How much non-movemnet are you talking about?
How small of a step move are you talking?
What is the resolution for that axis?
Do you have an idler on the belt to keep the tension rather consistant?
I say this because we did some testing some time back ( actualy when were doing thread testing ). We were checking the mechnical part of the system and trying to compare physical movement to steps actualy sent from the pp as compared to what the stepper recieved. He could follow the
electrical signal path anywhere along the pulse path. BTW, the pulse counter was custom made for the company. So each electrical component was isolated and related to each mechanical part or combinations. It was interesting to find that we could actualy find the number of steps " seemingly lost " due to varying belt tensions. Now we were cutting hairs, but never the less, interesting.
In fact we found that not all steppers are created equal such that when microstepping some steppers did better than others.
So large non-movements as they relate to the system can be measured conventionaly, but as you try find what part of the system contributes what fraction of the non-movement it can become very time consuming and difficult. You need to be meticulous as hell when measuring. Dial indicators
are calibrated ( we measured small distances opticaly/ eliminates the non movement of the indicator or non liner movement).
It is very easy to have backlash just do to bearing mounting and bearing preload can add to backlash. The screw can be non linear, the preload on the
nut can add backlash, inaccuracies in the belt ( inclucing timing belts ), lets not forget the gearing, etc, etc. So there are a lot of things that come into play and cumulatively they add up.
So i will not guess at your problem.
FWIW,
RICH
Guest
