Home Made 5-Axis CNC Head Is A Project To Watch

[Reiner Schmidt] was tired of renting an expensive 5-axis CNC head for projects, so he decided to build his own. It’s still a work in progress, but he’s made remarkable progress so far. The project is called Bridge Boy, and it is designed to use a cheap DC rotary mill to cut soft materials like plastic, wood and the like. Most of it is 3D printed, and he has released the Autodesk 360 plans that would allow you to start building your own. His initial version uses an Arduino with stepper drivers, and is designed to fit onto the end of a 60mm arm of a standard 3-axis CNC,  so technically it’s a 3+2 axis CNC. With the appropriate software, it should be able to work as a full 5-axis machine, though, and it should be possible to integrate it with a CNC that has a 5-axis driver board without too much effort.

This is definitely a project to watch, as his work so far is very nicely designed and doesn’t look like it would be too difficult to print and build once the design is a bit further along.

[Via Reddit]

47 thoughts on “Home Made 5-Axis CNC Head Is A Project To Watch

  1. This is freakin awesome. Definitely what 3d printers ought to be used for. I’m grabbing those plans for sure. These kinds of projects make machining accessable to many people who otherwise wouldn’t get to experiment, and that’s fantastic.

        1. Also might make sense if you want cheap hardware that you can use for to learn yourself 5 axis toolpaths. Or you can scale it up for education. Imagine classroom with 10 cheap 5 axis CNC mills (but then you have to hope, that it can survive lots of crashes).

      1. Apparently none of the machine snobs paid attention to the part where it is made clear that there is no intention to cut metal and equally apparent is their own lack of skill with a milling machine, you have to select a cut that the machine is capable of, look at the accuracy that can be achieved with floppy old human hands mounted on equally floppy arms.

      1. Why bother casting it? If you’re making this to put on a 3 axis CNC mill that means you have a 3 axis CNC mill. I don’t see any parts that couldn’t be made on a 3 axis mill with some smart work piece clamping and tool paths. (You might need an indexing head to do the belt pulley.

  2. Regarding the component you are worried about shearing when printed in a certain orientation:

    I would split the part in two through the middle length wise, then glue them back together using a high performing glue like superglue. That way you don’t have the layers working against you.

    1. That doesn’t fix the problem because the next layer then becomes the weak point. The correct way to avoid weak spots caused by layer orientation is to orient the part in printing such that the stress is never across layers. The parts he has made here are basically ‘L’ shaped. There are three ways to orient this for printing:
      1) Vertically, so that as it is printing it forms an upright ‘L’. This will break at the corner.
      2) Laying on its “back”, that is, where the shorter leg of the ‘L’ is sticking up. This will break at the corner.
      3) Laying down flat, where both legs are flat on the print bed as it prints. The stresses do not pass through layer boundaries, and the part will not break at the corner.

  3. This is going to look really brilliant, right up until the tool bit touches the thing it’s supposed to machine.

    At that point, it will explode, because you’re trying to make a machine tool *out of plastic*. Not even just plastic, but 3d printed plastic.

    Unless you only ever want to machine styrofoam, it’s pointless.

    1. I’ve seen people successfully machine things with very weak contraptions. Is this yet another “I know it all” outburst or do you have enough experience and knowledge to actually know this to be true? IOW can you show that the forces when cutting anything stronger than styrofoam will make this fail? Given your use of the word explode I think not.

      The problem that is obvious is that of precision: rigidity or rather the lack of it will make the cutting tool wander, smart software can compensate for this slightly (essentially calculating deflections and trying to counteract them).

      But not everything need extreme precision.

    2. This is so predictable. Every time someone comes up with something resembling a mill or lathe FOR CUTTING SOFT WOOD AND PLASTIC, using any material other than steel or cast iron, weighing less than their father’s first automobile, all of the wannabe machinists on the Internet declare it useless. And then the people who build these things, cluelessly go on building them and making use of them anyway.

      1. Not all of us are wannabes. Getting paid as a machinist does technically make me a professional, doesn’t it?

        That being said, this project seems like a lot of fun. Its all about what kind of material you intend to cut, how many you want to make, and what tolerances you intend to keep. It gets really expensive really fast at the limits.

  4. “With the appropriate software…”

    That could be the killer. Anyone know of any affordable 5-axis CAM software? I’m guessing that if the OP has been renting a head then it’s not an issue in his use case.

  5. “let me just turn this off” . Bro, just turning the power off like that on a RPI is the best way to corrupt your uSD card. It’s not like Arduino where you can just turn the power off. Think of it like a computer, you just cant unplug it.

    1. New RPIs aren’t nearly as bad as the old ones for corrupting on power loss. I still try to turn them off, but I’ve also just pulled the plug plenty of times and they’ve never ate the SD card like the old ones would do every single time.

      1. Most likely they cut out the CPU before the SD card then. SD cards are extremely prone to corrupting when they have a power loss during a write. And, as you corrupt the internal administration, the SD card becomes useless afterwards for anything reliable.
        So, if you cut out the CPU and keep the SD card powered for a bit longer, it can finish it’s write action and corruption be much less likely.

  6. It occurs to me to wonder if FDM printing could take advantage of additional axes, to print objects with challenging geometries, without the need for supporting structures. Of course, the nozzle always has to point approximately down, but this can be managed by mounting the hot end on an X-Z carriage as is typical, and putting more degrees of freedom on the build plate, some geometries could be accomodated.

    Or, how about a 3D printing “lathe”? This would work by first printing a core on a base oriented in the typical way, then turning the base 90 degrees so that core is horizontal. The base is then rotated for the remainder of the print, building outward from the core.

  7. What a cool project! Good luck with it and keep us posted. The belt cover idea is a good one as swarf will go everywhere. I’d also be interested to see how hot the spindle motor gets even with the fan on top. Hopefully there is enough moving air to keep the nylon from becoming too flexible. Great project!

  8. If it is anything like the one that came with my CNC3020, the accuracy of those cheap spindles is greatly improved by preloading the bearings. Ideally with some shim stock, but just a wave washer works well enough for me.

    As for the durability of the 3d printed parts, maybe once it is finished the most critical parts can be cut out of solid plastic.

  9. I think the fact that the guy modeled the pieces with fusion 360, went ahead and got the parts printed and assembled it into a working machine is just fantastic. The amount of time and thought that went into this is substantial. After messing around with Fusion 360 myself and making a few simple models, I can tell you Reiner is on a level I can only dream to be on someday…. Anybody making negative comments about quality content such as this are just trolling and they themselves know it.

  10. Great project. I have been into CNC as a hobby for many years and would like to try using 5 axis. The barrier to me is the cost of 5 axis cam software. Is there 5 axis CAM software that is cheep or open source that a hobbyist could use?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.