3D Printering: The Combo Machines Cometh

There’s only so many ways to squeeze hot plastic out of a nozzle, and eventually witnessing the explosion of 3D printer designs over the past few years gets just a little repetitive. What then, is someone who dreams of a technological utopia, Star Trek replicators, and making a few bucks off a Kickstarter to do?

The answer, of course, is a combo machine. Where the Repraps, Makerbots, and the very high-end Stereolithography machines can only do additive manufacturing by laying down plastic or resin layer by layer, these combo machines can also remove material, be it plastic, wood, or metals such as brass or aluminum.

microfactoryThe first of these machines is called the Microfactory, and is designed around a simple concept: 3D printers can already position a hot end anywhere on its x, y, and z axes, so why not add a simple milling attachment?

In reality, the Microfactory is much more complex than that; it’s designed to be an all-in-one solution with a computer included in the chassis. Just add a monitor and a keyboard, and you have a fairly capable desktop mill and 3D printer. There are even options for a welded steel frame and a pump that sprays coolant all over the sealed build chamber, so these guys might just know what they’re doing.

fabtotumIf it’s features you want, FABtotum might be for you. The idea behind this machine isn’t just, ‘let’s add a spindle to a 3D printer’. No, FABtotum also includes a 4th axis and a 3D scanner, greatly expanding what a desktop machine shop can do. There’s an Indiegogo campaign running for the FABtotum that’s 92% funded with more than a month to go.

These two machines look like they’ll be successful, but it’s hard not to draw parallels between this and other combination machines on the market. Look around any machinist’s forum, and you’ll see cheap combination lathe and milling machines vilified as both terrible lathes and terrible milling machines. Likewise, even  the highly regarded Shopsmith, with a 60-year history, is generally seen as inferior to a separate table saw, drill press, and sander.

How successful these combo machines remains to be seen, but it’s likely they’ll only be used for extremely niche cases – the Manhattan apartment workshop, or where portability and size greatly outweigh capabilities.

30 thoughts on “3D Printering: The Combo Machines Cometh

  1. I actually own two real milling machines (a Arno universal manual with iso40 spindles) and a cnc bridgeport interact that is qc30 in the spindle. I was waiting for this to come around tuit happening, and it has. Ive always viewed makerbot etc as low resolution hobby devices and never bothered with the whole 3d printing fad so far.

    Now, I envision a day when a extruder head in a iso or nmbt or qc fitting plops right into the spindle of a actual real cnc milling machine as just another tool system, uses the existing accuracy and ability of the cnc mill, but adds a capability to additive print plastic and sla parts to the machine. Just another item in the carousel for bigger machines. Real hybrid machine capability.
    Someone roll that out on kickstart, I’ll be there with funding…

    1. What needs to funded? The solution is simple build an extruder head as though it would mounted on a traditional 3D printer. Attach the setup to a 3/8″ rod, insert into mill and bobs your uncle, you now have a 3D printing machine.

      There is no real reason why any “cnc” is not interchangeable with router/milling bits, a plasma cutter, or a 3D printing head.

      1. yes I can diy my own but lots of detail into how to mount to the tooling, supply electrical power to, or have a feed system, and it has to cope with accidentally being rotated at 10k rpm one day etc. What I mean is a production complete extruder head inside a casing ready to just bolt in without all that learning, that machinists and engineers would buy as yet another piece of tooling, not a diy project. Not many cnc operators make their own tooling from scratch, so why do we have the expectancy that they are going to make their own extruders from some article in make or on a website or the like. Maybe the issue is speed. I read that additive printing is slow, who would want to tie up their cnc machine to print one part every few hours. Perhaps thats the blocker to a viable product to market right now. Maybe it’d be a toolroom or jobshop get out of jail another string to the bow product only rather than intended for production work. The problem with speed is the machines are so expensive you have to keep them working as many of the hours of a day as possible on short cycle high profit work to make the money to pay back the massive loans they require to purchase. Not really hackaday territory, we’re more like jobshop engineers where someone wanders in with a piece and we have to figure out how to fix it with what tooling we have to hand.

        Also I can come up with a good reason why my cnc mill isnt interchangeable with a plasma cutter, the table. I’m certainly not cutting anything with spoil being blown onto the precision xy table ruining it. I have bolted an arm to the table to carry a plasma torch off to one side on a mesh table for some basic cnc plasma out the same machine, but I wasn’t happy about the dirty environment that created near the machine.
        A router is generally just a metal cutting mill with less precision/rigidity/repeatability and capability and high rpm spindle than traditionally found in a mill but spindle speeds have been going much higher on newer machines, so yes a cnc mill will act as a router, but you could make the distinction of the mill keeps the spindle still and moves the table, and the router keeps the workpiece still and moves the spindle around.

        1. Of course that also raises the issue of the table itself being unheated in a cnc mill, and being requried to be so as otherwise the heat would dimensionally change the table slightly with expansion locally. So a insulated from below bolt on table platform would be required also, not just an extruder head.

      2. @Davidb,
        yes there are several reasons any “real” cnc is not able to take a print head and start printing.
        1. Real cnc mills control 3 or 4 axis’, mostly, and by servo motors, and have no capability to control a couple stepper motors for the extruder(s). Nor is that a factory option on my 2 Haas cnc mills.

        2nd. the controller in a real cnc mill is set up to work with a certain number of axis, usually up to 4 axis, (on some machines 5 axis) and no more. So even if you install an extra axis board and servodrive, you still have to find a way to get the control software to control it. Yes I know Linuc cnc, and I know mach 3, but they are not installable on a Haas Mill so I am out of luck in my case and for most other commercial cnc mills too. You would also have to install encoders for the stepper motors, because that is how the software in the cnc mill knows it has reached positions.

        3. you are going to void your warranty, but this is hackaday, right?
        we break warranties all the time. Not if you buy a $100,000 mill and they will not come fix it when it is broken because you have modified it. You want it fixed when it breaks, so you can make parts and sell them and continue to make payments and live indoors. And eat.

        Also, are you willing to slow down your 3d printing to the speed of a cnc mill? Except for very expensive high-speed milling machines,
        mills are slower than 3d printers in rapid motion and in feedrate, because you are throwing around a lot of mass and that just goes slower than 3d printers which have very little mass on the printhead. Imagine taking double or quadruple the time on some 3d prints that already take hours to print.

  2. Great idea, as long as it’s primarily a mill that also happens to do extrusion printing, and not the other way around. A decent mill needs more rigidity, power, and accuracy than an extrusion printer; so a multipurpose machine should be designed with milling as the main goal, with no compromises. It will also be slower by necessity, but that doesn’t affect the final output of an extrusion print (or scan), only how long you have to wait for it; which is an acceptable compromise in my book.

    I have yet to make any purchase, but my plan has always been to get a mill – then add attachments for extrusion, laser, camera, etc., building them myself if needed. I won’t consider a dedicated extruder, or any multipurpose machine that’s only a half-arsed mill.

      1. Interpreting that as I want perfection is just silly. For instance the micRo v3, if it still existed, appeared to meet or exceed my personal expectations for a base mechanical platform.

  3. I agree wholeheartedly that the main focus should be on milling first, then add the 3d printing capabilities. it takes more to mill than to lay down material. a flimsy 3d printer frame is not going to be able to mill very well.

  4. im glad to finally see an article with more then one and a half paragraphs of useful TEXT that i can READ quickly and universaly.

    refreshing, informative, and accessable on any internet connected device, including text only screens.

    this article got to me at
    H.yper speeds bcuz it was
    T.ext, that was
    T.ransfered through a


  5. Many people say “Just add 3D printer bits to a mill”…
    It’s true that the 3D printer frames are not hefty enough for even half-decent milling.
    That’s because 3D printer frames are designed for high speed and high acceleration.

    You wouldn’t want a combo machine that in “milling mode” was too weak to cut through some plastics.
    Conversely, you wouldn’t want a combo machine that in “printing mode” was so slow that it would take multiple hours to print out even the smallest parts.

    The two types of machines have some topological similarities, but in practice demand very different optimisations.

    1. You can see that in the price of the 2 different systems.

      The Microfactory is $4.5k which shows they focus on the CNC milling side – you can also see the larger cutters.

      The FABtotum is around $1k, and appears to have a dinky little spindle driving tiny cutters.

        1. You could turn up the acceleration even more if you want to pull out more speed from your Ultimaker, the default is a bit conservative.

          But yes, a decent mill would make a slow printer, and a nice printer would make a lousy mill. It would be ‘smarter’ to combine a lasercutter and a 3D printer, as those both can use the high speed high accurate light movements.

  6. I’ve seen one 3D printer in action first hand, a Solidoodle 3. I wasn’t impressed with the way its build platform is only supported on one side the the leveling system relies on some rinky-dink springs and screws. The surface finish of objects has a repeating wave pattern, probably due to harmonics of the cogged belt drives and motor steps, plus the not so rigid build plate mounting allowing it to wibble around just enough to make its motion visible.

    A 3D printer should have a build platform firmly supported on at least two opposing sides and the construction should be precise enough that the build plate is held flat and level without needing adjustment. Should any adjustment be required it should be made with screws that hold firmly in both directions, with lock nuts – no depending on some small springs to hold it up against some cheap stove bolts pushing down.

    I can lay hands on a bunch of 4040 aluminum T-slot extrusion and can easily get precision ground and chromed rod and all the other bits to build a rigid 3D printer. I just need a project that absolutely needs a 3D printer.

    I’m working on updating an old CNC knee mill from 1990 with 21st century motors and controls. Could make an extruder head mount to fit into the #30 taper spindle. NMTB 30, also fits CAT 30, BT 30 and old stuff like Kearney & Trecker 2D collets, any #30 taper if I want to make a drawbar with threads to match.

    1. Actually, I HAVE a 3d printer, and those “rinky-dink springs and screws” are built in such a way there extremely reliable. We only have to tune ours about once every 3-4 months, and we use are’s a ton. Also, are build platform is only supported on one side, however we have never once had a problem with that. Even when using nearly the entire platform. I was at first worried about those seemingly weak parts, but now using it about a year, I see it really isn’t that weak at all.

  7. The answer, of course, is a combo machine.

    I dunno, they seem like answers in search of the proper question. Like Sporks. CNC requires so much more horsepower and rigidity and than 3d printing.

  8. The Fabtotum is great!

    It has a build area larger than the Replicator 2, yet it costs half as much.

    Let’s compare all the features of this model to the alternative of separate machines.

    Replicator 2: ~2,000$
    Matterform scanner: ~600$
    MyDIYCNC: ~750$

    3,350$ vs. 999$, plus saving space, as well as the fourth axis of the CNC mill.

    Which one do you prefer?

  9. Why not try to cram a laser and a waterjet in there too?

    Seriously, the different requirements for each (rigidity, strength, speed) mean that a combo can never really be great for either task. If you’re willing to accept that then great, but personally I’ll be sticking to separate machines. Even just sharing the electronics (stepper drivers, etc.) wouldn’t really be worth the trouble.

  10. The Fabtotum looks like it’ll be a nice device just for the combined 3D printing/scanning functionality. Add PCB milling functionality and it’ll be great. Forget about any more difficult milling; it’s not really built for that.

    For “proper” combo jobs, the solution is simple: get a 3D printer. Get a mill. Print a part, move the part to the mill, mill the part (possibly while printing another one). Not cheap, but a proper mill is never cheap.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.