Gregg Eshelman wrote: > Can it handle using the linear encoders along with the rotary ones? Yes, but that uses up two encoder input channels per axis. For a 3-axis machine, that would be 6 encoder inputs, so you'd need two boards. But, they can be set up that way with a "daisy chain" cable on one parallel port. > I just like the idea of something directly sensing the position of the table > and quill instead of relying solely on the inferred position using the rotary > ones. 'Course it also has the single limit switches on each axis for hard > limits to travel. > On a small machine, with decent ballscrews, there really is no advantage to dual feedback. Large machines suffer expansion of the ballscrews, so the dual feedback can be helpful. Large, as in travels over, say, ten feet. If you have a lot of backlash, dual feedback will not really solve it, either. the table is still not fully constrained by a loose ballnut. > Nothing wrong with triple redundancy! :-) I've read the analysis of the > failures with the Therac 25 and its reliance on only software to ensure > nothing could go worng. Fortunately this machine won't be doing anything with > high energy radiation... > Still killed at least 3 people outright, and maybe several more over time! But, LinuxCNC has shown itself to be enormously reliable. I haven't had a real failure of the system since 1998!
Jon ------------------------------------------------------------------------------ Introducing Performance Central, a new site from SourceForge and AppDynamics. Performance Central is your source for news, insights, analysis and resources for efficient Application Performance Management. Visit us today! http://pubads.g.doubleclick.net/gampad/clk?id=48897511&iu=/4140/ostg.clktrk _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
