On 7 April 2010 20:04, Andy Pugh <[email protected]> wrote: > The spindles are linked in HAL, not on G-Code. This means that they > keep synch even when there is no program running.
This is what worked: using ddt to spot a change in the tooth count, then resetting the encoder and latching a new offset into the stepgen using a mux2 wired to itself as a floating-point version of the sample-and-hold function. (there is no way to reset a stepgen to zero, which would have been simpler). # Electronic gearing of stepgen position to encoder position net invratio invert.0.out => scale.0.gain => ddt.0.in # the ratio is linked in the postgui.hal to invert.0.in net spindlepos hm2_7i43.0.encoder.00.position => scale.0.in net gearpos scale.0.out => hm2_7i43.0.stepgen.03.position-cmd net aenable axis.3.amp-enable-out => hm2_7i43.0.stepgen.03.enable => hm2_7i43.0.gpio.044.out # Reset Logic. Zeroes the encoder and stores the current stepgen position as an offset. # This is to prevent the gear rotating on tooth-count changes (often quite interminably) net scale-change ddt.0.out => abs.0.in net abs-scale-change abs.0.out => comp.0.in1 setp comp.0.in0 0.01 # this is looking for changes in 1/teeth as the raw value is not available until postgui.hal has run net reset-spindles comp.0.out => hm2_7i43.0.encoder.00.reset => mux2.0.sel # this bit toggles high on a ratio change net sample-hold mux2.0.out => mux2.0.in0 => scale.0.offset # the mux is wired to itself as a floating point sample-and-hold net gear-spindle-position hm2_7i43.0.stepgen.03.position-fb => mux2.0.in1 # wire the stepgen position to the sample and hold, and thence to the scale -- atp ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
