Positioning System Motion Requirements

Last month we reviewed positioning system performance terminology—repeatability, accuracy and resolution. The next logical step is a discussion of the motion requirements associated with positioning system performance.

There are five typical motion requirements performed by positioning systems. The most common positioning applications simply require repeatable, point-to-point motion, with the destination point being approached from only one direction. For example, typical high quality leadscrew or ballscrew driven positioning systems, without linear encoders, are unidirectionally repeatable to within 2-3 microns (as defined by ISO 230-2:1997(E)).

Motion applications requiring unidirectional accuracy are somewhat more difficult because the motion system has similar motion requirements to unidirectional repeatability, except that the positioning system must now also be used as a measurement device. Motion systems that exhibit excellent unidirectional repeatability can often be used in applications requiring unidirectional accuracy because the inherent positional inaccuracy can be compensated for by offsetting the target positions. Typical high quality screw driven motion systems, without linear encoders, are unidirectionally accurate to within 6-7 microns (as defined by ISO 230-2:1997(E)).

Positioning applications requiring repeatable point-to-point motion, with the destination approached from two directions, demand bi-directional repeatability—a common requirement for high throughput applications. Bi-directional repeatability is considerably more difficult to achieve than unidirectional repeatability because typical mechanical systems introduce two additional sources of error: backlash and hysteresis.

Backlash is observed when the commanded motion direction is reversed, but the positioning system does not react accordingly. In this situation, a rotary encoder mounted on the motor shaft would indicate that the system was moving in the opposite direction, but the actual motion of the system would lag behind the rotary encoder by several encoder counts. Some system components, such as gearheads, introduce backlash that is highly repeatable. Good motion controllers are designed to compensate for this type of backlash. Other sources of backlash, such as ballscrews or leadscrews, introduce backlash that is not as repeatable. Preloading the screw nut usually eliminates this type of backlash.

Hysteresis is observed when the same destination position has been commanded from opposite directions, but the actual position the motion system differ by an amount larger than the backlash. In this situation, a rotary encoder mounted on the motor shaft would indicate that exactly the same destination position had been reached when approached from opposite directions, but the actual destination positions would differ by an amount larger than the backlash alone. This additional positional error is the hysteresis, which is caused by unseen clearances and elastic deformations. A linear encoder can be used to compensate for backlash and hysteresis in point to point motion applications. Typical high quality screw driven positioning systems, without linear encoders, are bi-directionally repeatable to within 6-7 microns (as defined by ISO 230-2:1997(E)).

The relationship between bi-directional accuracy and bi-directional repeatability is similar to the relationship between unidirectional repeatability and unidirectional accuracy. Again, positioning systems with bi-directional accuracy requirements are being used as measuring devices. As in the unidirectional case, highly bi-directionally repeatable systems can have their inherent positional inaccuracies compensated for by offsetting the target positions. Linear encoders, when used with highly repeatable mechanical systems with small angular errors, can often achieve acceptable bi-directional accuracy results for single axis, point to point motion applications. Typical high quality ballscrew or leadscrew driven motion systems, without linear encoders, are bi-directionally accurate to within 10 microns (as defined by ISO 230-2:1997 (E)).

The most difficult motion applications are contouring applications, as positioning systems are required to be accurate while in motion. Typical high quality screw driven positioning systems performing contouring operations will typically exhibit accuracy errors 2-3 times larger than when they were used in point to point applications. Therefore it is important to use the highest quality components when specifying a mechanically driven positioning system for a contouring application. Correct motor and drive tuning is also critical.

It is important to understand the limitations of a linear encoder in contouring applications. First, linear encoders and controllers both have bandwidth limitations that prohibit positional errors from being fully corrected while the system is in motion. Second, linear encoders measure the positional error at the encoder read head, not at the carriage surface (Abbe error). They cannot compensate for any angular errors or structural irregularities experienced at the point of operations (usually the table surface) or cosine errors that occur due to a lack of parallelism of the encoder and the travel direction. Finally, most contouring applications involve some sort of material removal or marking. If the motion system makes positional errors while moving or stopping (overshoot), the linear encoder error corrections will occur after the error has been made and the material has been removed. The linear encoder cannot un-remove the lost material. For more detailed information regarding contouring performance evaluation, please refer to ASME B5.54-1992.

There are five major types of motion applications. It is clear that proper components must be selected and their limitations must be understood before specifying hardware requirements for a given application.