I finally got around to building an MPCNC in my shop, and got far enough this weekend to do a test cut. There’s still work to be done – obviously the wiring still needs to be cleaned up, and I have the parts to install an electronically-controlled air blast system and T-track fixturing + spoilboard.
Rails are 36 inches long in both X and Y, making for a ~25.5" x 25.5" working area. Rail are made of 304 stainless steel, 1" OD / 0.75" ID; they’re VERY rigid and I’m hoping to be able to push it pretty hard with aluminum and perhaps even steel.
Steppers are 84 oz-in motors from StepperOnline, wired in serial using a homemade board (motors plug into header pins, all the serial wiring is done onboard, and I have screw terminals for the output to the drivers).
The spindle is the ubiquitous DW660; the bit being used here is a 1/8", 2-flute short carbide bit from KBC Tools. I have some other bits from KBC as well, along with some bits I ordered from Ryan a while ago.
I’m running this off a Raspberry Pi 3B (or maybe a 3B+, can’t remember), using a Protoneer CNC Shield using bCNC as software.
Yeah, I went big on the rails since my plan from the outset was to do at least aluminum and hopefully steel – I figured extra rigidity would help (or at least wouldn’t hurt). Did some test cuts in wood tonight and it’s performing great!
Installed drag chains and cleaned up the stepper wiring.
Designed and printed (several iterations…) of some brackets to allow me to install some aluminum angle extrusions that let the drag chains sit above the table. This will let me have my work piece pass “through” the machine if it’s larger than my work area in one dimension.
Installed an air blast system.
Installed a relay system that allows software control of the spindle and air blast system.
Built a touch plate for Z-axis probing during tool changes
Installed T-track fixturing and spoilboards.
Removed the bottom part of the toolless collet locking system on the DW660.
Pretty happy with the progress so far. This video is pretty boring as it was just a stress test to see how far I could push deep, full-tool-engagement cuts on a facing rough pass and to test my compressor’s ability to keep up with the air blast system.
Thanks Ryan! That was doing a 10mm DOC, which wasn’t QUITE full flute engagement (end mill has 1/2" flutes). I had the CAM set to 3000mm/min (which seems to run fine for trochoidal milling paths), but I was getting bad chatter on the conventional milling passes (the climb pass was fine) and I would have had it dialled down to about 45% at the time of this video (=1350mm/min). I later pushed it up to about 55%, which was kinda-successful (it worked, but got a bit of chatter when it hit knots). Surface finish was (surprisingly) very good given that I didn’t intend for this to be even close to a finishing pass.
Gotta say, I’m VERY happy with this thing. I love the design, and I’m really glad I splurged on the crazy thick rails. It’s chewing through softwood quite happily, though I’m still fine-tuning my feeds & speeds, and I’ve done one test cut in aluminum to great result.
Not unless they make building studs out of balsa. I’ve just been testing with whatever random scrap 2x4s I have lying around, so spruce/pine/fir.
I was just out in the shop running some stress tests to see how high I could push it. I just did some trochoidal milling of a 50mm x 100mm x 12mm pocket in a random piece of 2x4. It was running at 3000mm/min, 12mm DOC, 40% stepover, 1500mm/min plunge. It ran fine, no chatter. I had actually programmed it at 4000mm/min but it turns out my controller capped feed rate at 3000mm/min, so I’ll have to go tweak the controller settings to push it further. I have video but my phone died so I can’t upload it right now.
I’ve just been testing with whatever random scrap 2x4s I have lying around, so spruce/pine/fir.