Compensator
Introduction
A compensator can be used to add a delta offset on the position of one axis based on the position of one or two axis. An example use case would compensating the Z (vertical) axis for X-Y surface irregularities.
In order for your compensator to work you will need to externally collect data on the Z axis deltas for each of the data points in your table. These measurements could be taken manually or if your machine has a touch probe you could write a program to build your compensation table for you.
The compensator works by adding an offset to the position of an axis based on positions stored in a table. This offset is applied right before data is sent do the drive and does not change command position. How data will be retrieved from the table is defined by a min, max, and delta for each input axis.
Determining required configuration parameters
To configure a compensator, first define the X-Y area to be compensated (Xmin to Xmax, Ymin to Ymax). Then define the spacing of the data points points (delta) for the X and Y axes to determine the compensation table size.
The delta must be a multiple of the difference between the min and the max.
The range in which the compensator will be applied is defined by the min and max. The delta defines the amount of travel between the next compensation value in the table. Linear interpolation is used between points. To determine the number of compensation points required you can use the following equations.
1D Compensation: Points = (positionMax-positionMin) / positionDelta + 1
2D Compensation: Points = PointsX * PointsY
For example:
Axis
Min
Max
Delta
X
0
200000
50000
Y
25000
225000
10000
X_Points = (200000-0)/50000 + 1 = 5
Y_Points = (225000-25000)/10000 +1 = 21
CompensatorPoints = X_Points * Y_Points = 105
Pre-allocating memory
Once you know the number of compensators you plan to use and the number of points in the respective tables the functions CompensatorCountSet(...)
and CompensatorPointCountSet(...) to reserve memory in the motion controller for the compensator. This must be done before initializing objects that use dynamic memory(Like axes). This will reserve the space in memory on the motion controller. To check the amount of memory available for the compensator and other RSI objects you can check ExternalMemorySizeGet().
Configure Your Compensator
Once you have started the network you can call CompensatorConfigSet(...) to configure and enable your compensator. To disable your compensator you can use CompensatorDisable(...). For a full list of available functions see our API page.
Compensators use counts NOT User Units
Compensator does not affect your command position. The command and compensator output are combine right before getting sent to the drive. You can use CompensatorPositionGet() to see your the result of your compensator(s). The compensator will change your actual position Make sure your position error limit is larger than the difference between your largest and smallest compensation table values.
Multiple compensators can be applied to an axis. Overlapping compensators will add their result.
Example table
For example if you were at X=100000 and Y=115000 The value of -1800 counts would be added to your Z target.
Single Axis Compensator
It is also possible to setup a compensator with the same input and output axis. Such compensators must be configured to use the input axis's command position as the actual position would cause a feedback loop.
An example use case for this feature would be correcting position errors caused by mechanical imperfections in an actuator.
Here is an example of how a commanded position vs a corrected position graph could look for a single axis compensator.
Last modified 19d ago