Hi all, I am a University student and have experience in workshops and 3D printing. A project that’s been on my mind is that I would like I would like to design and build a desktop CNC machine, which led me to this website. Ideally I would like to just be able to change the tool (eg from milling to laser to pen etc) so I can make it multifunctional. I got this idea having seen the CNC Piranha fx. I have my 3D printer (an iMakr Startt, which I salvaged parts to create my own design), connected to octoprint via a raspberry PI. I’d like to use the same PI or multiple to be able to just load gcode (not sure if the name is different outside 3D printing) for each type of tool without unplugging and plugging in loads of cables. I have some idea about the x,y,z part, but was hoping someone may be able to help me get started or point me in the right direction. There are many threads on this forum and I imagine I can’t be the first person to ask a similar question, but having never been here before I’d appreciate being directed to the right place. Thanks in advance!
People have asked it a lot of times before, but I don't know if it's ever really been accomplished to any kind of streamlined, satisfactory level. I know as of maybe 3 years ago, the standard answer was just "don't even bother trying". It's technically possible, but to do it well it'd require an understanding of machine tools, CNC and probably editing firmware and writing gcode macros that a lot of people don't generally bring to the table when it comes to basic extrusion-based hobby machines. If you really want to go for it, look at toolchangers on proper mills. They use tapers, which are easy to aim at when you're trying to swap out tooling. You could break the relevant conductors out into "slip rings" running around the taper, which carry power and data lines and interface with corresponding rings or possibly sprung contacts on the inside of the female taper on the z axis head of the machine. If you design your "interchangeable tools" such that the relevant part- 3D printing extruder nozzle, laser diode lens, milling spindle motor (small, brushless) has its cylindrical axis directly in line with the center of the taper bore, it becomes irrelevant what tool's in there in terms of orientation or x/y offset. The only thing you'd need to do is save a series of gcode z-zero offsets to use for each tool, which you would add to the header of your gcode file. Alternatively, a tapered- or partially tapered, like an R8- dovetail. Which would eliminate any issues with torque or motor size when it comes to the milling spindle, but you'd need to add x or y values to your standard offsets for each toolhead. And in either case, you'd need some kind of drawbar mechanism- I feel like I've seen a lot of people do magnets, but i don't think they have the repeatability to be of any real use. I also don't know how effective a dedicated 3D printing firmware might be when it comes to interchangeability. A generic gcode parser/acceleration planner/motion controller like grbl might be more use, and use your general purpose computer to do the specific stuff prior to sending the g-code. I don't know if this is a viable option (I'm pretty sure it could be), or if maybe something like using Klipper (probably the best 3d printing firmware, by many accounts) with some additional modules that you can call as necessary would be a better way to go. Definitely something to look into- it's not just a hardware project. If you're engineering, a fellow E-eng, physics or comp sci student might be able to help you with some of the more esoteric stuff, depending on your level of comfortability. It's a complex project, though, I think.
There are a couple of issues here you need to understand. First, if you intend to cut anything hard with it, a CNC by necessity becomes a brute force machine. It needs to be very stout and rigid and have the strength to move the bit firmly through the material. On the opposite end of the spectrum 3D printers and lasers have no resistance to their movement other than their own inertia. These systems need to move with speed and grace and making them any more stout than necessary increases weight which has to be overcome with additional power. So combining the two is like having a sumo wrestler performing ballet. It can be done but it's not pretty. Of course if you are not cutting anything hard you can go fairly light with the system and all this becomes a non-issue. The second issue has to do with Z-axis depth. 3D printers want it maximized and CNC machines want it kept to an absolute minimum, especially if you intend to cut anything hard. This is fairly easily resolved though with a vertically movable build plate. One such example can be found here. I don't believe that system has been tested as a CNC machine yet though so you may want to wait for confirmation of viability before you jump in.
That was it! I was struggling to remember the fundamental reason why we'd always just said not to bother. Yeah, the travel-rigidity-power triangle is brutal. I was assuming it'd be an overbuilt printer or a very light foam+balsa mill, but I'd forgotten how to put that into words.
Originally, when I first started planning my CNC router I was planning to do what you want to do and make an all in one. I realized that if I wanted to be able to print all types of materials, it would have to be enclosed, and my z-height, as a printer, would not be good for anything very tall, or if tall, it would most likely suck at aluminum cutting. Also, my CNC router is out in the garage were it is cold. The printer is in the climate controlled house where I can easily keep my eye on it during long prints. I am now building a separate belt driven laser cutter because my screw driven router would have been to slow for my liking as a laser cutter/engrave. The only downside to multiple machines is the extra space they take up. The upside is that it got me to throw away crap and obsolete tools which were taking up real estate. Also, I can 3D print stuff while using the router to cut things. Since 3D printing takes forever, this comes in real handy, not to mention the time it would take to swap over tools and make sure they are square to the machine bed. You can now buy 3D pinters for under $200. A pretty good one for under $300. With another $20 in hardware store parts used to make it more rigid, it will print beautifully. I wanted to build one out of Openbuilds stuff, but my daughter - who was 8 at the time - was already designing her own things in Sketchup that she wanted to print for games she invents. I wanted to encourage that and I knew that due to my lack of available time, it would take forever to get a build done. Therefore, I bought one of these less expensive printers and modified it. It prints as well as the Lulzbot Taz 6 I have access to (sometimes) at work after I modded it. Although the Lulzbot has a bigger work area. Hmmmm... now that I have a 3D printer, I can 3D print the parts for the Taz and using Openbuilds extrusions, build one.
Thank you all for your responses! I really appreciate it! Apologies for the delayed response I had some assignments due. It is a very interesting point regarding the righty aspect of milling requirements vs 3D printing. As mentioned previously my idea to combine all sorts of tooling came from seeing the Piranha cnc fx. I think my requirements for different tools would be: Milling: aluminium mostly (do you class that as soft? I would), composites (eg carbon fibre) Laser cutting: max 10mm acrylic, general wood engraving etc Other: anything useful! I actually have my own 3D printer already for about £100 (IMakr Startt). It required a lot of tweaking and modification to get a quality output (I enjoyed the process though). It was this process that got me into the milling /multifunctional idea. I am thinking it would be best to look at a milling machine which could have a laser cutter attachment or other end effector and remove the printing aspect? Any further advice of existing projects would be gratefully received!
Aluminum is hard milling by extrusion machine standards. Soft by actual milling standards, sure, but to use those standards you actually need box ways or linear rail. The ability to mill aluminum requires leadscrews or ballscrews, no belts. Much more heavily built than your typical wood router machine, unless you're taking very light cuts. Composites are brutal on tooling and need beefy air filtration and enclosures, but nothing specific regarding the machine itself that I recall off-hand. Laser cutting 10mm acrylic is gonna require a CO2 tube laser- acrylic actually linearly propagates that particular 10.6um IR wavelength and it cuts nicely. To attach that to a mill, removably, you're talking fiber optics, and I don't know off-hand how well IR works in fiber or how practical it is even if it does. Cutting 1-3mm acrylic (as long as it's black) and engraving wood, you can get away with a pop-on-pop-off 3W laser diode module. A 6W one might even get you up to 5mm or 1/4", I'm not sure. You're still talking z-passes and possibly some slightly melty edges though, it doesn't like violet as much as it likes IR, and you're still unlikely to be able to do much with white or clear acrylic. Then there's the speed factor- lasers like belts, especially tube ones, for etching speed. Leadscrews aren't going to go so well here. With diodes and slower feedrates, it might not be as big a deal. But then there's vibration. Mills generate a lot of it, and lasers really don't like it. So there'd need to be some separation there. At this point, you may as well just build a dedicated machine. Lasers and 3D printers can go much better together than anything can with mills/gantry routers. Fundamentally different machine design considerations. But even then, there are substantial differences between lasers and printers when it comes to format. The Piranha is really just a lightweight basic woodworking machine with a bit of 3D printing that I doubt anyone really uses. The laser module is just for engraving the wooden parts you make on it. It's more or less the same as an OX or C-Beam XL, as far as I can tell. It could probably do some light aluminum work, though the surface finishes might not be super fantastic. The 3D printing could do some simple PLA stuff, but decent 3D printers are really dedicated machines with heated build plates, heated enclosures, clean & dry filament storage, various sensors, etc. Not really ideal for a dusty, moisture-laden environment. Basically, they made a machine that doesn't quite do anything really well, but somewhat does three things ok. This is always going to be the compromise with multi-machines. What you need if you're serious about the requirements is a beefy mill with ideally ballscrews and linear ways, a "real" 3D printer (I'm not entirely convinced yet that with the proliferation of $500-800 units it's not just better to buy one, but you could probably do linear rail here for less than the commercial units), and a tube laser cutter (maybe a modded K40, depending on your cutting area requirements). All separate, possibly even in different rooms, with enclosures, exhaust and air filtration on the mill, maybe on the laser (if you're cutting MDF, PVC, leather, some other stuff), and maybe on the printer (maybe not filtration, but exhaust helps with the plastic smell/fumes). Maybe a compressor for the mill and laser too. In other words, heavily compromised at best, severely hamstrung at worst. Always better to do dedicated specialty machines.
