After the success of the last cnc machine, I thought you'd be interested in this project log, start to finnish, full construction detail, all aluminium/steel, cnc router/light mill. I don't want to spend a load of time explaining about CNC, so if you need some background on computer-numeric-control then check out my wooden framed router. Of course, I expect a lot of questions, that's ok. I have the need for a stronger machine now, with the desire to do some full 3d wood carving, so the idea of an uber-accurate, non-flexing framed router was required. So far, I have planned out the x/y axis, the z-axis being drawn up as you read this. Take a look: The rails along the sides are mounted on their edges, and being 40mm hardened steel shafts sitting on 40mm A-frame aluminium, they will not flex one bit. The bearing blocks are also aluminium, with "re-circulating" ball bearings on their insides, to ensure smooth travel, even with 100+ Kilos across them! I just got them cut to length this morning, I have two lengths of 1500mm for each side (x-axis) and two of 850mm for the upper cross rails (y-axis). Being hardened steel, there were a wee-bit tricky to cut - my hacksaw never scratched them, so I took them over to my fathers workshop, where he cut them with his big dremel Amazing to watch, the sparks flew a full six feet behind him! I'm off to the scrap yard tomorrow, to pick up some more 1/2" aluminium plate, which they sell for £2 a kilo. I have some already for the gantry sides (the gantry being the whole y-axis assembly, which rides on the x-axis rails - you follow?!). You can see a couple lengths of phosphor-bronze in that shot, it will be used to make three anti-backlash nuts which ride on the leadscrews. The leadscrews which have not turned up yet!! I am using proper acme-thread 18mm, 4mm travel per turn leadscrews, driven by nice 24v 180w, 350 oz/in torque servo motors to drive each axis. You can see the massive 24v 25Amp PSU that I got from ebay to power these motors - they are 7.5 Amps each, and I have one for each axis. They will eventually have encoders attatched - these work like a mouse wheel in that as the motors turn, the little wheel in the encoder counts the steps, then feeds the info back to the little "geckodrive" motor controller which works out how far the leadscrew has travelled, thus ensuring that if I command the machine to move, say, 100.01mm, it will move 100.01mm. With the old stepper drive system, this feedback was not present, so when you commanded the motor to move, there was no guarantee they actually moved the exact distance. Anyhow, the encoders were bought on ebay again, but have not turned up yet from the USA. That's enough to get started, I'll get some scrap tomorrow, then start putting that base framework together.
Detn8r - you can! Look at the old router, built from MDF and ebay parts. Just needs a little research! Atomic - I am planning a 4' x 2' spindle travel, so the framework/slide assembly will total 1500mm x 900mm. Unfortunately this is the biggest feasable size that will comfortably fit into my shed! maybe next year....a new garage?!
Just a quick update, this CAD is taking sooo long! It's important though, when I need some measurements, I can just measure straight off the drawings - easy! As you can see, the z-axis (up/down moving section) is coming along nicely. Just gotta add the leadscrews and the belt drives to the motors, and I'm done here. I've been to the scrap yard twice this week, and no aluminium plate so far! Oh well, I'm off down there again tomorrow.
That's awesome. Can't wait to see how this comes out. A friend of mine designs those linear bearings and also ball screws. Curious about this project because I'd like to make a decent sized 3-axis mill for shaping hard foam to prototype composites and stuff.
I was mightily impressed with your wooden CNC machine kong, and if this is of the same quality then it should be absolutely stunning. Will you be able to post a video or something at the end of it cutting a logo into a piece of MDF or something? Like the ones linked in your other project log.
Will you be able to mill copper with it? If so you could get into some funky waterblock production looking good so far
I was very lucky to find those huge linear bearing rails, it will make the machine super-rigid. Therefore i have planned the rest of the machine to be stiff/heavy, so I can the headstock of my mini-lathe as a milling head to do some waterblocks. The router is good for thin metals, but heats up big lumps of copper/aluminium too fast to be of any use. We'll all just have to wait and see how well I can put the beast together! oh yeah, just got back from the scrap yard, and still no aluminium plate! I'll leave it until monday now, and perhaps have a go at making that z-axis in the meantime.
Have you thought of using coolant for when/if you mill copper and aluminum blocks? Makes a royal mess trust me but works great.
Ooh, coolant! One of the reasons for wanting this machine to be all aluminium is due to the fact that coolant is eating my MDF cabinet for my current router! If I could afford it, I would go for a system like the Trico Microdrop, or even a cold-air venturi. Yep, coolant, 'tis the juice of the devil!
This is one project I’ll be watching closely. I used to run a CNC lathe and engraver when I worked in a machine shop. Now your project has be wanting to build my own engraver. Yeah those are great systems for coolant just way to much $$$. I can see how the coolant would kill the MDF pretty quick. In the past I have used sanding sealer to waterproof MDF for outside cabniets. Should be able to find it at any home improvement store.
Cheers squirrelly, the MDF is actually painted, but that coolant works it's way through little cracks, screw holes, you name it! I have had soem fun cutting up some ali plate today, even with the proper metal-cutting blade on the bench saw, I can't go too fast. I cut the plates about 2mm per pass, stopping every now and then to let the metal cool down a bit. Too hot and it grabs the blade and stops the saw! Unfortunately, my neighbours don't appreciate circular saws this time of night, so I've packed up for now. I just had time to fit the blade into the mitre saw, so I can get the lengths cut to size tomorrow. It's then a lot of drillign and tapping, real boring stuff, but nice to see it coming to a start at last! Oh yeah, I have finnished the CAD design, everything bar the leadscrew bearing assemblies and lead-nuts. if anyone wants the DXF, let me know. It's 900k in size, so don't really want it clogging up my server!
I'd like to take a look at the CAD drawing for the mill there. What software do you use? Also, if you're going to be running any amount of coolant, it'd be good to build in a catch tray and possibly a recirculation system. What do you plan to use for drives and controllers? I'll have to talk to my friend at Thompson Ball Screw to see what he's got alying around
No problem, send me/post your email and I'll get them off. like I said, there are a couple of things missing - leadscrew bearings, nuts, and the leadscrew -> motor pulleys. It is designed in Turbocad standard, it is pretty basic, won't even let me designate materials for the renderings! Also bear in mind, it is my first attempt at 3d cad! The motors are 1700 RPM, I will be running them on a belt drive, with a 1:1.5 reduction, giving a speed of 4.5metres/min with the 4mm per turn leadscrew. They will be driven by gecko 320 servo drives. These run on a closed loop, with encoder feedback from HP agilent encoders, which have 256 CPR, quadrature output - sorry for the tech, some won't understand, CPR = Counts per rev, like the resolution of the mouse, the gecko's then use a 10x multiplyer on this output to give some ridiculously fine resolution of the leadscrew which I haven't got around to working out yet! It will be driven by a Linux PC running Linux EMC. OOh, Thomson ballscrews! Wish I could afford them! - A ballscrew is a type of threaded rod, which is ground into a hemispherical thread, the nuts are loaded with ball bearings which ride in this groove, creating a very smooth and accurate motion. Unfortunately I cannot afford them as they are about £100 minimum per axis! As for the tabletop, it's already planned - in my head! I intend to have a solid plate, then the T-slotted top will attatch to this, the coolant can then flow down the channels into some sort of collection system.
Thanks, Kong. If you can send it to marcello@lunagenlabs.com I'd appreciate it. I've been using Rhino3D for all my CAD work for about two and a half years now. It's very powerful for 'pure' modelling. Ie, no parametrics or animation. Anyway, back to the machine.. if the position is coming from the encoders in the drive, will you have to compensate somhow for backlash? I see you're using acme thread with anti-backlash nuts, but it seems unlikely that it'll be zero-backlash. Keep up the good work. And be carefull cutting that aluminum!
