Neil, if you every decide to build another version, contact me because I will share with you the files in stp format because yours was one the builds that inspired me to push the design further! Yes, I'm of the breed of designer and engineer who finds it hard to stop. My wife nailed it on the head this weekend. "I thought that design was completed and you have spent another 10 hours in the world of the computer tweaking the project " I will show some great new end plates designs in a few days!
Awesome, thanks! I see you have a great wife like I do. In fact any woman that marries an engineer is a SAINT!
My wife is an interior designer by training, so she understand why it's hard to stop designing when you have time available. I have ordered the extra linear bearings from China. So the size constraints are defined, but that leaves me with 3 weeks more tweaking!
The Chinese certainly come up with some bad designs - they are probably just fine for cutting foam products but I wish they would not make claims that they are good for hardwoods, metals etc. Just as a 'by the way' I came across this design ( V-Slot principle ) which, surprisingly, does work extremely well with a laser (which has no tool loading) but it could certainly do with some improvement to make it last long term. Tweakie.
I worry about eye safety with those unprotected lasers, but mechanically they're pretty straightforward. This stuff is getting to the price point where just about anyone can get into it. Sorta like what happened with digital cameras -- the quality vs cost ratio is getting really high.
Hi, I am amazed at your concept! This is my first post here. I am currently building the OX, bought in kit form on SMW3D. Just need to figure out which electronics I am going to mate to it. I guess I am somewhat normal that I am already thinking of changes and upgrades. Your machine is exactly what I was thinking about (no, not the design etc. which was beyond anything I could imagine - so THANK YOU!) but having something extremely rigid and precise for aluminum milling. How big/powerful spindle will you be able to put on your machine? Essentially, I would like to be able to use the bigger end mills and surfacers to mill blocs and hollow out blocks of aluminum. Thanks again for sharing your experience and images of your build. It is inspiring. I just need to figure out which CAD package I want to commit to. I am thinking of Fusion 360 as there is a CAM component incorporated in it. I think it is also parametric. Currently quite comfortable with SketchUp. I design all my woodworking projects in it. Thanks for comments. Bertrand
Really liking the steel frame update to your design, Been toying with the idea of building a smaller cnc for the sole purpose of cutting 1/2 and thinner cast acrylic and 1/4 and thinner 6061 aluminum. Originally considered the usb x6 1500gt But not my first choice if I can build something better. I look forward to following your build.
Welcome to the forum Bertramd and your kind words. Building a standard OX is a great learning experience and really worth the effort. Thinking and designing mod's is also lots of fun. The standard belt system is a serious weakness but inexpensive and gets you started but plan on going to the double thickness or replacing it with leadscrew or rack and pinion. I originally wanted cnc to machine 3mm carbon fiber, but since learning more about Cnc machines I decided it would be good to have a machine that can easily cut aluminum. I use Solidworks for design, aspire and mach3. Currently learning the cam and machining aspects. Solidworks is an outstanding 3D modeling and engineering solution but expensive for hobby person. It's own of the easiest 3D systems to learn yet very, very powerful. I plan to use the dewalt 611 or Bosch 2 1/4 hp routers to start off with because I already own these.
Hi Steve, Thank you very much for your answer. I am unsure what you mean by "going double thickness". I just ordered my Soothieboard to finalize my build. Can't wait! May I ask what you are planning on using for motors and board? Will you be using NEMA 23's? I'll look into Solidworks but depends on cost. Having the CAM aspect integrated would be better as I am only starting in this hobby. For now anyway, I need to configure my setup and learn how to use it. One baby step at a time. I'll start learning about how you are planning on doing your modifications. I don't think I've seen it but how big will you new machine be, meaning how big a piece can you work? Thanks. Bertrand P.S.: I'm also in Canada, Montreal.
Hi KYO, pleased you like the steel concept. I agree you can make a better cnc than the x6 q500gt for $1800 The nice thing about building it yourself is you pick and choose components, size and design. After much thought and planning I started to ask myself why use aluminum for the base structure when steel tube is much stronger, stiffer, cheaper and resist vibration more. The only advantage is aluminum easy to drill, tap and bolt together plus you don't need welding skills. I will continue to use aluminum for the X structure to keep this releativrely light in weight.
Hi Bertrand nice to see another Canadian here. A few members have used two GT3 belts per axis to reduce the possible stretch. Prauk did and excellent build write up and video. Http://www.openbuilds.com/threads/the-buffalo.731/ I bought Grecko G540 because this company has a very good repution, plus wiring and electronics are not my expertise. http://www.geckodrive.com/geckodrive-step-motor-drives/g540.html Motors I purchased 380 oz steppers, knowing I would upgrade from the GT3 belts. http://www.cncrouterparts.com/380-oz-in-nema-23-stepper-motor-14-shaft-p-151.html The new machine, I'm still not sure if I should jump up to nema 34's? But thought I would see how the nema 23 work first. I don't need a large machine, so decided 1000mm by 750mm bed size was good enough for my purposes. Sadly Solidworks starts round $2500 so beyond most hobby people. I have worked in 3D software development for 25 years on surface modeling software (Alias surface studio and Maya) Solidworks is the best of the best CAD solution for professional product design engineers in my opinion. I have been using it now for 12 years for my wood working and sim racing projects. If your friendly with a Student engineer, they get a wonderful student price for a fully functional version.
Hi Steve, Thank you very much for your answers. I'll be finishing it and hopefully get it to run when I get my Smoothieboard. I'll strat my thread at that time. Unfortunately, I don't have a student friend in engineering. Thanks again. Bertrand
Man, you deserve the tittle Master! If I wanna go bigger and stronger, I'll most likely will give your design a go!
I am very impressed with this build. I'm amazed at your design. Can't wait till this is complete. What do you think your total cost will be?
Thank you Paul & xLordXSidiusX Ball screws are around $150 each for 1000mm, linear bearing are around $80 each for 1000mm I'm also looking at alternative solutions because I'm considering offering complete engineering drawings for this machine because I think there is a real need for a strong reasonable cost CNC machine. These are interesting alternatives, which I'm currently getting prices about... http://www.igus.com/wpck/3555/drylin_t http://www.igus.com/wpck/3596/drylin_w http://www.igus.com/wpck/7834/DryLin_Anti_Backlash_Muttern
Hello Steve, Thank you very much for the information. This is a treasure to have for planning a future build. Trying to find adequate parts is difficult when you don't really know what you are looking for. Any additional sources you want to put in is welcomed. I found this one but they seem to be quite expensive: http://www.pbclinear.com/Pages/AllProducts It may be only marketing speak but one company said that having found bars unsupported increases flex and the bars themselves may not be very strain to begin with so it is better to go for supported bars. I believe this is what you have in your design. What to you think of the suggestion of going with square bars instead of round for the load issue? Thanks. Bertrand
From what I have read supported rails are much better than un-supported which I why I purchased supported. From what I have also read that square bars are much better for even load distribution but the price difference is much higher. I honestly don't think my machine warrants the price difference because I don't machine stuff full time. The igus bearing looked interesting, because the double rail provides precise alignment for the long Y axis but I'm waiting to see what the price is...
Steve, Brand new to the forum and the idea of building a cnc myself, but have to say your design and research is really impressive! The site is very cool too to credit the openbuild team and the other builders who freely contribute their ideas. I've learned so much in the past week lurking and reading through the "ox", Neil's frog, Area 51, as well as yours. Thanks! Looking forward to seeing the final design and build. Best, Jeff
Thank you Jeff, it's been really fun designing and evolving ideas. I really enjoyed the standard OX build it was a very simple process using Marks design just cutting, drilling and tapping and assembly. My latest design made from steel tubing is a little more difficult for most people, but a fully welded steel structure will be extremely rigid. I agree lots of really helpful people on this site always willing to help out and educate us novices!
Nice job Steve. I wanted to go with a rail system myself but the cost was outrageous, at least for what I found. My Ox still has a little deflection so I haven't attempted aluminum yet. I'm really interested to see how your machine handles tougher materials.
Thank you Scott. I'm off work at the moment with vertigo issue, so I had lots of time to request info from a few manufacturers. Wow, as soon as you look at the major suppliers the prices are a real shocker! $1500 is typical just for the linear rails & bearings for my small machine... Agree 100% really outrageous. The ebay open supported rails from China are good enough in my opinion at $90 for two... A few nights ago I found a local sell off on new European Bosch Rexroth brand which are very tempting... Because they are just a little more than the Chinese. I priced the steel tube today, this was very reasonable for 4" * 2" & 2" * 1" 1/8" wall thickness.
Steve, The steel tubing looks like a really cost effective way to bring rigidity. Sorry to hear of the vertigo, I had that two years ago and couldnt get off the couch for a week. I was talking to a colleague yesterday about solid modeling and he mentioned "onshape". I thought I'd pass it on for people without access to solidworks. Its new parametric modeling done through a web browser ... I know crazy. It was developed by one of the solidworks guys. They have pro versions for a monthly fee but the cool thing is there is a free version that would probably work great for quite a few people on here. It looks like a pretty full featured program. The downside is they limit how many private assemblies you can have so the rest would be public, which might not be a limiting factor for most. Hopefully this isnt old news, didnt want to hijack the great thread but Bertrand was asking about a modeler. Best, Jeff
Hi Jeff, a sincere thank you for posting the new web product information called "onshape", it looks really amazing especially when you consider it runs in a browser. I signed up at, it's great you can learn for free using a fully featured product! Parametric modeling takes a little getting used to but is the way to go when design parts that need's to fit together. So much better than surface modelers like SketchUp or Rhino unless you are creating complex freeform surfaces. It's funny that some SolidWorks developers have started a new company, the original SolidWork team worked for PTC and they broke away to start SolidWorks. Vertigo is nasty illness, I'm now on week 4, last time I had it was many months... It's really slowing down my CNC build progress. Thank you for the concern! Yes, steel tube is cheap & very strong. Under a $100 for all 1/8" wall tube... Steve
Steve, That is a great price. My father was a welder for 40 years before retiring. One regret I have is not cohoercing him to teach me the skill when I was younger. Hmmm, maybe it would be worth picking up a welder and having him show me. Always a good skill to have. I read in one of your posts...you're an old alias power animator guy. I used it back in the sgi days. Great nurbs modeler. Rhino is pretty sweet too for surfaces but for this stuff parametric is the only way to go. I was curious how hard you found wiring your first ox you made? I'm no electrical engineer although I've done plenty of home wiring. That's a concern I have in starting an endeavor like this. Vertigo totally sucks...it was a week long hangover without the drinking I wasn't 100% for weeks. I hope it subsides quickly. Jeff
Jeff, Welding with a modern mig welder is very easy to learn, traditional stick welding is much harder. Your Dad could teach you the fundamentals of mig welding steel in under an hour! IF you build a CNC from steel find good welder who knows how to deal with distortion and welding all the joints in a good sequence. Dad will know what I mean! Welding the cut parts for a skilled welder should be under an hour, so should not be expensive. Drill all the holes required before assembly... I served apprenticeship in the steel industry, then switched careers to computer graphics back in the 80's. I worked in the Alias development team for 16 years, which was lots of fun. Nice to hear you used PowerAnimator back then. I subscribed to Onshape it is excellent! Thanks again for the tip. It has many of the fundamental concepts of SolidWorks with a few new twists. I'm also not an electrical guy, so I bought the Gecko 540 all in one unit and have no regrets... A few simple wires and away you go! I have read so many people having issue with the much cheaper solutions, and I still feel it's worth the $279 for 4 drivers in an all in one solution.
Hello Jeff and Steve, Thank you very much for the recommendation. I will be checking this out. I'll be using it to learn parametric CAD. Thanks again for all the great help and time to provide information to newbies like me. I'm still waiting for my Smoothieboard which seems won't be getting to me before another 2 weeks so no working OX for now. Have a great day. Bertrand
Your most welcome Bertrand, happy to help out and share. Parametric modeling is ideal for engineering modeling where many parts need to fit together or can be related to each other. For example you create a simple sketch (2d drawing) with a square and circle and then use this drawing to create a 3D surface using a function such as extrude. You then add that part to an assembly. Now you can create another part in relationship to the first part and add constraints such as co-linear, perpendicular, equal, or measurements, etc. At any time you can modify ANY part and all other parts will automatically update.... In the case of a CNC machine I authored a sketch with a rectangle 4" * 2", added some radius curves on the corners, and offset the curves 1/8". Then you chose this sketch and the extrude function to create a 3D model called a part. Next you add the part to an assembly. Next I import the rail again and create constraints between these two parts. The two rails are parallel with a distance between them, aligned the end & top surfaces with two more constraints. Any surface in a model can be used to author another sketch to produce more 3D content. By using surfaces as construction plane's all the 3D content remains related. So if you adjust the width of the machine then the complete machine will automatically update! Awesome... All the main steel frame is one assembly. The X gantry is another complete assembly, Z axis another. The linear bearings for each axis are another assembly. If you model it correctly you can change any part and the full machine will update automatically. The downfall of parametric modeling is when parts are modified and the relationships between all the parts cannot be meet! Which sometimes means finding the invalid constraints is hard!
Hi JustinTime, I agree that working with parametric models takes a little getting used to especially if you come from the world of surface modeling like Sketchup or Rhino. It requires thinking in a totally different way. You need to think about the separate parts and relationship of these parts together. STL export should always be flawless because it's a solid volume model which is watertight, unlike a surface model. It's funny I find SketchUp very awkward to work with, I love the sketch tools in SolidWorks & OnShape and the dimensioning of drawings. When it comes to CAD tools everyone has preferences. No right or wrong solution.