CNC Router experience

Fixing and making things, what tools to get and what skills to learn, ...
Post Reply
AxelHeyst
Posts: 2118
Joined: Thu Jan 09, 2020 4:55 pm
Contact:

CNC Router experience

Post by AxelHeyst »

I found a thread on 3d printing, but I'm interested in CNC material removal from wood specifically, at a desktop scale. Does anyone have a CNC router or mill that they use for wood? Have you used it for 3d shapes, not just 2d? Did you buy premade, kit, or complete DIY? Any other experience to share?

Context: I ran out of ideas for my family for Christmas, so I told them I'd make a custom, heirloom grade wooden Settlers of Catan board for the family by next Christmas. I'm interested in getting (building, assembling) a CNC router (mill?) to do the job, and if I can make something cool enough, sell copies and other designs for fun and to recoup the costs of the machine.

(CNC machine skills would complement my existing 3d modeling skills, I think.)

User avatar
Sclass
Posts: 2791
Joined: Tue Jul 10, 2012 5:15 pm
Location: Orange County, CA

Re: CNC Router experience

Post by Sclass »

I built this back in 2000. I occasionally cut plastic and metal parts. The only time I cut wood is when I crash the tool into my sacrificial base.

The mill was originally built to manufacture electronics parts in 2000. It turned out that the throughput was too slow and I used other techniques. In the meantime I used the machine for fixturing my assembly line. You have your thinking right. These tools are great if you want to replicate multiple copies of something. Otherwise the setup is such a chore it really doesn’t make much sense to invest all the time.

This was a big deal when I had customers ask for additional copies of custom HW I made for them. I billed the same job but I just churned out more parts. In one situation I had a client’s business boom and he needed to massively parallelize his assembly line. Being able to easily replicate precision copies of the same design was a hit. It was like printing money.

For fun the machine really hasn’t seen much use. I replicated some unobtainable Mercedes parts and sold them for awhile.

The build.

I started with a Taig Tools mill. I liked it because it was small and very stiff for the money. I used DC servos and pulse encoders from a discarded gene arrayer. The encoders were mated to standard CAM decoder amplifiers. All that is driven by a parallel port on a DOS PC with hobbyist gcode interpreter SW.

Fast forward 23 years. People use Linux CNC and it is way better than the CAM SW and HW I used. My motors and mill are still unbeatable by most hobby level machines people make. Stiffness and acceleration from DC drive is everything. Premade stuff has varying levels of rigidity and varying precision.

YMMV. I am cutting small precision parts the size of half dollars. The machine requirements will be different for big wood cutting machines.

At some point I’ll get on the Linux CNC train. If I make more stuff. Right now my SW tool chain is very awkward using free gcode translators and Freecad. I have the tool chain for circuit boards working pretty smoothly.

Some photos and a Mercedes Key head project from a few years back.

Image

Image

Image

Image

Image

Image

Image

Circuit boards

Image

Image

Image

User avatar
Sclass
Posts: 2791
Joined: Tue Jul 10, 2012 5:15 pm
Location: Orange County, CA

Re: CNC Router experience

Post by Sclass »

Forgot to say, all this stuff is 2D. When I first built the machine the 3D modeling software had a steep learning curve and it was expensive. I designed everything with 2D cad and cut out simple shapes. When I finally got a Master Cam pirate copy I found it very cumbersome to design 3D contours and machine setup and toolpath generation was longer.

Then the actual 3D cutting took forever. I simulated a few parts and they took hours. It was then I realized why real 3D CNC systems have 240V spindle motors and giant servo axis drives. You need to cut fast if you want to get anything done in your lifetime.

I use the machine to do utilitarian tasks and I don’t really care about 3D contoured surfaces. If I need a part with 3D functionality I’ll make multiple parts as 2D plates and glue/screw/weld them together to make a 3D part. It’s faster and easier.

Here is an example of 2D CNC assembled into 3D. Not my parts but a good example of what I mean. It is much faster than hogging out a solid block with 3D geometry. It’s simpler design and cutting. It wastes less material. It needs simpler cutting tools. It’s faster.

Image

I found this image of a small PCB being made last year. I posted this stuff up in the diy electronics thread. This is easy to do using Flatcam software and using Gerber files for geometry.

Image

jacob
Site Admin
Posts: 15907
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 77
Contact:

Re: CNC Router experience

Post by jacob »

I get the impression that commercial desktop sized wood CNC cutters are often used for sign making, carving out names, etc. At least they're advertised that way.

You could easily make a Settlers of Catan board with a scroll saw. Scroll sawing also offers some 3D options. Due to the nature of the saw blade/process, the 3D needs to be---there's probably a term for it---extruded all the way along the given axis. A chess set is a common 3D project. Otherwise, 3D shapes can be assembled by gluing together slabs of plywood which are then sanded down. Main benefits is less waste and less setup. Main downside is that everything becomes piecework; there's no automation.

BTW, commercially you'll be competing with 3D printers. A quick google shows a fancy Settlers board on etsy.

AxelHeyst
Posts: 2118
Joined: Thu Jan 09, 2020 4:55 pm
Contact:

Re: CNC Router experience

Post by AxelHeyst »

The kind of board I'd make would be targeted at people who wouldn't buy a Catan board made out of plastic.... think "Jumanji" materiality except Catan. But you're right, a lot of people make custom Catan boards, not exactly a blue ocean market. It's more of a well, if I'm making one for my family, I might as well see if anyone else wants to buy copies, which I can make at low reproduction cost once the initial design and troubleshooting design work is done. Since rough tolerances is acceptable (even, perhaps, desirable, to reduce the "this thing was obviously made by a robot" vibe), the sort of stiffness required for precision parts making might not be necessary for my application.

The scroll saw is interesting - I didn't know you could make stuff as 3d as you obviously can. Those chess sets are impressive. I will look more closely at that. Marginal cost of reproduction is high, though, I almost certainly would just one and done the board for my family.

Sclass - very cool, thank you for the download!

The level of sculpting I have in mind is similar in kind to this. The basic 3d modeling is trivial for me, although converting the sculpt file to CAM formats and efficiently is the bulk of the learning curve I think. e.g. a bulk material removal pass followed by high-res pass so the machine doesn't take 12 hours to cut a single hex, as Sclass mentioned.

I am tempted to get one of these $200 CNC router kits as a learner machine, and use the experience to decide if I want to upgrade or do a serious DIY Opensource hardware rig build. I've started keeping an eye out on eBay.

User avatar
Sclass
Posts: 2791
Joined: Tue Jul 10, 2012 5:15 pm
Location: Orange County, CA

Re: CNC Router experience

Post by Sclass »

I think at $200 it’s worth a try. Wood is more forgiving than metal. In your video it isn’t clear if it is sped up. I think it is. The tool is very long and slender. I would expect a lot of deflection if the feed rate is high.

Download some other people’s gcode. Hogging with cover passes or not, it is going to be slow going. Especially on a flimsy machine. By design you must move slowly. Simulate their code in real-time on a web based simulator to get a feel for what you’re up against. If the machine is too flimsy you can use the machine to make reinforcements. It’s an interesting concept a lot of cheap CNCs have. They ship just good enough to make improvements to themselves.

I was looking at doing CNC wood block engraving years ago. I had this interest in reproducing images like these in wood for block printing. I lost interest. It was easy to get vector data but translating it into tool paths that looked good on wood was too difficult given the SW available at the time.

Image

AxelHeyst
Posts: 2118
Joined: Thu Jan 09, 2020 4:55 pm
Contact:

Re: CNC Router experience

Post by AxelHeyst »

Sclass wrote:
Sun Feb 05, 2023 2:30 pm
Download some other people’s gcode. ... Simulate their code in real-time on a web based simulator to get a feel for what you’re up against. If the machine is too flimsy you can use the machine to make reinforcements. It’s an interesting concept a lot of cheap CNCs have. They ship just good enough to make improvements to themselves.
Fantastic advice, thank you.

I also thought of an end of life strategy for the unit:

1. Get CNC Router
2. Learn workflow
3. Use it to make whatever I want to make
4. Be done with it.
5. List it for sale at whatever I paid for it locally, targeting homeschool nerd family with kid(s) interested in it. List it for cost and offer to provide support / lessons on the workflow, if they're interested.
6. Secondary yield: Potential relationship with local homeschool family/ies that might do other cool projects in the future.

Post Reply