The drive runs very well in EPOS4's Profile Position Mode, but when operating with a master (e.g. Beckhoff TwinCAT or maxon's MasterMACS) the motor's motion looks rather jerkily in "CSP - Cyclic Synchronous Position Mode".
- What is the reason for this?
- Have the control parameters not be tuned well?
The "PPM - Profile Position Mode" as well as the "CSP -Cyclic Synchronous Position" mode use internally the same position and current control algorithms and the identical control parameters. If an unsteady motor run only occurs with CSP mode, the cause cannot be the controller parameters or the results of the auto-tuning. The root cause must be related to the "interaction" or data exchange between master and EPOS4.
The difference between PPM and CSP is that in CSP the external master (e.g. Beckhoff TwinCAT + NC-Library or maxon's MasterMACS) performs the motion calculation (or so-called "trajectory generation") and has to provide updated "Target Position" values (-> object 0x607A) to the EPOS4 in a precise cyclic clock.
More information about the different operating modes and the tasks of the master and EPOS4 for each mode can be found in the following linked Support Center document and the PDF attached there at the end:
-> EPOS4 / IDX: Control design, Operating modes and Bus cycle rates
The EPOS4 position control loop cycle time is 0.4 ms, i.e. faster than the communication bus cycle rate. Due to this the EPOS4 performs an interpolation in between the "Target Position" values provided each communication bus cycle. The difference between the last "Target Position" and the new "Target Position" is divided into intermediate positions based on the configured "Interpolation Time Period Value" (-> object 0x60C2/01). These interpolated position data values are used as "Position Demand Value" (-> object 0x6062) every 0.4 ms cycle of the position control loop. This ensures that at the at next expected new "Target Position" (= communication bus cycle), the drive has reached the last "Target Position" value at the exact time and a continuous smooth movement is ensured.
For correct interpolation and a steady continuous smooth movement, the "Interpolation time period value" (-> object 0x60C2/01) must(!) be configured identically to the effective communication bus SYNC cycle of the master (PLC). A missing or faulty configuration leads to the fact that the "Target Position" is wrongly interpolated and the position control loop does not get fitting "Position demand values" at the right time. From the user's point of view, this leads to a visibly and audibly jerky motor movement. In detail, each movement based on an incorrect "Interpolation time period value" consists of a short movement start and subsequent stop within each bus cycle. The "Target Position" is reached at a wrong (usually too early) time and the EPOS4 has to wait for the next update of the "Target Position" (= cyclic data transmission).
Main cause of problem:
The most common cause for an unsteady motor movement observed in CSP mode is the missing or wrong configuration of the "Interpolation time period value" (-> object 0x60C2/01).
The "Interpolation time period value" (-> object 0x60C2/01) must be configured according to the bus cycle time of the EtherCAT master.
Further potential problem causes:
Further potential causes for an unsteady motor operated in CSP mode can be:
- A master's bus cycle which is not precisely maintained or even some bus cycles are completely missing.
In EtherCAT based systems the EPOS4 typically reports an "EtherCAT communication cycle time error" (0x8184), "RPDO timeout" (0x8250) or also "EtherCAT PDO communication error" (0x8280) message in case of missing or strongly deviating bus cycles periods.
- A master's "Target Position" values, which are impossible for the drive train to follow physically or which are completely wrong.
For the correct integration of the EPOS4 in the system manager of a Beckhoff TwinCAT master please refer to the notes in chapter "6.3 Beckhoff TwinCAT Integration" in the "EPOS4 Application Notes Collection".