This final project at MIT turned out quite nice. It is a CNC mill that cost under $100 to make. The tolerances are pretty tight as you can see in the pictures of the PCBs he has milled. He shows that he can even mill mild steel. It is a pretty brief writeup, but you can download build instructions and pcb files.
[via HacknMod]
Actually this machine is quite outdated. The latest one is Mantis9-1: http://makeyourbot.org/mantis9-1
Am I the only one who is always really sceptical about actual prices on projects like these? You see projects all the time claiming “Build XX for only $X!” but when you actually start looking at the parts, costs seems to rise real quick like. Often they rely on having random bits laying around the house which I think is misleading.
Not saying this is the case here but I have a feeling it is.
Regardless, great job, always nice to see random CNC machines popping up all over the place. It’s amazing how relatively good tolerances you can get with parts that look almost like random junk. Personally my own (quite a bit more expensive than anticipated) CNC is still a pile of parts after a full year of inactivity. I suck :(
This is pure awesome even if it costs $500.
I must admit I am a bit skeptical as well. It looks like he used ‘real’ linear bearings and acme lead screws. I would love to see more pics. Specifically the lead screw, nut, bearings and linkage to the stepper motors.
I built a CNC very similar to this about 2 years ago. The biggest issue (and the one that eventually forced a rebuild) was the that the epoxy I used kept cracking off after a small period of use of the thing. It looks like these guys used epoxy in exactly the same manner that I did so I wonder if that is going to be a problem for them…
Got to love the CNC builds! :D
It would be nice to see the structural parts on some we-manufacture-hardware-for-you service web site. If other people are thinking what I am thinking, then we want a quick path of least resistance to building one for ourselves. Then the machines-making-machines chain reaction can carry over to people who do not have the means for conventional mechanical manufacture.
Is it just me…. or is hackaday now just makezine.com regurgitated?
what are those IC’s on the bottom of the parallel board?
I enjoyed it.
Onanimous, thanks for the update.
It does leave out the cost of wood, adhesives, copper board, etching solution, and transistors.
Also, if you are a stickler for this, the fact it is controlled by parallel port instead of microcontroller.
That said, I’m curious about the BoM for the more recent version where he has halved the # of parts.
hehehe
“over and out”
Why on earth are they using epoxy to hold the bushings on instead of something like a conduit clamp? Ala http://www.ted-kyte.com/3D/Pictures/Conduit%20Clamp.jpg
That’s a nice little build. It would be great for a desktop area, especially if it had a dust-collector vacuum.
According to his small build summary it is based on
this project. That link has way more detail for replicating your own. I would love to build a desktop cnc, can anyone recommend a good source for hardened steel rods?
I think whenever you have a project where you claim it can be built for X dollars assume the following:
-All materials must be purchased (typical used price with shipping from ebay is acceptable)
-All required attachments (such as power adapters) must also be purchased
too often do I see projects labeled with a price and once you start realizing the guy who built it basically sourced parts from a huge bin he had on hand and didn’t include any of those in his cost estimate…
you can get hardend steel rods from old scanners. some of the older hps and the like used around 1/2″ rod for the guide.
my $45 one is better than this :/
I too am guilty of often not including the market price of parts I have on hand in my builds.
I think the proper way to say it is:
My X cost $Y, not including Z{1,2,3,4,5}
Oh how wonderful, a CNC mill with a accuracy of estimated +/- 1mm, which doesn´t have a feed strong enough to even cut aluminum, how nice.
With a bit of experience I can get the same results with my 50$ wood router.
If you spend a bit more money at ebay, you can get a used mill which was once even useful for something, and just needs a bit restoration.
This is what I call a real CNC mill hack
http://www.cnczone.com/forums/showthread.php?t=28668
@JustMe: really? You can mill 10 PCBs to 1mm accuracy at 30 inches per second? I’m impressed, although not as much as I would be if you actually had a project of your own to offer instead of cutting down the efforts of others.
@harrison
Older inkjet printers and scanners typically have one 8mm hardened steel rod per machine. You’ll need your own bearings for sure. The ones inside are either plastic or small brass ones, which don’t last long.
Any new printer that I have seen just uses a folded steel plate as a linear slide, which the plastic carriage (holds the ink) rides on top of. This is totally unideal for anything precision.
– amishx64
@JustMe
I want to see how well that freehand routed circuit board is coming along. You gave a link to a 3,000 lb milling machine in comparision to a desktop hobby router. Build a hobby cnc and use it to drill a couple hundred holes in a part, and be glad you didn’t have to do it by hand.
You need a pcb mill to make the motor driver pcb for the pcb mill? Talk about bootstraps!
The Mantis 9.1 is the latest version. But no parts list is published. I found a photo of the parts list at a FabLab site, but not enough resolution to read.
I find it weird all the links go to the old one project, instead of updated to the new project.
This machine will be usefull to make printed antennas. Can i get 1 from you?
And this thing actually feeds the motors from the PC’s parallel port?
Isn’t that a bit crazy?
I would expect that the parport is used for controlling the motors through another circuit not directly.
from what i can tell his breakout board drives the transistors for each phase of the stepper motor (linear drive) using a separate power supply, by doing it this way he saved a lot of money and created a much simpler circuit but sacrificed performance and lost many parallel port pins. the transistors (100ma?) look a little small for the steppers he is driving and could cheaply be upgraded to drive the motors (50ma – 2a?) with full current. i would also include current limit resistors for the parallel port pins and the steppers so that you dont have to find a specific voltage power supply (3v, 4v, 12v exc..?) and protection diodes for the transistors to save them from the inductive load.
Anyone have a reference to how he might be driving the steppers from EMC2? EMC2 sends step/direction (2 pins / stepper), not 4 phase / stepper control like his ‘brute force driver’ board insinuates.
I am really tempted to build this one but I am having big issues sourcing the raw materials (rod, bearings…) considering I live in Europe (France) and McMasters will not ship to my place. Anyone can think of any Euro-centric shops for this kind of stuff and maybe a BoM? Although I am willing to work on that last part myself if I can find good suppliers.
Thanks!