Now that v2 will have more backlash compensation would it be possible for LL to put together a demo that would illustrate the the precision & repeatability of the machine? Spec sheet motor ratings of .00001 angstrom nano-stepping are seldomn helpful, when the reality is many of these desktop machines and mini mills easily will slop .005" or more with .002-.003" being considered respectable.

Have you guys made any complex and tiny parts that could shed some light on the limits of this machine?

Yeah, where are the nano-machines built using the micRo?

Actually we have measured the tolerance several times, and I have posted several pics/vids and pictures somewhere deep *shudder* in the micRo thread. But coincidentally just yesterday I decided to mic the "trophy" again from the German Troll Debacle and the parallelism tolerance over 5" in 1/2" thick aluminum was <.001", probably in the range of .0002"-.0003", but my calps can only display .0005" increments. The RA (square tolerance) was not measurably out (no measurable error). This is from a machine that was hardly trammed and calibrated, and in which all axes did not have lash compensators or shaft collars.

With V2 I expect the tolerances to be very high, on the order of <.0003" or better. While it may seem fantastic, you may note that we use a 12tpi lead screw with 8x microstepping (1600s/r yields .00005" step resolution, ), a precision tool plate base, very long threads for the blok bores (more screw contact means higher tolerance), and we have greatly improved tolerances by going to our own linear bearing system.

Do you map the screws?

Acme screws have more lead error than .0003 per 5"

QUOTE (arizonavideo @ Sep 17 2009, 02:58 AM)

Do you map the screws?

AV: You're just chumming for my crude humor here aintcha?

Well, first you'd have to qualify what you mean by error. If the error is in the pitch or thread spacing, then the error will repeat accurately which is not the same as lash. If the error is is the screw diameter and enough lash compensation is used it does not effect the outcome. This is why I mentioned the long threaded bores we use, as they tend to "absorb" these small differences in tolerance. (The smaller the "nut", the greater the error.)

"Lead Error

The deviation from nominal lead resulting from manufacturing variations. Standard lead error tolerances for Ball Screws are .010 in. per foot. Lead errors for Acme, Hi - leadŽ® and TorqsplineŽ® series leadscrews are .009 in. per foot with precision tolerances of .006 and .003 available in selected sizes. All tolerances apply plus and minus, however, thread rolled product errors are generally the same throughout the entire screw and can be easily compensated for with appropriate profiling software."

Our screws are CNC machined out of billet, not rolled. The lead error over a foot is < .001". We detect the error by the loss of black oxide on the screws.

Cheers~

bc

You are using ground Acme screws?

If so you should brag about that fact a lot more because they should cost a lot more and be much more accurate than a rolled screw.

I have never seen any of the "normal" leadscrew makers selling ground Acme would you mind telling us who makes them or even a reseller?

Not ground, machined on a HMC (I'm guessing with a stock feeder, continuous process), but I need to verify the exact method.

I'm sorry, I though you said ground.

I have not seen machined Acme before but I did find that Nook sells them and they have a lead error of .002 per foot which is a nice screw.