On 4 December 2013 22:19, Leonardo Marsaglia <leonardomarsagli...@gmail.com> wrote:
> I managed to move the motor and do some tests but I can't get rid of the > oscillation from the PID. If I follow the classic way of tunning the > algorithm when I increment P I don't see any oscillation until I disturb > the motor. When that happens I get a really strong vibration and right > after that a following error. What I'm saying is I can't even start to test > my error to decrease it. Firstly I would set the folllowing error to about 16, two full turns. (if you get more than that, then you probably do want to stop). Then see if you can kill the oscillation with some D term. Tuning a bare motor is often hard even when it is a real servo, and an induction motor is likely to be worse. One problem you are likely to have, and which I had, was that my VFD didn't do anything at all unless the frequency demand was more than 5Hz. The solution to this might be some sort of inverse deadband, so that even a small velocity demand is at least 5Hz. (1V to the VFD or thereabouts) One way to do this, with a fair bit of tunability is with the "lincurve" HAL component. You could set that up to give no output with +/- 0.1V demand, 1V with >0.1V demand with X = -10, -0.1. -0.1, +0.1, +0.1 +10 and Y = -10, -1, 0, 0, 1, 10. -- atp If you can't fix it, you don't own it. http://www.ifixit.com/Manifesto ------------------------------------------------------------------------------ Sponsored by Intel(R) XDK Develop, test and display web and hybrid apps with a single code base. Download it for free now! http://pubads.g.doubleclick.net/gampad/clk?id=111408631&iu=/4140/ostg.clktrk _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
