Good morning all,
I've got a rather frustrating problem with a pair of Dugong DG2S-0080 servo drives driving a pair of Kelling 34-170 brushed servos. The issue is that my positioning error is never centering on final position. This system is on a lathe. The Z axis has a 1000 count line driver quad encoder, and the X axis has a 2500 count line driver quad encoder. Both have been confirmed to be good (actually, the 2500 count encoder is brand new). There is no appreciable noise in the system which I've confirmed with a scope. I have tuned and tuned and re-tuned again these drives. The only way I can get the position error to stay somewhat close to 0 (+/- 10 counts) is to turn the Kp and Kd gains up to the point where things start getting "buzzy". The critical thing to note here is that NO MATTER WHAT THE Kp and Kd tuning, Ki inputs (integral, error correcting over time) does absolutely nothing to the curves. This doesn't seem possible to me since it should just be math driving the transistors in the drives, but if I didn't know better I'd say the integral term in my drives was broken. The configuration software for the Dugong drives allows inputs of Ki from 0 to 65500. The response graph at Ki=0 and Ki=65500 look nearly identical with neither of them eventually landing on 0. I have these same drives in my mill with some older DC servos but identical encoders and I have never had this problem.
Any thoughts would be very much appreciated.
SG