On Tue, Aug 10, 2010 at 07:53:48PM -0700, Neil Baylis wrote: > Chris, > > would you care to hazard a guess at how much phase shift you see from the > 7i48 input to the velocity input at the servo amp?
I tried to measure the delay in this setup. I scoped the real tachometer (which is now hooked to nothing else) vs the dac output, and issued a move 1.5" positive then 1.5" negative. I doubled the acceleration to 60 inch/sec^2 to try to get a better slope to measure the delay. I can't identify any significant delay with this setup. Upper trace is the real tach, lower trace is the dac output. http://timeguy.com/cradek-files/emc/real-vs-generated-tach.jpg Here is another shot with 1ms/div horizontal and 0.1v/div vertical. This is the beginning of a move. You can clearly see each servo cycle. I have about 1 div vertical separation and at the beginning of each servo cycle it looks to me like the two signals match (of course as the cycle goes on, the generated value gets stale and falls behind but that is unavoidable). http://timeguy.com/cradek-files/emc/real-vs-generated-tach-upclose.jpg I think the noise is a feature of my very old digital capture scope; I'm not sure how much of it is real. Anyway I'm still inclined to call this configuration 100% successful. Chris ------------------------------------------------------------------------------ This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
