I posted a while back making some small parts in 17-4ph stainless on a $500 desktop CNC.
The removal rate was slow but acceptable. However, I had some mic6 offcuts laying around for another project, so I used the machine to upgrade itself.
20mm mic6 side plates, nearly doubled working z height.
1.5kw vfd spindle replaced the old 300w dc motor.
In aluminum, my material removal rate is now 25 times what it was stock, and no more chatter. Should speed up my projects quite a bit, and any other upgrades seem moot at this point. Absolutely incredible to hear this thing sing with just a new spindle and some scrap aluminum.
That’s awesome. What brand is the desktop CNC?
This is an anolex, their offerings have changed a bit since I purchased.
After seeing this thing rip with the new side plates, I think these simple extrusion and side plate gantry designs are 10000% percent the way to go for home gamers. All the newfangled kickstarters with the bells and whistles don’t seem so easily upgradeable where it matters compared to this: rigidity and spindle power.
This gets you in the game for cheap, and you use the machine to upgrade its own side panels when you get tired of long cycle times and needle-thin chips. Slap on a cheap Chinese router spindle and you are flying through material (for a hobbyist) for under $1k.
Fascinating. What’s the difficulty like with using a CNC mill like this versus using a printer and slicer? I’ve been eying a desktop mill for making the parts I can’t make on a printer, but I heard the software side of things isn’t nearly as developed as it is for 3d printing.
It depends on your project needs and cad capabilities.
Project needs: most gun stuff you would make with a CNC (rails, optic cuts, slides, fgc9 type trunnions etc) do not need super tight tolerances, and as a hobbyist you don’t care too much about cycle times. You do need proper measuring equipment: at least a small machinist square, and a dial indicator to make sure your machine is trammed and square.
CAD capabilities: if you use fusion capably already, then it will be pretty easy. Simplify your designs to not use fillets, chamfers, or crazy 3D surfaces, or your cycle times will shoot through the roof and you’ll have to change and relocate a lot of tools (think releveling your bed and rehoming your toolhead several times in the middle of a print). Sometimes I have to make a copy and change the model a bit to get fusion to apply toolpaths only in a specific area. Once your tool path is created, you need a post processor to write the gcode for your flavor of machine, GRBL in my case. Rather than transfer it directly to the machine, you use a gcode sender like Universal Gcode Sender to send the gcode to the machine.
The biggest thing to figure out is feeds and speeds, but that’s very analogous to dialing in your extrusion and layer height. You just take a sample block of material you want to cut, and change 3 variables until you have good chips and no (or acceptable) chatter for both facing and side cuts. It is a bit of an optimization thing that takes some experience to figure out on your own, but look at what other people are doing with similar machines and dial it back a bit to start.
Once you got that down, This Old Tony on YouTube has a great video about squaring stock, and then you’re off to the races.
Oh, and don’t skimp on your vise and how you fix that vise to the bed. Most likely a small toolmakers vise is going to be the best bang for your buck.
Machining is similar to 3D printing in that you can get up and cutting with relatively little knowledge, but it’s different in that the more complex your projects, the more tips and tricks you will need to make it work. but most of our projects can be massively accelerated and improved with the most basic of CNC operations. Hell, it’s stupid easy to make a SS on this thing, especially now.
That’s super informative! Thanks so much for explaining in such depth.
For sure, I want to push this community to utilize cheap home machining because it pushes the envelope wayyy further in making gun control laws a futile effort, and opens up so many cool project opportunities. Especially with this machine design, it is very defensible as it is open source and easy to diy from the ground up as long as your extrusions are cut square and you can procure the parts.
I have quite a few projects lined up to accomplish some big goals, the slide I’m making now is just a fun one to get started.
Actually, you know what? Here’s how to make a super safety.
You need to wiggle the geometry in cad a little bit to be able to use 1/8” end mills. I think there are already some models floating around with the changes, but basically the slot for the lever needs to be widened a bit, and the pocket needs to be changed to radiuses that can fit the 1/8” tool. The bulge on the end of the lever needs to be thickened to match. Add a 1/2” or 12mm square boss on the non slot end so you can hold it in a vise. Orient the square boss so that if you lay the whole thing on its side, the safety detent grooves and trigger interfaces have no overhangs. Maybe I’ll just post it.
Use a 12mm square stock of preferably 17-4ph as you can heat treat it. Stand it upright in the vise. Mill out the diameter of the of the safety with a rigid tool, like 6mm or 1/4” end mill. Your tool stick out should be a little more than the length of the SS for clearance.
With the same orientation in the vise, use a 1/8” bull nose end mill (bull nose is less likely to chip a tooth) to mill out the lever pocket.
Flip it on its side in the vise so that where you want the detent and trigger interface geometry is straight up. Use a 1/16” ball nose endmill with 3D adaptive clearing to mill those pockets.
Saw cut off the square on the end and boom, you have yourself a homemade SS in heat treatable steel. If you want to heat treat it, use the H900 process, it’s stupid simple if you have a kiln or oven with the right temps.