Fun With a $200 CNC Machine

In reply to Keith Tanner :

It took an hour to cut, but most of that was me telling the machine there was way more stock to remove than I actually gave it. If I optimized, I could make a second in 25 minutes. 

And just to prove the fixture building was worth all the effort, I made a keychain for the Buick.

And with that, we're caught up to reality. I've figured out you really can get a CNC machine at this price, and figured out that it can even cut aluminum.

Now it's time to actually make stuff with it. I'll keep this thread updated as I continue to learn how to be a machinist! 

Mr_Asa said:

I wonder if an upcut bit would work.  Sounds like you might have downcut bits and they're digging in

Looking forward to the solution

I may have these backwards.  I can never remember them correctly. 

How thick of a piece of aluminum do you think it could chew through? 

In reply to Somebeach (Forum Supporter) :

Given enough time, the only limit is the work area. Theoretically you could make a brake caliper on this thing if you were willing to wait a week or so for it to run the program. Thicker materials just mean more passes with the tool. Case in point: To cut the bolt holes for that fixture, it removed something like 14mm of material.

In reply to Tom Suddard :

Thanks!

I finally had some free time to get back to this, so I bolted some aluminum down and hit "send." Wish me luck, because if all goes well I'll have a new aluminum part in about 14 hours.

Any guesses what it's making? 

License plate frame!

Since it can cut aluminum (slowly), am I right in assuming this could be used to machine delrin or some other stronger plastics?  Just thinking it could be really useful in making brackets for lines or other parts inside an engine compartment where a 3D printed part might not hold up.

Luggage tag?

Yes, it wouldn't have a problem with Delrin or similar. It would even cut those at a reasonable speed!

I woke up this morning and walked out to the garage to grab my mystery part... only to realize the ancient PC I've been using as a controller crashed and caused the job to fail. Darn. I'll move the router over near my "good" shop PC and take another shot at it.

Looks like an e30 door handle trim piece. 

Very cool!

Would love to see the gcode, does the machine generate any for you to see or is it all conversational and hides the gcode?

Post more screen shots of the setup/programming process. 

Tom, your next upgrade is obviously a Pi running Octoprint.

Mystery piece is a bracket for a...I want to think it's something electronical to talk to the driver. But it's tiny.

It would be really interesting to take a number of items and discuss the best way to make them and why. Would this piece be better made out of a 3D print? Or, if it's basically two dimensional, would a laser be the fastest and cleanest way? And when you're looking at one-offs, how does the CNC solution compare with good old fabrication? 

We tend to see everything as a reflection of our tools. You know, if all you have is a hammer. But that can end up with us forgetting about other options. I've seen a lot of designs in the Miata 3D printing group that would have been a lot faster and easier to make via fab, such as a mirror delete plate. You can make one from alloy scrap in 10 minutes with a drill and an angle grinder. Or you can draw one up in CAD and spend a couple of hours printing one. Obviously not all of us have all of these tools available to us, but an article or examination of this sort of thing would help people choose what kind of tool they want.

In reply to Keith Tanner :

I know I have 3d printed stuff just because I had a printer and wanted an excuse to use it. It doesn't happen as often now that I've had it a while but when I first got it the printer was the tool for every job.

Hah, so I already have a Pi running Octoprint for my 3D printer, but I prefer a full-size PC to run this machine. The offline controller it came with works, but I really like being able to visualize the origin onscreen, and adjust with the keyboard before running a job.

And yeah, you're completely right about choosing the tools for the job etc. etc.

Since nobody guessed it, I'll give the answer away: This is a door handle bracket for my 350Z. I prefer OEM door handles on a race car since I've found they make one-handed door opening easier than pull cables. The factory 350Z door handle bracket mounts to the inner door panel with four tabs that are at three different heights and spaced quite weirdly, so this part should mount the factory handle to factory holes in the door and still clear the door bars on the cage.

I'll explain the process in detail in an upcoming 350Z update, but I used the 3D printer to prototype and test for fit and function, and after three versions I'm confident enough in my design to make a real piece out of aluminum. I'm doing this all in CAD instead of the old-fashioned way because I need to make one for the passenger side, too, which should be as easy as copy/paste, and I can just hit the send button if anybody else wants one.

Toyman01 (Moderately Supportive Dude) said:

In reply to Keith Tanner :

I know I have 3d printed stuff just because I had a printer and wanted an excuse to use it. It doesn't happen as often now that I've had it a while but when I first got it the printer was the tool for every job.

Since my 3D printer access isn't very convenient, I find myself evaluating whether a part is best done that way or via the laser. The laser's faster for both design work and output - but of course it's limited in what it can do. It does make for some interesting design exercises, though, trying to figure out how to make the same thing in a couple of different ways.

Yeah, the 3D printer is the best toy ever. Here's how I mounted the lithium battery in the 350Z without replacing the extremely lightweight plastic battery box or having to make new cables. 

I have definitely gone down the rabbit-hole on hobby CNC.

This was my first machine - a Taig 2019-CR.  A nice little mill that I set up with steppers and Mach3.  It would cut aluminum slowly.  I ran flood coolant on it, which helps a lot with clearing chips.  Failure mode - the cutter would stall out, as would the steppers.  Stop, reset zeros, run again.

I briefly put together this router from take-off industrial parts (ignore the giant mess).  This one cut faster - because of the ball-screw axes.  It mostly did wood with it.  I tried aluminum and managed to bend the spindle on that little Craftsman router.  Failure mode: E36 M3 flying off the machine into my face.

Next I got this Gorton manual mill, with plans to add CNC capabilities.  Turned out it was a bit of a basket case, and I had to rebuild the whole thing.  I was putting the finishing touches on the cnc system when (foreshadowing) the next machine came up for sale.  Failure mode: never working in the first place.

Then I got this puppy.  Anyone who's following my TVR build will have seen this one in action.  Mid-80's Journeyman 325 built by Tree Machine (RIP).  5 hp of "Doesn't berkeley Around".  Failure mode: Breaks All Things.

So yeah, this kind of thing can escalate.

In reply to Keith Tanner :

So what you are saying is I need to buy a laser? 

Toyman01 (Moderately Supportive Dude) said:

In reply to Keith Tanner :

So what you are saying is I need to buy a laser? 

 

No, I'm saying I need to buy a CNC mill.  

HEY! EVERYBODY STOP SHOPPING FOR CNC MILLS UNTIL I BUY ONE!

I don't want any extra competition.

Oh, and here's a shot of the CNC's new home next to my fast shop computer. And by fast I mean it was $65 on eBay but I put a $17 SSD in because I'm not a monster. 

So far so good! I use a cheap webcam to keep an eye on it while I'm working.

A bit more on cnc cutting:

Biggest deal is holding the work-piece tight!  Any kind of wobble or flex will catch the cutter and ruin the cut.  On a bigger machine the cutter will break or the part will exit the machine!  In the least you'll get a ton of chatter that will make the cut look horrible.

See about getting a little milling vise for holding stock in your machine.  Tape works fine until it doesn't - I'm not a fan.  For flat stock, I like to start with a couple pilot holes and then use those to bolt or screw the work-piece to the table.  I do that with aluminum a lot.  It'll mill fine when screwed tightly to 3/4" plywood.

Second-biggest deal - cutters.  Cheap ones suck.  Good ones don't have to be expensive but make a huge difference.  For aluminum you want "high-helix" cutters with fewer flutes.  I like 2 or 3 flute cutters.  Steel (not that you'll be cutting steel on this little guy) wants more flutes and different cutting edge geometry.  I'm very fond of MariTool end mills.  Made in Chicago, not very expensive.

Spindle rpm and feed rate are the next big deal.  Every material and every tool are going to be happiest at a specific RPM and linear feed rate.  This establishes what machinists call "chip load".  Basically how much material gets scooped away with every tooth rotation.  There are formulas, if you're interested.  Broad view, you may be better off running a pretty high linear speed but not cutting very much depth with each pass.

This is a big deal when I run aluminum vs steel.  For a 1/2" cutting in AL, I might go 6000 rpm spindle speed and 90 ipm feed rate, with a depth of cut of 1/8".  For steel that might look more like 1500 rpm and 20 ipm.  And less feed rate.  Plastics are often pretty forgiving, but they tend to like slower spindle speed and heavier cuts.

For you, the lack of rigidity and power in the machine may be the limiting factors.  You may not have super good control of feeds and speeds either.  But hopefully this gives a bit more insight into the thought process and trade-offs.

TVR Scott (Forum Supporter) said:

For you, the lack of rigidity and power in the machine may be the limiting factors.  You may not have super good control of feeds and speeds either.  But hopefully this gives a bit more insight into the thought process and trade-offs.

Thanks  , do you need to  cool the cutting tip with air or coolant ?

In reply to californiamilleghia :

Depends on what you're cutting, how deep, and how fast.  You can cut without but you can go faster and remove more with coolant.