|
|
Home |
|
|
Poor static stiffness is where the motor position or velocity are disturbed to an unacceptable level by low-frequency or DC disturbances. These disturbances may come from the load, such as in a machine tool where a vertical axis holds a heavy work piece against gravity. They may also come from the transmission where friction disturbs shaft torque as the motor comes to rest.
Poor dynamic stiffness is a related problem. The problem of poor dynamic stiffness is observed with high-frequency disturbances. However, many causes of and cures for poor static stiffness are the same as those for poor dynamic stiffness. In Kollmorgen's ServoExpert, the assumption is that problems of dynamic stiffness occur at frequencies near, equal to, or above the velocity loop bandwidth. Problems of static stiffness are assummed to occur well below the velocity loop bandwidth. The distinction is somewhat arbitrary. If you are not sure, you should investigate both problems carefully.
The primary causes of poor static stiffness are low total (motor and load) inertia and low servo gains. Feed-forward gains have no impact on static stiffness. The primary servo gain that effects low-frequency dynamic stiffness is the velocity loop proportional gain. For example, the gain KV on the ServoStar S, CD, and SC series drives when configured as PI or PDFF/PI+ controllers has the dominant impact on dynamic stiffness. Actually, this gain scales the entire control output and effects both dynamic and static stiffness.
The velocity integral gain (equivalent to the "P" gain in a PID position controller) also strongly effect static stiffness as will the position loop proportional gain (roughly equivalent to the "I" gain in PID position controllers).
If you want to verify that dynamic stiffness is at the root of the problem you are trying to solve, lower the velocity-loop proportional gain a small amount (say, 25%) and see if the problem worsens. Repeat with velocity loop integral gain and position loop gains. If all these gains effect the problem, it helps to verify that poor static stiffness is the problem. If only velocity loop proportional gain has an effect, the problem is more likely to be from poor dynamic stiffness.
| Rate your level of concern about this problem: | |