School still locked down but some Students/Alumni(prior students) want to do some So we embarked on a project to try to build a CNC - a Simple one. We don’t have much money but we want to see if we can do something that can help us with certain tasks like for example cutting a 1/8 in aluminum plate to be the belly pan - hopefully better than a student with hand drill and sawsall does. So here are some pics (of course we use 3dp to help lol)
Initially we were going to belt drive but its pretty big. The dimensions are going to be about 4x4 foot the cutting area up to about 38x38 in. So some tests showed that belts would stretch too much so we are going with a 12mm Leadscrew and a DIY no backlash nut (made out of 2 brass trapezoidal nuts that go with the leadscrew.
Th Y axes will have a motor on each side. The X axis most likely 1 and so the Z axis. It will be driven by a 24 Volt PSU and a 32bit SKR 1.4 controller running hopefully reprap or if that dont work marlin And we are going to use GRBL to make the Gcode
The idea is to mount different tools to the contraption. Like a hand drill for drilling holes, we have been donated a Hand held router which we plan on mounting and a dremel. If we find some more $$ we might get a proper spindle and chuck. And mount a shop vac we have to remove the debris. So far we are about $350 in it and hope that is it (plus the existing tools) Idea is to have an interchangeable Z carriage to which the tool is mounted and then you fix it to the CNC and Home it.
Is it going to work? Who knows. In the worst case if it doesn’t the SKR 1.4 will power a printer the 2020 and threaded rods will wind up in a robot and the steppers are going to work on a 3DP just fine - so does the PSU. So we will either have a <$500 CNC that works or some parts to build a robot and maintain the printers. How long will it take? who knows?
Very nice! Good investment in a summer project. Even if the router only cuts plastic plate it can be incredibly valuable for a team. You might find yourself designing robots to use more polycarb, delrin, hdpe, uhmw, pet, and other great plastics which can quickly be machined on a simple router for most FRC needs. Not to mention you can make some very nice wooden parts out of plywood, especially Baltic Birch (great for bellypans). Also a great chance for kids to learn CAM software which can lead to some well paying summer jobs and careers!
Best of luck! You may want to source more of your parts from places like Alibaba. I’ve seen some very cheap spindles that online reviewers found to actually perform pretty darn well.
What else will you do to ensure the frame is as rigid as possible, as well as the screws for each axis?
My limited knowledge of 3DP Cnc tells me that it’s very possible that you’d have to replace those parts frequently. Nonetheless, I’m very excited to follow this project, because not only would it be a great learning experience for us to build one, but we also want a cnc. I look forward to seeing your progress!
That is one of the ideas. I have experimented with wrapping both things like cut hardboard and 3dp stuff into Fiberglass and so far it looks promising currently using polyethylene resin (cause its cheap) and once all the bugs are ironed out probably the “good stuff/epoxy”.
Here is one of those pieces glassed with 2 layers of 10oz Fiberglass on each side the holes were cut with an exacto knive while it was still kinda wet before hardening using the 3dp holes as a guide
In the meantime there have been tests with foam, card board, hardboard and misc plastics etc using anything and in different combos of 17 oz matte and then covered in 6-12 oz cloth (depending what was on sale)
This was one of the less successful tests a crack at a test of a bellypan. The thing that did not work those nipples were to poke through the cloth and then you can push through the plastic and put a bolt through so you dont have to do all the xcato knife work. It worked fine for a little bit the problem was pushing the nipples through the matte took too long and it hardened before pushing them all through and then there was no time for rolling - so it looked terrible - but it gave some good data on having a 3dp core with a layer of 17 oz matte and 2 layers of 6oz cloth on each side. - Very strong
So one reason for the CNC is to get a big piece of something easy to cut but not that strong and then cover it in FG. (CF is the same process but way to expensive for us ATM) I figure we can do a skateboard frame in FG for way under $100 with a 3dp core (HIPS) acting as a skeleton and precision guide with some help from Hardboard ($10 for a 4x8 sheet) and foam and then covered in 1 layer on each side of matte and 2-4 layers of 6-12oz cloth. Structurally you are about approaching a canoe here.
We might fiberglass some of them also they are holding 2020 aluminum extrusion in place plus in some places reinforced with 1/2in 16 gauge square steel tubing. Now we had a similar construction (instead of steal 1/2in al. 16 gauge tubing) survive a whole day of playing “defense” last year at an off season competition. So the logic was if it survives bumper cars with a bunch of angry robots lol then it should be able to drive a tool around - but we will see.
I’ve been involved in a couple iterations of some printed/accessible CNC’s over the last few summers, as part of a research lab I’ve been lucky to be involved with. Let me know if you want more info… anyways looks great but my three bits of advice are: use as big of extrusion as you can get away with, what youve got now looks a little thin and with beeft extrusion even mostly printed machines get darn stiff, secondly the beefier the connecting bits the better (though thats fairly evident), and finally preloading rolling elements does a ton for smoothness and stifness (thinking about it from a kinematic perspective constraint wise is always handy when doing this).
Thank you the rolling assy is pre-loaded. IOW it runs way smoother when everything is mounted to it. The brackets are not the final version. Its more efficient in print time to get everything to fit first. The wall thickness right now is 5mm as that is a good balance between strength and printing speed. IDK yet if we will print some of the parts heavier or wrap them in fiberglass. The X and y carriage should be easy to wrap the legs on the corners are quite complicated and probably difficult to wrap. We do not have infill as everything is printed with enough perimeters to make a solid part (no air inside) currently we are using hips but we might use some other materials for some of the final parts. So right now I am leaning towards abs for the final x and y carrige Assy and Nylon for what will hold the brass lead screw nuts (as you can make them slightly off and use the springyness and toughness of the Nylon to make a 0 backlash nut (hopefully)) The corner legs probably stay HIPS as the motors mount there and hips does well with vibration and still gives more stiffness than most Nylons plus is much cheaper. Many/most parts will be hybrids like our robot Like you can see that the X carriage is a 2 piece assy with the inverted U channel design and the holes that u channel will tightly fit a 1/2in steel 16 gauge square tube and he holes are there for self tapping screws to or maybe nuts and bolts to hold the plastic to the square tube. So It needs to be strong in certain places - but yes I am happy for any pointers we can get.
Oh and should it work - all drawings and inventor/F360 files and firmware etc will go on grabcad
Another cool trick we learned is that if you design your brackets and things in the righr manner once you have a working router you can upgrade eveything by milling the bracket pieces out of phenolic and reinstalling them. We found phenolic to be the sweet spot in terms of being really easy to mill with non rigid machines, cheap, and an order of magnitude stiffer than 3dp.
GRBL is actually a controller pretty much equivalent to RepRap firmware and Marlin. It cannot generate Gcode itself. GRBL is pretty much exclusive to CNCs, where as Marlin/RepRap Firmware are designed for 3D printers (although I have seen some folks have some success with running them on CNCs).
If you chose to go with GRBL, the controllers can be pretty inexpensive and run the larger nema 23s (rather than the 17s that usually go on printers). I once built one for like $50. I used these drivers and an arduino uno. You dont have to, but you will have a much easier time wiring with one of these.
For controlling the machine and generating gcode in GRBL (and I think marlin) I like these softwares: https://software.openbuilds.com/
Also, when you brought up 3D printed CNC, It made me think of a project that I saw a little while back. The root 3 uses Wood and 3D printed parts to make a pretty capable CNC. I have not personally built one (although strongly considered it once). It can maybe help provide some inspiration.
Hope this helps This sounds like a very fun project for your team!
Thanks for all that info it should help bunches. I saw this and similar controlles and it would have put the machine way out of our budget range. Going with leadscrews was almost impossible. But lots of great info
We did something similar back in 2016. For the controller, We used a Arduino UNO and an DVR8825 driver board like this.
Many choices for kits. We used v wheels and belts. For Cam and controller software, I’m a big fan of Estlcam. https://www.estlcam.de/ I find that young students catch on to it very fast. We also use it on a big router we refurbed. For the big router we use an Arduino and TB6600 drivers mentioned above. You might want to look at Openbuilds for parts to purchase and ideas. https://openbuilds.com/
We got DVR 8825s too. only currently we got an SKR 1.4 board scheduled for it as that can be used in a 3DP as a replacement too. As for the limit switches we got both some mechanical and some optical ones
This is a 1500mm x 1500mm CNC router that we (FRC# 4336) built a couple of years ago. Nox CNC Router
Lots of good info on the openbuilds.com site.
We used belts for the X & Y axis. Here’s a video of the belt and stepper motor. The belts are overlapped on themselves, so only the small loop for the stepper motor pulleys are free hanging. This works wonders on reducing any stretch. We have used this router to mill out numerous aluminum plates: 1/4", 1/2" and even 3/4". Tolerances have been good enough for gearboxes and we can easily make a hand pressed fit for bearings.
Ours is kinda based on this pic only DIY with the resources we have and the size we need - like I’d love to use linear rails instead of 2020 extrusion with wheels
Let me know if you need some plates milled for your machine. If you can provide the CAD, we would be happy to help out with the milling @ aluminum cost + shipping. Your tall Y-axis plates would probably be around $20 ea +shipping depending upon thickness and exact size. We now have 2 HAAS Mini-mills. So tolerances are VERY close.
Appreciate the offer and we might take you up on it. Right Now those endplates are 3dp and we see how that goes. We will see what we can do and we want to try it - you could say - the reprap way. Which means some parts that might need improving we will improve with the machine itself. The goal is not just to have a CNC we can afford but the whole learning experience - the journey to a CNC you might say. And then there are other challenges it might help with that would fill another thread. Like keeping kids excited about the team without knowing if there will be a competition or when we can meet. Right now it looks like that - at best the team can first meet in person after christmas (unless there is a vaccine widely available way before then) District is talking phase 3 for indoor after curricular activities when we are on full schedule. Right now the challenge is to find ways of keeping the team engaged which is challenging. Right now there is some excitement - working on the competition robot is still painful so its probably more a way right now to keep the team alive. And to show to some potential past and future sponsors that we are resilient and worth supporting and working hard. But I digress. We and I appreciate any help but I also want to make sure some ideas will get a chance of getting tried. If for nothing else than the designer can see his/her part perform however it will perform
You chose stepper motor wisely buddy. I appreciate that and you do some fantastic work that need to be appreciated.
The best part is that the stepper motors are simpler in their design and easier to use, which makes them immediately a better option for the people who are new to the world of CNC machining and there are many factors which comes into consideration while choosing them like -
Torque rating
Unipolar and bipolar
Power supply
There is a complete article on Stepper motor which might help many who are willing to build a CNC machine like you.
This sounds like a very fun project for your team!
