On Wed, 2008-01-09 at 03:23 +0000, Chris Morley wrote: > Hey Kirk what kind of linear encoders were u using? Do you know > what kind of signal they produce ? square edge (quad) or sin ?
My scales have quadrature output. > Sin wave signal encoders can be subdivided to produce more counts. > .0005 inches per count is too course for pid tuning. I agree. > Also I found > that when i stopped the slide would very slowly oscillate. > eventually I decided it was because The PID component which has a > deadband setting didn't work as I thought it would. If I remember > right it kills the P but not the I or P. so the I and D would > slowly more the motor though the backlash till it moved the slide > past the deadzone. I played with the deadband a little at 1e-06 and 1e-05, but apparently these numbers are way too small for my .0005" scales. > But my backlash was something like .005-.007 , > probably more because the encoder head was not bolted down very > well. This was just a test-I wanted to see something move! Same here. > When > I finish the machine The rotary encoders will be on the motor > (Thats the oem way), and I will fix the backlash, which I'm pretty > sure is a bearing. My Hardinge lathe had resolvers and tachometers on the ballscrew ends, which I replaced with with US Digital 10k pulses/rev. I am using the same arrangement on the Bridgeport. I see it as a compromise between the servo motor shaft and the real table position that the linear scale measures. After fussing with the linear scale without much success, I jury-rigged an encoder on the servo motor shaft to see if my tuning problems were from the linear scale, and it tuned well. (see fourth picture down, here: http://www.wallacecompany.com/machine_shop/bridgeport/ ) > I will also toy with dual encoders since the > mounts are there using the idea talked about - setting P with the > linear scale (at least .0001 inches per count I would like more) > and I and D with the rotary. At the very least i will try mapping > the back lash with the linear scale. > > Chris Morley I was hoping to do more with the linear scales, but at this point, it is much more important for me to get the mill into production. Though it looks like you will be finishing the work for me. Please keep us informed. I wonder how linear scales and measuring equipment are checked or calibrated? It would be nice to have an independent means to classify and measure ballscrew error and backlash. -- 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
