On Tue, 2008-01-08 at 12:30 -0600, Jon Elson wrote: > Dave Engvall wrote: > > Hi, > > > > It is indeed possible to use linear scales with emc. With backlash > > they do not tune as well as encoders on the > > ballscrew. If the ballscrews are very tight the performance should be > > excellent. Recently there was a proposal by someone on the list > > (JMK?) to use encoders on the ballscrew/servo motor for velocity and > > rough positioning and then use the linear scale to drive final > > position. AFIK there is no software at present to do implement this > > but it is certainly a cute idea. > I believe you could probably mock this up with HAL components > (possibly might need to add a new math function along the way, > but the set of components is getting quite robust) to merge the > two encoder inputs. The math shouldn't be very complex. > > But, a caution: Knowing position via the glass scales doesn't > eliminate the backlash problem. The real problem is that the > position of the table is not constrained in both directions by > the servo. it is only constrained in one direction at a time, > and the motor cannot hop from one side of the backlash to the > other instantly, therefore cutting forces and inertia can flip > it from one side to the other faster than the motor can > compensate. This can lead to messed-up parts, broken tools and > general foul language around the shop. So, don't think the > glass scales are some kind of panacea that allows you to do > precision work with sloppy leadscrews. > > Jon
We have (safely) assumed so far, that WGB isn't using stepper motors. Because a fairly stiff mechanical system is necessary for a good finish, and a rotary encoder can easily put a ballscrew well within the positional uncertainty region, which affordable scales can't do, I find it hard to justify the effort to be able to mix feedback. It would be interesting to find the least invasive ways to eliminate the backlash. I don't know what it's called, but I have seen a system that utilizes matching sections of a leadscrew as caged rollers arranged around the circumference of the leadscrew, similar to a planetary gear. Two caged assemblies in tension would take up the backlash. I like this design because it looks like it could be made by most home shops at reasonable cost. I tend to like solutions that trade time and talent for cash. -- Kirk Wallace (California, USA http://www.wallacecompany.com/machine_shop/ Hardinge HNC lathe, Bridgeport mill conversion, doing XY now, Zubal lathe conversion pending) ------------------------------------------------------------------------- Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
