Hi, The most scientific approach to tuning Gecko's is to open the case up and find the set-point pins and apply an oscilloscope to view the set-point waveform.
I'm buggered if I'm going to undo all the wiring and scrape off the heatsink compound now just to get to the screws at the back.
So I'm like you, fiddling in the hope I get it right.
I belive that the gain control is like an adjustable spring, it gets the motor to position quickly but overshoots abit then undershoots back and forth until it settles on the correct position just like a car with no shock absorbers.
The damping control is exactly like a car shock absorber and controls the springyness so by adjusting this you should get it to a point where when the axis moves to a position you just get the smallest amount of overshoot then the system settles on the correct position straight away.
The trouble is that the two controls react together so adjusting one will need adjustment on the other.
The dithering you mentioned sounds like it is over-damped or too much gain. More probably the former, my guess is that the axis almost reaches the position but not quite so it is hunting back and forth until it settles. I may be wrong i'm still learning too.
It would be nice to here from Marris on how to get the best tuning without opening the case and what you would expect to see/hear if it is not right.
Currently I've only finished one axis so I'm still in the building stage. The axis seems to work ok when the motor is at a standstill it makes an urr,urr noise every 10-20 seconds or so. When I rapid and adjust the gain it doesn't seem to make a difference, mine is nearly full up. I have noticed if I get the damping wrong the whole mill makes a noise like I'm dragging a table across a wood floor.
What I would like to know is, do these adjustments only affect the drive on a few encoder counts, so rapiding up and down whilst adjusting wont make any difference.
What's your opinion?
Cheers  |