When I say small, I mean small — think of a bed size of three inches square, and you’ll have a reasonable idea of the scale I’m considering. My requirement is for very fine detail, working on steel, & I’m thinking very high spindle speed with very fine diamond-tip tooling. The really unusual part is the axis setup : one horizontal translation axis, one vertical translation axis, and a horizontal rotary axis. An advantage of this (I think) is that, supposing the rotation element to be based on a chuck, there’s never any ambiguity about where the center of the workpiece is, so long as it’s cylindrical at any rate.
Can anyone point me at good resources for this kind of thing? I welcome comments as well, even negative ones, since they do sometimes help me clarify my thinking.
What RPM and Horsepower will my spindle be? This is determined by tools you are using and the materials you are cutting. This tool will help you find answers to this question: CNC Speed and Feed Calculator and Formula
I would not try and reinvent the wheel, I assume you want a good tool not a hobby onto itself so I would start by reading up on DIY CNC mills here http://www.cnczone.com/.
From what I think you want to do with the mill I would probably modify a Grizzly G8689 to be a ball screw CNC and add a 4th axis and at the very least replace the spindle motor probably more. The primary concern I have with using any DIY build for extremely hard to machine materials is that rigidity and vibration become a really big issue so you will need to keep that in mind.
Six of one, half a dozen of the other, really. It’s intended to perform a highly specialized operation, so I don’t mind spending some time reinventing, as long as it’s not the square wheel.
So far as rigidity and vibration are concerned, I’m seriously thinking about getting a nice scrap of granite from one of those countertop places, or something like that, to use as a base. My hesitation about modifying an existing machine, beyond that very concern, is simply that I don’t need two linear axes in the horizontal plane, & I’d kind of rather put together parts that I think will suit, than have to take completely apart & butcher the result of someone else’s idea of how to perform a totally different set of operations. “Very little travel, superfine steps, & extreme rigidity” isn’t a featureset I’ve seen for desktop CNC, although I’m willing to believe it’s out there.
I actually saw an article from Make this morning talking about a 4th axis CNC. Bart Dring, the guy that invented Makerslide created a Delta CNC router with a horizontal rotational axis. http://makezine.com/2014/03/28/cnc-machining-contest-winner-viva-la-four-axis-delta-router/ I know this particle set up isn’t what you were looking for, but it goes into how he modified Repetier firmware to change the extruder travel to degrees rotation for the 4th axis. Pretty cool, might give you some ideas…
With a 1 mm ball-end mill, which is what I would expect to use for a rough-cut tool, the feed-&-speed calculator Frank pointed out gives me 0 HP & 10 000 RPM (assuming diamond tool working against hardened tool steel), which must be off-scale values. It also gives a feed of 16.75 mm/min, which is higher than I would have thought. Of course, if I’m cutting a spiral from the center to the edge of a 40 mm piece, with 100% overlap on the tool passes, that’s still a lot of millimeters!
Playing around with the calculator it seems to be doing the math right however the breaking torque of the endmill is so low it is outside of rounding error for its spindle motor torque calculations. Based on playing around with numbers it seems your ideal cutting RPM (75% of spindle max, this is to ensure that you are actually getting the RPMs you think you are and have sufficient torque in reserve to maintain the RPM under load.) should be around 34000 rpm for a spindle capable of running at around 46000 rpm. You can do the math yourself here is a starting point for that: http://en.wikipedia.org/wiki/Speeds_and_feeds
Nice find! that would be a great starting point! Also microstepping does not increase accuracy or repeatability it only improves ‘smoothness’ between steps.
Good point Frank. The micro-stepping increases resolution. If the stepper
motor has 200 steps per revolution and you use 1/8th micro steps then it
now requires 1600 steps per revolution. If 1 step equated to 1 inch, this
would now be 1/8th of an inch.
I was definitely planning on using reduction gearing, yes.
And as much in the way of existing parts as I could, although I’m not sure if that cross-slide table is exactly the right thing. It would allow me to have a rigidly mounted spindle, though, with only the vertical axis of motion. Something like that certainly seems like a good choice.
