with a machine, so you must come up with solutions for those problems. I already done my research for those kind of problems, but that are problems where the most people not think about). If the grow of the ruby is leftsided and your lap is turning right it wont cut, if you stay putting preasure on that you properly destroy your lap. These kind of problems cant be handled by standard software!

 

[vonG]

Was more thinking about how I would be able to run my own machine, with gcode or other standalone program. Now I think I know how to do it.
The problems you are talking about I already have solutions for, atleast on paper. But I know it works because other automated machines use atleast some of the ideas.
The other untested ones are just bonus I believe.

I dont have any problems with using my machine semiautomatic, as long as it produces better polish than commercial cut stones.

Hopefully my project will be on the way soon, just have to find the final parts. Dont have any deadline for it, and that is probably the best way to go and get the best deals on parts.

 

[Nena]

Just buyed 2 new spindels 20mm x 1000mm speed 16mm per turn. I shall post my partslist soon including a photo. Another problem with g-code is that it is object orientated. To change a stone from 20mm to 1.1mm is just matter of a blink of a eye! Software is the key for faceting!

Some parts are so expensive... for example, one type bearings of $300 each and i use 8 pieces... so are there manny more things that a hobbyist cant buy. But i think the biggest thing is you need to go full for it! Get the guts to build one! No one will stop me building one! The coolest thing i will build is the automatic dopping system, this is a part i need to let it make in rvs (stainless steel), and the elektronics i need to build for it :- )

I have never seen a hobbyist building a cnc faceting machine, but i know a few people who have made one. But all those people are verry limited in actions what they can do with it. I could also use lasers for scanning the size what i can cut the stone to :- ) im not limited by software!

Ill keep on building, this week wil i make the first parts for the x y and z axis, including the motors.

 

[vonG]

Haha, you almost got me excited now. I am really looking forward to see your machine work it's magic when it's finished.

You are absolutely right in not holding back on any expenses on the machine you are building. There is already a lot of money in it, so why stop now.

I hope it will be as good as you hope in the end, you sure have the software and knowledge to expand it further too.

Shouldn't there be a way to grind off small parts at a time, even if it is object oriented. Lowering the z axis microns at a time. Have some ideas to solve it in another way if that doesn't work though, but don't know if a standard gcode software can handle it. Will just have to see when I get there.

It isn't that easy or "normal" to build a cnc faceting machine I guess, but can be a fun project even if it doesn't work in the end.

Lasers sounds nice. A bit expensive maybe, but hell why hold back.

 

[Nena]

First...

Im going to build it as a CNC Milling machine. Then i will build all other parts. But in the mean time i could go programming the machine :- ) Milling 3d parts is cool to! But i dont think i will do that. I hope it wont take long to create software for controlling the machine but i already have some elektronics that i can use to write my own software for.

We will see, if i can get it work as a milling machine then i can go finishing it for a faceting machine :- )

I hope you can build one to!

 

[Gilly47]

WoW did not realise that Gemology could be this intricate

ray

 

[vonG]

Gilly47, looking rather complex when Nena shows his programming skills.
But faceting competition grade gems manually isn't that easy either.

Nena, are you going use your mill for some of the parts that you need? Like the arborplates for the laps and the automated dopping system.

 

[Nena]

Hehehe, this is just easy talk, the complexity is still to come :lol:

Yes, some of the parts. The most how ever such as the construction parts makes a friend of mine. Becourse he uses a milling machine from over 1 million $. This will cost me around the $500 the cutting head (twice) will cost me around the $200 each. I do this becourse he is a machinist and he's making aircraft parts. Ill do the drawings he's going to mill the parts with high precision!