Siemens SINAMICS S120, Function Manual

Page: 630 / 1094

Function modules
8.15 Cogging torque compensation
Drive functions
630
Function Manual, 11/2017, 6SL3097-4AB00-0BP5
8.15.3 Filling compensation tables
Cogging torque compensation is executed via a table p5260 which, depending on the
position of the motor measuring system, is read out and precontrolled. The table is entered
in Nm for rotating motors or in N for linear motors.
Configure settings to fill the compensation tables.
The following parameter settings are important when filling the compensation tables:
Para-
meter
Bit Index Value Significance
p5251 0 - 1 Activate the option “New slow learning” By activating the option, the correspond-
ing compensation table is deleted with the learning process. During the learning
process, the cogging torque compensation is automatically deactivated.
So that the acceleration torque does not distort the results, this option may only
be activated after the normal speed is reached.
With slow learning, the motor must be operated at a very low speed (1.5
-1
) or
velocity (0.1 m/min). During the learning run, the cogging torque can be measured
when moving the motor and stored in the table.
Additional values are averaged when passing through several periods. The num-
ber of average periods can be viewed in r5254[0], the actual table index in
r5254[1]. The table is only completely filled when r5254[0] ≥ 2.
1 - 1 Activate the option “Supplementary slow learning”. This setting is used to fill the
compensation table in several steps across multiple learning processes. This is
required, for example, with a linear motor. During the learning process, the cog-
ging torque compensation is automatically deactivated. This supplementary learn-
ing is only possible for non-direction-based cogging torque compensation
(p5250.1 = 0).
During supplementary learning, gaps in the table are filled in. Where there were
previously already values in the table and new values are measured due to over-
lapping, an average is used (old value/new value).
An overlap is necessary for supplementary learning. Without overlap, the values
of the supplementary learning are rejected, and only the values from the 1st
learning operation are kept.
0.1 - 0 Shuts down the slow learning. Following this, the mean value of the measured
values in the table is determined and removed.
2 - 1 Removes the mean value of a compensation table. This setting is required if
compensation values calculated through learning must be manually corrected at
the start/end of the compensation table. The values in the table to be corrected
must be deleted from the range at the edge of the table!
p5252 - - n Used to define the length of a compensation table. When specifying the length,
only powers of two are permitted (2
p5252
). To achieve a sensible compensation
(for linear motor), at least 10 compensation values are required for each pole.
•
Example: linear motor: p0315 = 30 mm, traversing distance = 1500 mm. As a
consequence, this results in p5252 ≥ 1024 values.