leggazoid Wrote:
-------------------------------------------------------
> Would either of these mills work for a reprap
> machine?
Based on the review, the KX3 looks to me like a good basis for repStrapping, although IMHO, it's rather expensive overkill -- unless you wanted an off-the-shelf CNC machine, anyhow. -- Not that that'd be a bad thing at all!
You'd also have to be up for doing the wiring/switching between stock CNC operation and reprapping -- unless you were going to forgo the machining option for pure fused-filament fabbing. In which case, please let me know how much you want for the spindle.... ;-)
I don't know the CNC software that's bundled with this mill; the review *says* it's general purpose.... But that can mean different things to different people.
Before plunking down that kind of $, I'd check into all the details of that software (e.g. how does it handle the Mcodes of interest -- those we use for extruder control. How open/documented/hackable is it, generally.) Similarly, can you get enough info about the electronics to make them do what you want? Ask *before* you buy!
The KX3 apparently has ball-screws, which is a *nice* step up from ACME threads (on many mini/benchtop mills.) The backlash from ACME screws is a real problem for CNC machining. (The backlash can cause tool breakage, if not dealt with carefully, especially when doing a contour that has axes reversing direction during the cut -- e.g. cutting a circle.) It's not clear (at least to me) whether the KX1 has ball or ACME. I believe Nophead's setup uses ball-screws, at least for his horizontal axes (& maybe vertical also.)
One thing I didn't see either on the HF page or in the review is how far the axes move. I'll bet the KX3 has a machining volume significantly larger than Darwin's.
The envelope for the smaller machine (KX1) is probably closer to Darwin's nominal fab envelope. But don't take my guesses; get/read the full specs before buying!
Note that there are less-expensive CNC-ish options. CNCing a (less expensive) mini/benchtop mill can be done -- even easier/cheaper now that we (reprappers) have a low-cost (arduino-based) G-code interpreter, with source-code -- and the *developer* taking part in these forums! (And, with the Sanguino, more RAM and flash memory for adding new functionality.)
I'm doing something along these lines -- a repStrap that (I hope) can do both fused-filament fabbing and conventional (chip-making) machining, similar in spirit (if not the same hardware) as Forrest's Tommelise
And similarly, trying to do it without totally breaking the bank. I bought two axes worth (and no spindle), by buying a "compound table" for the X and Y motions:
[
new.shars.com]
(Shars p/n 202-2302)
Grizzly: [
grizzly.com]
and
Enco: [
www.use-enco.com]
have similar ones, as well as smaller models. And these are not the only vendors selling these. Palmgren makes a better-quality, more expensive one.
I have another linear slide from my junkbox to do the vertical axis.
The steppers from rrrf.org have more than enough torque to drive this. I may upgrade the ACME nuts with anti-backlash ones from [
www.dumpstercnc.com]
-- or try software backlash compensation. (Easier to do with a low-force process like extruding, vs. high-force cutting away chips.)
Such an approach is far more feasible with the source-code for the processor driving the motors available to read and to modify.
I'll probably use a dremel/rotozip, or maybe even a router as my tool spindle.
With a sturdy mechanism, I have some mass margin for the end tooling.
That's still a ways into the future:
I just built/tested my first two stepper controllers in the last couple days, and I haven't started modifying the compound table, either. Still, it's a start.
HTH,
Larry