Picture of Build a Foamcore CNC Machine
Build a low-cost CNC machine in your kitchen, for under $120, using tools obtainable at your local art and hardware stores.

This project seeks to lower the ground floor of CNC machine construction. It utilizes foamcore - an easy to work with and cheap material - for its structural components. The most important advantage of this approach is how quickly new ideas can be implemented. Modifications can be built nearly as fast as they can be thought of because the material is so readily cut and glued. It's also very easy to repair - I shipped my first machine out to Maker Faire in San Mateo (from Boston) a few years ago and it (naturally) arrived completely crushed. Half an hour later, with the help of a hot glue gun and a few scrap pieces of foamcore, it was working as well as ever.

On the flip side, the constraints of foamcore as a material have led to mechanisms which are different than on most CNC machines. The drive system uses shafts which are directly driven by stepper motors and are stabilized using tensioned wire rope. Because the foamcore cannot be cut with high accuracy, the accuracy of the bearing system is independent of the accuracy of the foamcore.

A much more complete writeup of the project from a technical perspective can be found here:

Watch it in action!
This video shows an earlier version of the Foamcore CNC, to which some modifications have been made.

Foamcore CNC from Ilan Moyer on Vimeo.

How this Instructable is Organized
The Foamcore CNC is comprised of five categories of elements:
  1. The structure - including the top plate, back plate, bottom plate, and legs. This is built of foamcore and comprises the stationary parts of the machine.
  2. The table, which includes a set of struts and a counter-spring mechanism to help offset its weight, along with a drive mechanism. The table moves up and down.
  3. The XY shuttle and it's associated shafts, bearings, and drive mechanisms. This is the exciting part of the machine which is able to (fairly) precisely move a toolhead under computer control.
  4. The toolhead - i.e. the business end of the machine. I will show a few examples of both a pen and a ketchup dispenser, but in the spirit of how quickly ideas can be tried with foamcore, I'll leave the toolhead you design up to your imagination. In the simplest case you could just glue a pen to the shuttle.
  5. The controller. Unfortunately this would be a very long instructable indeed if I went into building a controller as well. For now I'll provide pointers to other projects, but one day in the not-too-distant future I will document a simplified version of the controller which was used in the video above.
[STEPS 1 - 2]: Materials and Tools.

[STEPS 3 - 8]: Laying out and cutting the structure.
[STEPS 9 - 11]: Laminating plastic to certain key areas of the structure.
[STEP 12]: Gluing the structure together.

[STEPS 13-14]: Laying out, cutting, and assembling the table.

[STEPS 15-21]: Fabrication of miscellaneous components, mounting the motors, and stabilizing the axes.

[STEP 22]: Building a rudamentary shuttle.

[STEP 23]: A brief discussion of the toolheads.

[STEP 24]: Relevant projects and work.

Most of the effort of creating this instructable went into the additional documentation which goes along with the step-by-step online format. These are PDF drawings attached to the relevant steps, and include information such as a visual BOM, layout drawings for each part, and lamination and assembly drawings. The complete set of drawings is attached to this intro step.

Why this Instructable?
I built the initial version of this machine several years ago as my first assignment for the MIT class "How to Make Something That Makes (Almost) Anything." Ever since, I've been meaning to publish this design so that others could experience the enjoyment which I felt as I cut and glued this contraption into existence. With the current empowering trend towards digital design and digital fabrication, the distance between our hands and what we create seems to be growing steadily. I've been immersed in a digital workflow for the past six years (both in school and professionally) and can say without a doubt that this project has brought me the most joy of anything which I've created in that time. This realization both puzzles and slightly troubles me, since it certainly isn't the prettiest or most durable thing which I've made.

The Epilog Challenge
I almost always need a deadline to push me to finish a project. In this instance, my motivation has been the third Epilog Challenge.

One thing which the experience of building the Foamcore CNC has taught me is the importance of the tools at hand. The fact that I had reliable access to a hot glue gun and an Olfa knife made it much easier to tackle this project. Similarly, having my own laser cutter at my home would help many of my projects come into existence which don't have the urgency necessary to overcome the activation energy of tracking down somebody else's laser and setting aside solid blocks of time (i.e. my personal projects).

Finally, and perhaps most exciting to me, I would like to be able to produce small volume runs of my CNC machine designs (not of the foamcore variety) upon which I could base a small cottage industry. One example is the variety of toolheads which I developed for the Fab-in-a-Box project, which are all made using laser-cut acrylic:

Remove these adsRemove these ads by Signing Up
surpcrepair3 years ago
just to give you another idea on this project I dont know if you use fiberglass at all but something that would take more time but still be very cheap and add a huge amount of strength to your foam core would be to go over all of your foam core either in epoxy resin you can spray it over all of the foam parts and test the strength. Epoxy is pretty strong on its own you can do some test to see how strong it would make the foam core. Or a longer process and make it even stronger would be to fiberglass and resin the foam it will make it very strong and you could easily use a dremel or something for your cnc cutter rather then using a marker or something light, you would probably have to put a metal base on it though to give it some weight to not move around.

I may try this project as i have time and If I do I will be testing how just the resin holds up to save time and energy because fiberglassing these parts would be annoying but worth themoney savings over buying all wood or metal pieces and having to cut them and all that.

Good job!
imoyer (author)  surpcrepair3 years ago
That's very interesting. I have experimented a bit with making composites by laminating aluminum sheet on to foamcore (which makes great shelves by the way) but I haven't yet experimented with epoxy. If you do decide to try it out please let me know how it works! And if you have any questions while building the machine just shoot me a line.

I have used cardboard boxes to make snake cages before, and they really hold up extremely well. It was a working alternative to wooden cages that couldn't hold up to the boa's learning that he could just flex his muscles and basically press on the box and eventually distort the window frame or some other area and get out. Reinforcing the cage then finding the snake had gotten out again, sometimes without any damage tot he cage at all, was a problem.. A Houdini snake makes for an interesting conversation piece for a new guest! (...found in the tub behind the shower curtain for instance by new girl friend...:-). Point being, the cardboard was easy to work with and design, then when you have it the way you want it, you use the resin to saturate the cardboard and the structural design inherent of the material becomes resilient to flexing and distortion and ultimately kept you from tripping over Houdini on the way to the bathroom in the middle of the night!

Oops, just wanted to add to the other ideas of structurally suitable alternatives to the foam takes a lot of resin to saturate, not fill the cardboard to the extent that it needs to be that it is still fairly costly in time and investment and the end result is not all that eye appealing and would have to be further worked into what you could just do without all the extra effort with the foam core, it just might be more rigid than foam and could serve as a component for the higher stress areas. I didn't want to give the impression that I just suddenly wanted to reminisce on an old memory about snakes in a thread about a CNC machine!

Your composite sounds like homemade "Dibond." I believe it's made from an aluminum, polystyrene sandwich. What thickness aluminum do you use? At $180 US for a 4' x 8' sheet, I think I might be inclined to try your method for some of my art projects.
imoyer (author)  spiderham3 years ago
That does look pretty similar to what I made. I can't remember the exact thicknesses that I used, but I believe it was around 1/32" on the side under compression and much thinner on the side under tension. I was building shelves so it was clear which side would be stretched and which would be compressed. The reason for the thicker material on the compression (for shelves, the top) side is because it will buckle if it's too thin. To laminate them I used paper-based double-stick tape but I think that a large sheet of transfer adhesive will work just as well. Good luck with your art projects!
Wondering why you would want to even mess with foam core due to the fact the price difference between that and MDF is negligible per inch/foot. Is it simply due to the ability to do without a table saw/etc..

I am only asking because I am using the outline to design a wood laser engraver for smaller areas. I am assuming the benefits of the MDF as far as a more rigid material per price would be more beneficial.

Great Instructable and excellent play by play!
This is great! I think I'm going to use this concept to make a laser cutter that I desperately need! So if I were to do that I wouldn't need the Z axis. In that case, all I wouldn't have to buy would be the third stepper motor and the 3rd steel axle am i correct? I will most likely buy everything at once so i dont want to mess up the order.
Yeah. but you might want it if you want to upgrade it later on. You would also need a different substance for the foam core, it would be too light and not accurate enough

What laser are you going to use? How powerful will it be?
Hmm most likely just a generic 445nm 1 watt blue led. And what material would you recommend? I'm trying to keep this as cheap as possible.
led? Do you mean laser?

Hmm, you should try plywood, it might work, just get the same thickness as the foamcore
ah yeah laser. sorry my bad. And thanks ill give it a shot
Make sure to tell me the results, I am thinking of making a 3d printer using this design
Alright no problem. Unfortunatly if i do decide to build it it wouldn't be until next year because I'm thinking of making this my senior project!!
Well then I guess I will have to experiment it myself:)

I was looking at another design, it used snap on segments, and you could use a router or the like to build it
It is also smaller, it could be put on a desktop
would this work for the foamcore CNC?:
If not, what would?
Yep, or any 3d cnc board will work
Could you make it a plastic extruder with a hot melt glue gun and make it a 3d printer?
I was thinking the same thing!

I found this repstrap that uses a hot glue gun instead of abs:
A great solution to uneven and possibly inaccurate cuts would be to cut this foam-core with a cnc machine. ^_^
dombeef3 years ago
I have to make this!
Have you tried a extruder to a 3d printing machine? Then it would be a very low cost 3d printer
imoyer (author)  dombeef3 years ago
I've extruded ketchup and pudding with the machine, but never anything serious like plastic. If the weight of the extruder were kept down, like how the Ultimaker guys do it with a "Bowden cable", it might work.
you could try a soldering iron?
I'm not sure there would be enough stability and durability in your model to successfully hold not only the weight of a plastic extruder+hot end, but maintain a high enough degree of accuracy. But as a food extruder, why not? You should make potato paste to extrude little figures and such for deep frying.
Dumchicken3 years ago
is it for foam core or is it mad of foam core?
Look at it, it is foam core
I guess that implies to your username.
REA Dumchicken3 years ago
you could probably cut foam core with it lol
mmaattanen3 years ago
Foamcore? Wow We use that in my sign shop a lot never thought to make a CNC machine with it. If you want something that can support more weight and have better durability and stability I recommend using Sintra boards, you can get them fairly easily, little pricier but it'll hold up a lot better than foamcore in the long run.

Sintra is just a bit tougher to cut but can be cut using a razor blade, that's how we go about it unless we're using a 10ft x 10ft board with 6mil then we use a power saw of some sort.
Do you use DiBond in your sign shop? You can easily fabricate many structures from it. Take out a slot from one skin on the table saw and then bend at the slot.
dombeef3 years ago
For the aluminum shaft could we use wooden dowels instead? That would bring the cost down extremely
imoyer (author)  dombeef3 years ago
I don't see why not as long as the dowels are straight and have a consistent diameter which fits inside the nylon bushings. Your idea would significantly reduce the cost of the machine - for the price of 1/2" wooden dowels it's certainly worth a try.
dombeef imoyer3 years ago
Thanks, when i build this, i will try that out
ah-el3 years ago
j3xt3r3 years ago
I've only quickly run through this and am liking what I see, although my lovely wife might object to me taking over the kitchen table with another project (we really need to move somewhere with a garage lol) but I have one teeny tiny issue with it and I'm sure yours wouldn't be alone in this and that I wouldn't be the only one thinking it either.

Where's the metric measurements?

I understand that imperial is still the incumbent in America but a far greater proportion of the planet has moved onto the metric system to the extent that my kids only know imperial as the system that Grandad talks about. Personally I find it to be an easier and more accurate system but that can be attributed to it being my first language so to speak.
Any CAD program will convert dimensions to and from both systems of measurment.
Is that a mac i see in the video... Any chance you could post info about the software you use? I'm having trouble finding a suitable osx application for my own projects. Nice work dude!
Look up reprap, try the software they use
ARJOON3 years ago
the axis is diffffffffffffffffffficult to make. nice tible
ljacox3 years ago
Awesome project! I enjoyed reading your clear instructables directions and watching your movie. Your project makes low cost CNC accessible to everyone. I'll have to try building my own foamcore CNC machine this weekend! ; )
Laura_j3 years ago
I just read an article on Kerdi-Board ... basically reinforced foam board. In the article they were using to build a tub surround using 2-in. thick for framing. Might be a bit expensive for a small project but if you were using leftovers from something bigger?
They've got a site. I've not been there.
Dumchicken3 years ago
mmmmmmmmmmmm motor on a stick
hey i love the cnc machine and i am going to biuld it i was wondering if insted of steper motors could you use the lego robotics motors and use the brain to control it ?
imoyer (author)  goosetheawsomescout3 years ago
I don't see any reason why not. I don't know anything about those motors, though, but if they spin and you can adapt them to the 1/2" shafts on the machine (or use different shafts) the machine should move. It requires very little torque. Please let me know how it goes if you do give those motors a try.
mycroftxxx3 years ago
I remember when i first saw pictures of this project, I was stunned at the audacity of it. Then my brain turned over and I saw a chance to come up with something really really silly. This project inspired me to consider doing an X/Y hot-wire foam cutter. I had seen one retrofitted from printer parts that an at-home engineer would use to cut out wing spars for his RC work. It would slide pieces of foam back and forth over a roller, while the wire slid back and forth on the X axis.

What would you say to trying to recreate this general project as a foamcore-based hot-wire foam cutter? You could submit it to as RepCRap, the self-REPlicating Cutting RAPid prototyper, capable of producing just as many of its own parts as a Mendel and much simpler, it would be the first open-source pun project.

Or, you know, you could just pretend you never read this.
imoyer (author)  mycroftxxx3 years ago
That's an awesome idea, and very funny! The stuff is super-easy to work with - you should go for it! If you do please let me know. I'll bet you could make it in a weekend :-)

Maybe two of these stages stacked on each other with a hot wire running between them? You could probably use the adhesive-backed delrin bearing strips to also hold together the larger plates that don't fit within the work envelope of the machine.
rblee3 years ago
If this thing works tolerably well, it's got to be the best technical instructable that I've seen. Something that does something relatively complicated, a 3-axis stage, using not only easy to obtain materials, but with easily gained skills and no roadblocks. $120 isn't much for a device like this either.

There are far too many instructables that casually throw in requirements like time on, and expertise with, a CNC lathe, or the ability to weld to exacting tolerances, which make them either difficult or completely impractical for many, not just because of the equipment, but because of the basic skills and practice required. Any chance of an Origami version? :)

Do you know how its resolution compares with a RepRap? Does the chassis flex much? I notice that you don't use much, if any, triangulation. Is the foam-cored board you used the stuff which is commonly called "Polyboard" in the graphics business?

As mycroftxxx pointed out, this could be Reprap for the masses, which would be ironic.

Well done. I'm very impressed.
imoyer (author)  rblee3 years ago

In a theoretical sense the resolution of this type of rotary-to-translational machine is limited by the stepper motor resolution and the diameter of the pulley (for belt-driven machines) or the capstan in the case of the foamcore cnc. I'm using 0.9deg stepper motors, which is half the pitch of the ones that reprap uses. My capstan is 0.5" in diameter, which couldn't be more than twice the diameter of their pulleys. So I'd estimate that their resolution is similar.

However my accuracy is probably much worst because the machine is not very rigid - as you point out. It does flex significantly even under the inertial load of the head moving around. Works fine for pen writing and laying down ketchup, but probably not for milling. You're right about the triangulation - that would definitely help significantly. The beauty of foamcore is that testing out some struts would be around 5 minutes of work.

I'm not familiar with polyboard but foamcore is a sandwich of paper-foam-paper.
Rob3113 years ago
People don't fully appreciate the potential of foam board. Its so simple, so light, and o so cheap. The only tool you need is a razor blade an hot glue gun.
imoyer (author)  Rob3113 years ago
Those are way cool! It's good to know that other foamcore CNC machines are out there. It really is a great material to work with. The cost of failure is nearly nothing, and you can build with it just about as fast as you can think.
dombeef3 years ago
Can you post a pdf of all of the plans? it would be easier to print
imoyer (author)  dombeef3 years ago
Sure thing - it's under the intro step.