One thing to consider when printing functional parts is most infill geometries will provide strength in only two axes. For most prints this is not an issue, but in functional parts it is something to consider. For functional parts I use gyroid infill to ensure maximum strength in three directions. With gyroid the infill percentage can generally be halved without loss of strength; important if you need to keep the mass of the part to a minimum for such things as 3d printer parts attached at the hot-end. The downside of gyroid infill is some geometries will cause a great deal of vibration when printing. Long runs of infill are fine but small areas of infill tend to vibrate the whole printer quite a bit if you have a moving bed. CoreXY printers have fewer issues, the only one being the chance of greater infill-ringing if the hot-end is very heavy. In all cases, printing the infill slower is the answer. (Some firmware and slicers can be trained to slow down when small areas of infill are deposited. Check your setup to see if you have this option.) Luckily, one other benefit of gyroid infill is its much shorter print time with fewer large hot-end movements, so slowing down the infill speed should not noticeably increase the print times. I have printed parts in PETG which prevent flexing of a steering shaft in a track focused sports car. With the majority of infill types a percentage of 60 to 80% was required. With gyroid only 40% infill resulted in a part with almost double the strength. Both prints (cubic and gyroid) were annealed at 80C for one hour after printing. For functional parts I always recommend annealing to prevent unexpected warping or shrinkage.
1:29 if you went to the effort to redesign this part, why not make the Z axis travel larger or at least spread out the existing linear rods above and below the screw? Would standard Nylon 6 be a suitable material or would it not be rigid enough?
Excellent follow-up video to your other mod video. This machine is just what I need to make items for my Etsy Shop. Looking forward to seeing the BOM. Keep up the good work.
Great video! Can't wait for all the details and files. As soon as you made them available I will start ordering stuff and try to redo what you did. Great work. Any idea when you have all the details?
Great idea for the 3rd rail, but it looks as though you added flex in the collar around the spindle motor by making it longer and thinner. In the video shown from the side, it looks like an even trade for no net gain.
Thanks Rick. No flex on the collar, it's 100% infill PETG so pretty solid. The flex comes from the rails still, while it did improve it did not eliminate. If you want to truly eliminate flex you'd want to buy thicker rods, or even better linear rails. However, at that point you might as well buy a bigger/better machine. My goal here was to improve the machine at relatively low cost. The main gain was being able to add a spoilboard, check out my more recent video on that👍🏻
0:37 you made third. i made second x and z axis sandwich spindle between. NO PLAY. it was so good i upgrade spindle turn got 4 axis added chuck have 5 axis.i believe i still only one LOL then i made all new bigger without linear rods and leadskrew.used linear guide rails and ballskrew less NO PLAY
I've thought of doing something similar, however I was always curious why they didn't take advantage of the aluminum extrusion right behind the rails, as the rails themselves are most of the flex a linear rail compared to a rod would be much stiffer
Man this video is old :) but just some design advice, never leave sharp inside corners. Add a fillet to improve the strength as the sharp corner are stress concentration points. Other than that good video.
well... the next upgrade would be to get the 3018 Pro, ProVer, etc, with the 20x40mm extrusion sides and bakelite or metal side plates... i've got the Pro, and there's just a little play in the z axis rig (which is a metal replacement one off banggood, with a '500w' spindle) and i'm considering replacing the rods with linear rails.
I was almost done doing the same upgrades when I found your series. Trust me I would have become a member on your site just for the time I would have saved drawing up and printing my own. I have a stupid question though. Are you powering your board on a separate 24 VDC? I am also running the larger motor on its supplied 48VDC PSU but I can’t find anywhere the max voltage for my board or any others.
Great upgrade; it's turned out very well. I'm curious what the control board and stepper drivers are - the drivers look like DRV8825s. How much current did you need for the larger Z axis stepper, or is that still "tuning in progress"? I notice heatsinks on the MOSFETs (I assume) but not the stepsticks.
By any chance do you know what kind of lead screw it has to drive the x and y i tried to upgrade mine and the lead screw were not the same i was able to run it but since i only changed the x the height was was off, im not even gonna try messing with the code for the x
This has bothered me for a while and maybe you can answer it... It seems to me that a z axis backlash spring should be pushing the carriage UP ensuring the z steps would push the tool into the work piece, ensuring thread contact. I consistently see these installed with the part attached to the z axis carriage on the bottom, like yours. This would put the pressure downward and any slack would be exaggerated. If the spring on the backlash spring cannot hold the weight of the z axis carriage you've lost the effectiveness of the backlash assembly even installed properly. The only way to ensure this is to have the spring push UP on the carriage, meaning the part bolted to it is on top. Does my logic make sense?
Any chance of getting the X carriage and 3rd rail mount stls? I'm getting ready to order a 3018 and want to have the parts printed before it arrives if possible so I only have to assemble it once :-).
Hey, I'm new to the CNC world. My "making" started with a Tronxy XY3. I hope to prototype with thick aluminum and thin steel. Affordably of course. Would this unit and all these mods help me be successful with those project materials?
Not without installing actual linear rails. He installed a third shaft guide, which flexes in all directions and are not linear rails. Adding a third one adds negligible rigidity to the system. Actual linear rails are mounted to the frame of a machine and are rigid in all directions. And the 3018's actually are very easily converted to have linear rails as they have two frame beams close to the X axis (one right behind the Z gantry which is ideal to convert to support the weight and forces from that gantry). Those beams you can mount linear rails to easily and make even more rigid by replacing them with a bigger profiles (like 4020's or 4040's) if need be. Shafts and shaft guides float between the frame and flex, the longer they are the worse they flex. Linear rails don't, they can take the force of a Z gantry pushing down, lifting up, torquing, doesn't matter what direction the load comes from. With the 3018, the full weight of the Z gantry is supported by 2 guide shafts and the lead screw, all three things that flex like mad. Adding just one Linear rail behind the Z gantry will make the whole thing exponentially more rigid and capable of putting force in any direction without flexing remotely as much, which makes milling aluminium possible. To do it at any decent speed, you may have to fortify the tower beams to the sides and then you best add a second linear rails to the upper beam behind the Z axis (at which point you really just need one shaft guide to buffer some vibrations) and either add a linear rail to the Z gantry or replaces both shaft guides with linear rails. At that point, the whole XZ system is solid enough to cut aluminium at a decent speed with a decent surface finish. The only prohibitive thing about doing this is that even cheap linear rails (which will already be a vast improvement over shafts) will still cost you 50-70€ for 2x300 for X and 2x200 for Z and you have to have access to a 3D printer or other way to fabricate methods to connect your gantries to them.
As I said in the video, it still moves but it's a lot less than before. It is a very low cost upgrade that slightly improves the capabilities of this machine.
This is good start for learning CNC machines and understanding how it's works. But there is big problem: You can't realy do this! Why? Because RIGHT feeds and cutting speeds depends from spindle speed, spindle power, cutting tools and material hardnes. This is calculation between tool toot count , tool form, cutting deep, cutting direction, cutting owerflow percentage, material hardnes coefficient, spindle rotation speed. For normal cutting all parametrs should bee right calculated. Finaly you see, that your spindle is to slow and for weak for cutting wood... and you tool wearing out wery quickly, heating up and screaming. I know what I say, because I made many woodworking machines(Excluding CNC) and know about parametrs above. I bought this machine , like you, for learning all about, but quickly understand, this is not possible with this small machine.Yes, you can adjust some parametrs for you machine, but not every time,and you experience will be wrong. This machine is good start to learn how to make one better machine, price is low as poosible and save lot of money for you in future, helps avoid big mistakes and loasing money(This is coolection of all impossible mistakes for CNC) Learn all this mistakes and DON"T spend money for this machine, collect for one new!
I agree sort of. My Dad was a MANUAL machinist for 40+ yrs and understood all of that. Time is money and cutting air or using improper feeds and speeds is not a wise use of the machine. For the size of this unit and it being specifically for hobbyist to learn the basics of Gcode commands, fixturing, tooling, etc, I think it's "good enough" for us without breaking the bank. Yes, we'll break a few end mills and scrap parts at first but we'll either learn from those mistakes or give up and do something else. Most of the time when I "need" to make a part it would fit on this machine. That gives me a chance to keep my chops up on F360, design the part, make it and have the satisfaction of doing so. I don't care if the tool paths are optimized for efficient cutting or it takes longer to complete the cut. I'll just have a cup of coffee while the machine does the work :-). I just wish he would offer the "upgraded" F360 parts shown separately instead of making me join his website. I have no interest in any of the other parts and can if necessary recreate these parts in F360. I'd be willing to buy just the X carriage, smooth rod plate, 300 watt spindle models. Not sure why he didn't take the opportunity to increase the z rod travel when he remodeled the X carriage. Looks like there's easily an additional 25mm or more to be gained. Did he add a bearing in the bottom to control the Z lead screw "wobble"? Don't know since I can't examine the model. I will if I have to redo the parts myself.
@@home-dp6oh Yes! This is good enough for understanding that this is yust toy...Nothing more. Biggest inventors problem is difference between manufacturing "laboratory model" and manufacturing production process... All you laboratory experience becomes unusable! The same will happen with this experience. Why? Yust one example, rigidity. When you start with real machine, all you problems will dissapear! There you spend lot of time for solving this problem, you should change feeds and owerflov, searching best way and making lot of machining pases. Again and again. And more... Yes, I agree with you that you never spend lot of money(I hope), but better learn all mistakes from this toy and save money for building one realy good machine. And you get mutch more satisfaction. Mans remain kids all time and should play with toys. Only toys becames more expensive and mans older...
