4.71.2
11/2/2020
SCN 3577 corrected a bug with hard coded filter constants that were incorrect when SERCOS was being used. The constants should be calculated based on the cutoff frequency and the sample time. The sample time goes from 1ms to 2ms when SERCOS is used so the constants needed to change for SERCOS.
SAMCO has a belt drive system that they were able to tune with V4.48.00 and not able to tune in v4.49.00 which is where is bug was fixed. After experimenting with their settings we realized that they found settings on the old software that were a balance between accuracy, smoothness, die stroke and the drive overheating. If we provide them a method to use the old hard coded filter constants, we expect they will be able to use the latest software successfully.
A new setup called Lag Command Velocity Filter has been added. Its default value is 35.6Hz. If SAMCO uses 17.8Hz when using SERCOS they will be back to the filter constant values that were hard coded. If they are using analog they should use the default values
When using SERCOS with a Die Accelerator there was a negative blip in the command profile at the start of the acceleration ramp. This blip did not exist when using analog.
The challenge for SERCOS is the motor encoder feedback is one sample old and the new command takes effect one sample delayed. The calculations attempting to account for this had a bug in them. They were attempting to make adjustments to the set-point ramp and when the ramp started.
The Set-Point ramp is now identical between SERCOS and analog. All adjustments for the old feedback and delayed command are now made in the loop. The error calculations in the loop are all made against the old Set-point which is also one sample old. The Lag Integral calculations that are removing the error will push the actual die position out ahead by one sample so that the actual die position will begin to match the new set-point by the time it becomes active and the die should be matching speed and position.
The only remaining customization in the set-point calculations for SERCOS is the tolerance test. Since we only know where the die was one sample time ago, the tolerance is compared with where the material was one sample time ago.