I've always had a hanker to make one of these, but without justification it would be a build without a purpose.
I've now found that a friend makes makes a wooden device for sale which has a lot of small holes in it and would be a good target for a CNC controlled drilling head.
Google shows a significant number of hits on this topic, with a significant number of design approaches which I'm sure I could use, but would be interested in finding if any one here has made such a device and what instructions were followed. The particular application for my friend is low accuracy but as I do make the odd PC board, a reasonable accuracy would be desirable.
You would learn a lot quite cheaply making something from MRMDF, using studding lead screws and roller drawer slides and a few bearings with a small/medium router as the cutting head. Model Engineers workshop ran a series a year or so ago along these lines IIRC. This should act as a good exercise and help convince you if you want to go up a notch or three to making a metal milling CNC or indeed converting a chaiwanese mill to CNC. Depending on your time/budget ratio, you can buy small machine ready to go and there is a chap who does conversions as a hobby from time to time and sells them through
formatting link
with everything up and running. Obviously you won't learn much that way but it will get you cutting metal quickly. I've a round-2-it project to convert my Myford VMC to CNC but not worth waiting for me to complete it for guidance :-)
Hi Bob The trouble is that there are clearly rather a large number of machines on the web that are MDF based, etc., and I was hoping I might find someone who had found one that did actually work. I'm sure they all do to an extent but some sort of personal recommendation is worth an awful lot.
If you want to find one the 'works' look for youtube links. FWIW I think they will nearly all 'work' but how well depends on the accuracy of construction. Getting three axes working smoothly and orthogonally should be reasonably easy for someone with engineering skills and replacement parts are cheaply made from more MDF if errors creep in. There are kits, ready made or DIY solutions out there. I've looked at a couple of forums but the problem there is that the photo links are broken where the authors have not continued to host them so just the text is fairly useless.
The moral seems to e that if you see a good write up in a forum, to archive it yourself just in case you want it later.
Andy - can you expand on that a bit please. I don't understand your first sentence; I now do have reason to build one and am seeking the sort of guidance you imply in your second and third sentences.
Searching on your the names in the 2nd paragraph identifies 3 sources of software, any one of which at this stage might be applicable once the applicable design is settled on, and one supplier of already built machines.
What I am trying to avoid is setting off following a well documented design on the internet, only to find that it is well documented, but actually a lousy design.
Sure - I've got a keyboard now that I didn't really have before (I was typing on a turkey-smeared something out of a cracker). Don't choose a design until you know what it's for.
I'd say there are broadly four or five popular designs you could go for. Most aren't even strictly milling machines - milling machines are difficult to build. It's more usual (even for working metal) to use routers rather than milling machines. These are faster and use different tool geometry, which reduces the side load generated, thus makes the slide and leadscrew arrangement easier to arrange.
Proxxon mini mill, with DIY steppers added. You can make this with screws, drilled adapter plates, soldering and shopping. An improved version replaces the plastic endplates with more rigid CNC milled aluminium ones. This is a high-speed small-diameter collet head using fixed shank diameter carbide burrs. Great for jewellery making or model making. Limited for working metals. They'll cut slowly, but are happiest in free-machining aluminium or brasses. Forget steel, or even aluminium or brass of the wrong alloy.
Larger mills, with stepper drives added. The favourite for this is the US Taig, as it's about the right size and is cheap. The downsides are lethally bad American electrics (don't go near their 240V version, it's just plain dangerous) and also the standard milling head is a bit on the slow side (which means milling cutters, which means high sideloads, which means a difficult drive system). They do also offer a higher speed spindle, but most people have gone to faster direct-drive router heads instead - the Kress from Wickes used to be a favourite, but now you have to get them from Germany. Oh, and the spindle bearings are crap, and the spindle motor is crap, but apart from that they do the job.
The generic Taiwanese model engineer's milling machine (Warco et al), retro-fitted with steppers. This is a world of pain and I've never seen one work right. The spindles are too slow, so you're stuck with 1950s tool geometries. These highlight the slop & backlash in the slideways and leadscrews. The torque needed is more than steppers can deliver, so skipped steps become a real problem. If you do try this, at least dowel the gibs beforehand and take adjustment seriously.
Bridgeports. Now you're dealing with serious kit and a budget that allows the steppers to be ditched in favour of high-torque servomotors. The slideways are well behaved, the gibs can be adjusted nicely. If you're lucky with your scrap hunting, you might even get to afford ballscrews.
Plywood or MDF constructed CNC routers. A lightweight 1/4" router head on a constructed carriage. Whether this works or not depends on your ability to design a sensible kinematic chain. It's a popular approach for making wooodcutting machines and they can be made cheap and large (usually with a belt rather than a leadscrew). OTOH, only half of those I've seen could machine a straight edge.
Electronics are generally bought-in, as stepper drives from someone like Gecko. Buy ones with opto-isolators - I don't know why they even do them without. You have the problem of how to connect them to the control computer - USB is best, parallel printer interface (used to provide a bunch of bitwise I/O) is OK, so long as your parallel interface actually works and at the speed it's claimed (USB to parallel doesn't).
I assume that you'll be running G code for control. This is universal for mills, although I don't think it's the best choice for laser cutting - certainly not for engraving. This then requires a G code interpreter running on a small computer (recycled ancient laptop is popular) which can be running Mach 3 (nasty bit of work, but very configurable for the drive parameters, and it's very widely used), could be proprietary for the mill & electronics, especially if you bought a recent Chinese turnkey mill. It could even be an Arduino running GRBL.
It's essential that your G code interpreter, and the stepper drives, can drive your mill to the right position. It fails to do this if it either gets the acceleration wrong, or if the steppers skip steps under load. A one-off machine will need a lot of tuning work on these aspects. It's helped by decent drive electronics, suited to the motors you're using.
The ease of practical use depends a lot on the software you're running here. Can you set the zeros easily? Can you drive the mill to a setup position easily, maybe even using a hardware joystick? Does it crash halfway through a long G code ? (This is unforgivable IMHO, as the whole point of G code is how little state it needs to maintain.)
You also need a generator for your G code, which itself depends on a "CAD file" to represent the dimensions of the object. Something like Cambam does both for you, being both a CAD editor and also a well- integrated G code generator for milling, with tool and material libraries to draw upon. You can also edit the drawings as DXFs (or SVGs) and then make G code from this using one of several command line filters. A process of great inflexibility and suck.
Good documentation will generally indicate whether it works or not.
If you've got a local Hackspace, there's probably a self-build CNC mill / 3D print / laser cutting group within that.
My point WAS that the DESIGN is less important than the accuracy of making one and the quality of teh bits from which it is made, and, finally the software and the electronics.
In the '80s (when I worked on the things) X ray lasers were made from a mystical American substance called "saran wrap". It confused UK physicists for a long time as to what the elemental composition of this stuff was, before it was realised it was just plain old UK cling film.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.