Superficially, it is easy to think about converting a 3D printer into a CNC machine. After all, they both do essentially the same thing. They move a tool around in three dimensions. Reducing this to practice, however, is a problem. A CNC tool probably weighs more than a typical hotend. In addition, cutting into solid material generates a lot of torque.
[Thomas Sanladerer] knew all this, but wanted to try a conversion anyway. He had a few printers to pick from, and he chose a very sturdy MendelMax 3. He wasn’t sure he’d wind up with a practical machine, but he wanted to do it for the educational value, at least. The result, as you can see in the video below, exceeded his expectations.
The resulting device can work wood, acrylic, and even aluminum. This required a few changes to the printer, though. A lot of the MendelMax is 1 mm sheet metal. [Thomas] replaced a lot of parts with beefy 3D-printed parts and MDF. He also reduced the Y axis travel a bit to make the bed more stable. The software was essentially left unchanged but has some configuration changes that he explains.
The first incarnation used an inexpensive rotary tool as a spindle. However, the stress on the tool would cause it to fail after a bit of use. [Thomas] got a 48 V spindle motor and a matching power supply that works well.
The results are good, although not perfect. [Thomas] still wants to switch out some belts and strengthen the Z axis more. Still, it looks pretty useful. We aren’t sure all 3D printers would be as easy to stiffen as much as the MendelMax. Still, this build proves that it can be done, at least for some printers.
A more common approach is to use your 3D printer to build a CNC machine. Of course, converting the other way (that is, CNC to 3D printer) is almost trivial. Another option is to mount a LASER which doesn’t have torque problems. Of course, you might have safety issues, but that’s another hurdle to clear.
The reverse is also true. guys are making the Mostly Printed CNC into a 3d printer that is doing more with higher accuracy than even $3000 printers.
Sounds interesting, do you have a link? I was not able to find something with a bit of googleing…
MPCNC is here: http://www.vicious1.com/
I know the site of MPCNC, but I did not find no pages of MPCNCs converted into a superior 3d printer. I only found one multi head 3d printer made of a MPCNC, nice but quite special…
That seems like a tall claim, but I think it depends on what you’re comparing it to. There are pretty sad machines with steep price tags attached, such as an AirWolf3D, at least the pre-Axiom models.
I have said it before and Ill say it again, dont expect much from those “spindles” that they are using there. They are awful. They are just a DC motor with a ER chuck slapped on with set screws. Honestly, a dremel is way better than one of those. We had one at work that came on a little router and it was pretty much instantly tossed and replaced with one of the water cooled spindle/vfd combos that are on ebay. Those are actually pretty good. The VFD is kind of junk but for that price you cant complain much.
Yeah, I’ve borrowed a cnc with that motor from a friend who didn’t even use it once and motor axle broke after about 1m of cuttng in wood. Axle was 6mm, but the bit was only 3mm and didn’t break. Good that motors aren’t that expensive. I can replace it and will never touch it again.
I did a quick test of spindle motors I have in stock. That test confirms your statement with some numbers:
https://www.youtube.com/watch?v=NhS6TWdcCA0
The spindle test is part of my ongoing Hackaday CNC project:
https://hackaday.io/project/10299-self-replicating-cnc-for-194-or-more-countries
Nice, glad someone did some tests to show what crap they are.
That was so informative [Al Williams].
I am currently making and reciprocated (folded) arc delta to drill holes in PCBs and I wanted to test it’s limits with light milling of wood and hopefully Delrin (Acetal) and even more optimistically 2-3mm sheet aluminium.
This one video nailed all the issues at once.
It also give me some faith that another project might work (when I get to it) because I have bought 16mm round linear rails for that, only to discover that more experienced people recommend flat rail.
You want to read this thread: https://groups.google.com/forum/#!topic/firepick/9Z0hCwVwwOU
then you might want to re-evaluate using a delta ;)
I’m aware of that project and perhaps some expression that it hasn’t (so far) lived up to peoples expectations.
My project is being built independently and I don’t have any exceptions except that I learn something from the experience.
I want(ed) to build a “cost down” an excluding linear rails and ball screws was the main focus. Admittedly I have spent money on other expensive items like planetary reduction gears but I am still with Version 1.0 so I hope (if this succeeds) to reduce cost later.
I intend for this to have a welded (or alloy brazed base) in the end for rigidity if it is to be used for routing. There will also be improvements in the arm structure but for now this is how it is as I am stuck on the kinematics at the moment because I want the kinematics to be calculated in integer math on an 8-bitter.
but just to give you some idea – this is how it differs from more common delta configurations –
https://cdn.hackaday.io/images/425381483269385086.jpg
I´ll eat my hat If this works.
I’ve always thought this would be best done the other way round – replace a spindle with a print head. Mills need to be much more robust and rigid than a 3D printer to get good results, so a mill should make a better printer than a printer will make mill.
That was my first 3D printer. My DIY CNC router has ballscrews, and I got very anxious watching it move back and forward in the same spot for 7 hours, worrying it would wear out the precision screws.
I ended up getting a delta kit.
Pressed send too soon.
My plan was to be able to rough out a 3d printed part, then finish it with the router – accurate flat surfaces, correct diameter holes and bosses.
cutting into material generates torque?
perhaps he could harness the torque generated to power the machine?
No worse than saying that a nuclear reactor “generates” energy.
I am pleasantly surprised that nobody called out the fact the 3D printer is already a CNC machine. Means we are making progress here at the comments section.
CNC and manual machininst for a living, about to build a Folgertech i3 Prusa clone out of 80/20.
Yes- cutting forces of even small endmills, burrs, and anything with side cutting ability generates a LOT of torque.
This is why spindle quality and rigitity influences surface finishes and cutting accuracy so much.
When I do this for a living, I never use an endmill longer than I need to, and I always chuck up on it as far as I possibly can,
but just beyond where the cutting flutes end, so I don’t create a stress point for tools to snap.
Norbert’s video above is a really good direct view of what makes certain hobby spindles suck. My local hackerspace
has a homemade gantry CNC that uses one of those ER-16 silver cylinder spindles from china, yeah it runs 12,000 RPM,
but it sucks for vibration, has high runout, and weak flexible design with the shaft necked down like that right at the motor.
From being heavily involved in my area’s 3D printer group, who has several members that write for Make magazine,
I’ve seen very methodically that most 3D printers want as little mass as possible at the print head for speed.
CNC mills want the exact opposite- they need mass and rigidity as a damper to isolate vibration from cutting and
create a smoother surface finish. Having a beefier spindle bearing surface and using ER collets (or ultimate- spin balanced shrink fit tooling) gives best results in milling.
Now if someone here could come up with a CNC mill or printer we could build that used the ultimate movement
tech for accuracy- hydrostatic ways (look up Kern Pyramid Nano cnc machine), then we could have no more screws
to wear out, and hyper accuracy…
I am sure Joe wishes not that there was an edit button. And so do I.
. . . Every time, only on this site, does that formatting issue occur. I can’t get rid of it. And it only happens HERE. I’m not an idiot. I’m a maker & I have been my whole life. And I hate posting here because this always happens but I love Hackaday and have for over a decade :(
wracked my brain for a few days but it turns out I can run cnc gcode relatively easily. Think I am going to go ahead and turn my 3d printer into a cnc lols
https://ibb.co/e5f1Lo here is a screenshot of completed gcode. admitted its off the bed there but not on my printer. just used a g92 to set it up
Hello, I have an 3d creality 10s Printer that I want to turn into a CNC . Could someone help me to choose a good Extruder for cutting wood and plastics up to 4 mm ? Thank you in advance.
It’s a lot harder than just bolting on a spindle. If you watch the video, Thomas had to beef up the frame and replace some of the drive components, and even then he wasn’t entirely satisfied. And the creality frame is a lot less sturdy than the MendelMax that he was starting out with.
You might be better off starting from scratch, and use your 3D printer for making bits you need along the way.
I was just watching this from CNC Kitchen last night, coincidentally. https://www.youtube.com/watch?v=smIi4zrfatw And the various machines designed by Norbert [Homo Faciens] are also worth a look: https://www.youtube.com/watch?v=smIi4zrfatw and https://hackaday.io/HomoFaciens