SparkFun Forums

[estuardo]

Hey Everyone,
I'd like to share with you a project we've been working on for the past ten months, a CNC Mill that is more user-friendly than traditional mills and comes ready to run. We did all the firmware, software, PCB design, and cut all aluminum parts locally. One of the things we wanted to do with it was make PCBs, and we already ran a couple of single-sided arduinos with 10 mil isolation. We have a tutorial in the works for doing 2-sided PCBs. We want to publish it soon but managing the kickstarter has been a little overwhelming.
A couple of interesting tidbits. We're using GRBL for the motion control, and we designed our own arduino shield that plugs right into an Arduino UNO. We might integrate everything into a one-board solution but this has been working very well. We added firmware updates to add tool-length probing to make it simple to set up a job (the machine automatically zeroes right above your work piece), so for PCB milling there won't be any fiddling.
We designed our own BLDC controller with closed-loop speed control (PID), also using an arduino, and we can do both sensored/sensorless commutation. Our Spindle runs between 2K and 12K RPM.

Our software can't take a board file directly, so we used http://www.pcbgcode.org/ to generate the gcode from an Eagle board. We'll be publishing the settings we used as well. Hope this is interesting to some of you, and please feel to fire away with any questions!

Kickstarter link: https://www.kickstarter.com/projects/17 ... d-cnc-mill
Video link of the Nomad milling a PCB: https://www.youtube.com/watch?v=ZReH7zQ3cpc

[Ross Robotics]

OMG, I want one. Bookmarked in my Wish List.

[Mee_n_Mac]

Also mentioned at HaD. Read their comments. In particular I would mention belt drive vs screw drive. Both can be done right ... or wrong ... the former is easier to do wrong.

Having said that ... looks interesting ... as does a $500 PLA 3D printer. I can see uses for either and both. Damn !!!!
And having said all that ... http://www.sculpey.com/??

[estuardo]

Yeah, we got a nice bump from HaD, and there were something interesting comments on using belts vs leadscrews. We've ran the machine for hundreds of hours and haven't had any issues yet - we've tested cutting aluminum and brass as well as a slew of other materials (I'm one of the co-creators so I'll obviously be biased, so take this with the appropriate grain of salt). We have the belts tensioned pretty well, and we're using high-quality reinforced belts at that. We do like the belts for very fast movements if you want a good finish on dense 3D toolpaths in minimal time, and that extra speed is pretty handy for milling PCBs. If you plan on milling steel or want to remove A LOT of metal quickly then no, this isn't the machine for that.

[Mee_n_Mac]

As I said ....

Both can be done right ... or wrong ...

People acquainted with timing on cars, belts or chains, can appreciate the benefits and frailties of the choices. Neither is a clear winner, both have their application. As I'm sure do lead-screws ... and some anti-backlash engineering. In the end testable results are what matters. Perhaps some new metric could be derived that fairly compares the technologies w/in themselves as well as to others ?

Like (I'd bet) others here, "I" are torn btw a good CNC machine and a 3D printer but can afford just the one. These machines, of course, serve both overlapping and disparate needs. I'm still working on my own "ultimate" PCB machine. I'm not sure a PCB milling machine is (or isn't) the best approach.

Though it's well outside of your needs, perhaps you could educate us where/when a CNC machine is better than a 3D printer and visaversa.

[Ross Robotics]

I already have a 3D printer, The Replicator, and if I had a choice whether to own a cnc mill or a 3D printer, I would choose the 3D printer. Lot more possibilities. Anyone that would want to own a cnc mill is going to use brass or aluminum. If you want to make project cases and anything that doesn't need a metal, will choose a 3D printer.

I am glad that they decided to go with belt over screw drive. I believe it would allow for less backlash, of course it would depend on how well the belt was tensioned and engineered. Looking at their videos, it looks like they are precise enough for common hobbyist. I do understand that they won't show a video of the troubles they had when milling the examples. But I suspect it is as much work to dial in a 3D printer as it would a hobbyist level cnc mill.

[lyndon]

Anyone that would want to own a cnc mill is going to use brass or aluminum. If you want to make project cases and anything that doesn't need a metal, will choose a 3D printer.

I disagree. At this stage of the technology, I would prefer the CNC mill by far.

I already have a benchtop manual milling machine. I'm very familiar with what they can do and the benefit that CNC would add is repeatability, speed, and automation. I can build essentially anything I want, in any material: it's just a matter of how long it takes (yes, I do work with steel sometimes). I have consided building a 3D printer, but it would just be a "gadget." At my day job, we have a $XX,XXX Stratasys 3D "printer" (It's so well made, I hate to call it a printer) and while it makes prototyping pretty fast, I have yet to see it make anything that a CNC milling machine couldn't have done much faster, even with a few manual interventions needed. Not to mention the poor finish.

I think 3D printers now are really in the "interesting" stage. They are cheap, because they don't need the rigidity of a metalworking machine, they make prototyping simpler (for some value of simpler) and they're new. It's a fascinating technology, but I can't see buying one right now. Not when there are so many services popping up to do the work for you.
Great DIY project though.

[estuardo]

3D printers are pretty amazing, and the great advantage they have over CNC Mills is how easy they are to use. Even a child can go on tinkercad and in just a few minutes make a little action figure. Within minutes and after only a few button pushes that same child can be playing with his creation. What happens though, when a designer wants to make 100 of those action figures? It'll take 100 times as long and it will cost 100 times as much as making one, it doesn't scale. Here is where CNC has the advantage. You can machine a mold and make hundreds, if not thousands, of figures very quickly. CNC Mills on the other hand can be intimidating for the average user. Feedrates, depth of cuts, tool changes, plunge rates, etc, etc, etc sound a little overwhelming. We tried to address that with our machine, and tried to simplify the process. You'll be able to pick a material, the desired quality, the desired machining time, and we generate all those values for you. We've been cutting a lot of materials and fine tuning values, breaking bits and gouging parts so you won't have to
Accuracy is another area where 3D printers fall behind. Making a yoda head or very complicated vase isn't a problem, but making a housing that has to fit perfectly requires tuning. We designed and 3D printed a Gamby housing for a project, and it took a couple of prints to get it right. There's material shrinking and curling that has to be taken into account. Not so with CNC milling.
I realize this post is sounding a little anti-3D printing, but that's not the case, I love our Makerbot and uPrint and wouldn't trade them for anything. They can't be beat when it comes to making a quick prototype, proof of concept, or one of those "impossible shapes" that can only be 3D printed. I think in an ideal world a 3D printer and desktop CNC mill would be on everyone's shop.

[Ross Robotics]

well, I'm on my tablet so I won't go into detail. But, you can't mill a cavity like you can print it. There is a ton more things you can do with a 3D printer, in my opinion. 3D printers for hobbyists has been out longer (mass produced).

But, I do understand that everyone has there own idea.

Once I get back from vacation, I will go into more detail.

When talking about precision, my printer does a fine job, .2mm.. I think that for a hobbyist and for what I need, that's more than enough.

[dean449]

Since 3d printing is additive it's very versatile. The downside is higher cost and durability of the plastic. I once printed a small frame out of abs only to find it wasn't stiff enough for the application. I found a good alternative is cutting small sections of metal stock out of steel and welding it.

[Ross Robotics]

Yea I have run into that situation designing joints for a robot and is pretty much why I would want a cnc mill.