Anyone built a CNC machine? Would love to build one in my upcoming uni break but know fuck all about them
>>971911
https://www.youtube.com/watch?v=WKmzlMvxC7I
look up this channel
this guy is pretty awesome,
I've been working on one for a year, close to finishing it. Didn't have much time to work on it due to uni and shipping times from China are shit too at 3 weeks.
Mine's controlled by an Arduino Uno using DRV8825 motor drivers and a driver shield, small NEMA23 steppers driving 12mm trapezoid rods powered by a 240W 24V PSU. It has a 550W router mounted and the frame is made of 3030 alu profiles, a 30mm version of the 8020 profiles.
Initially I was going to use 12mm steel rods with LM12UU linear bearings for mounting each axis, but Y and X bent under the weight way too much, so I splurged a bit and got a bunch of these:
http://www.aliexpress.com/item/SBR12-rail-L500mm-12mm-linear-guide-cnc-router-part-linear-rail-SBR12-rails/1313496588.html
These are solid as fuck. I want it to be good enough to mill aluminum though. I've spent around $600 on it, but a lot of that was on mistakes, if I knew what I was doing I could've gotten away with around $450.
However, if you only want a small CNC for wood, plastic and PCB, stuff like >>971929 or your picture is good enough and a fair bit cheaper; NEMA17 steppers and a dremel mounted. You may use belts in these cases but belts stretch, so you can't really achieve great precision in harder materials, aluminum is definitely out of question.
>>971911
I'm trying to do this also
right now I'm just working on the motor control circuitry and software
http://bcndynamics.com/docs/Manual-MaduixaCNC-rev1.pdf
>>971911
I've built two of them. Old pic of almost-complete machine. I've since added a t-slot table, got some important bits (VFD, drivers, the PC that controls it, etc.) stuck back on.
I had some issues with occasional missed steps due to interference on the signal lines, so I made an aluminum box to stuff the drivers in.
...unfortunately, I didn't lay it out very well and the signal wires are all bunched up together in places. Rather than fixing the interference issue, it made it so bad that the thing is unusable. I've been meaning to re-do that box, but I've been too busy procrastinating.
So, not op here, but am interested in CNC anyways.
I have a few I-Beams here and I wonder if they would make a useful base for a CNC machine. Like welding it together to create a portal, kind of similar to the top picture.
I have an angle grinder, welding equipment and elbow grease, but I don't know if it's viable, precise or sturdy enough to do it in this fashion. Any thoughts?
Most CNC Mills I see on youtube that are /diy/ are not really suitable for milling steel. I wonder if I can get that if I bolt the welded frame to a welded table, hopefully making it heavy enough.
As a flat surface for welding, I could possibly get access to a plain stone from a mason friend of mine. That or a new unused gravestone without enscription.
Any ideas or experiance?
>>972154
"Flat" as an average machinist defines it is only second to what a machinist making laboratory-grade equipment does. Unless it was specifically manufactured for extreme flatness (as is the case with tooling stock or surface plates), you're going to need to figure out some way to get it flatter than it already is.
If you're making a smaller one, it may be worth your while to fork over the cash for a surface plate. (I found a 24x36x4" one locally for $150.) You can use that as a reference surface for scraping and etcetera. Or, if you can find someone willing to let you use their's, that would also work (especially considering that, if they have one, they probably know a thing or two already and can help).
Or, if you don't care about the poor tolerances, there really isn't a reason you can't use them as-is. But at that point, there's not much benefit to using steel in the first place.
>>972161
Thank you for the solid advice. I have contacts to some precision merchanics who have access to mills and lathes, maybe they can help me out (one guy did some motorcycle part for me, one of two who offered to do it...)
Now theoretically and with a lot of time I guess it would be possible to flatten them with sandpaper...but it would take forever and would be prone to human error. Also the idea of grinding something down with sandpaper reminds me of the horrible workpiece of a metal dice for some reason. That thing came out abysmal...oh well into the furnace it'll go at some point. Leave no evidence of horrible work :^)
Maybe I could grind them down roughly with rather rough and later rather fine sandpaper, then start polishing them with a polishing machine. Though I don't have any real reference to "flat" here ...
As a matter of fact I do care about tolerances. Because, to be quite honest, if I go through with this I'd rather have a beautiful work of art and not some shitty half-useable piece of crap machinery.
I have to re-evaluate the thickness of the I-Beams but they were pretty thick, so I guess grinding them "flat" would be the way to go.
Just the thought of it...oh well, no use in rambling...
>>972138
damn you are good
how much did it cost? how long did it take?
>>972046
Are you following a specific open source design? I'm interested but I lack the knowledge at this moment to create one from scratch, without plans...
>>972609
I'm not, I didn't find anything suited to what I wanted to do so I roughly copied one design. I only have this very old picture, this was the first attempt with 12mm steel rods. Another big fault in this design was the Z-plate, it was only two plates of 3mm aluminum, it bent super easily and was mounted poorly with barely fitting steel tube holders. I made the new one out of 10mm aluminum with two alu fuel pump mounts holding the router itself.
>>972343
More than I'd have liked on both counts.
It's got like $2200 dumped into it. Which is quite good for something with the work volume it has, and I was extremely pleased to discover that it would handle steel (I had only planned on aluminum at best). But not cheap by any means.
How long is hard to say. That one was made using and with parts from the previous machine, which I started working on years ago. Additionally, I'd often let it sit for months at a time because I didn't have any motivation to work on it and was busy with other stuff, or I'd be waiting forever for some part from China. It's been worked on on-and-off for like two years, but it could have been done in far less time if I'd had my shit together.
