hw>Mom m Mom moved page S7 Controller Optimization to Public:S7 Controller Optimization without leaving a redirect |
m 1 revision imported |
(No difference)
| |
Latest revision as of 10:53, 9 March 2026
1 Usage[edit | edit source]
Quick tips on how to optimize a S7 temperature controller.
Some basic and valuable information can also be found here: ETS132_Controller_Setup_Rules
or here:
Cascade_Controller_Adjustment_Rules
Example Cascade Controller with electrical heating.
As you see the actual air temperature was quite unstable.
The controller output was fluctuating constantly between 0 and 100%.
So the behaviour was more like a P-control instead of a PI control.
Let's optimize with some basic rules.
Initial (horrible settings):
- Kp = 4.0
- Ti = 5sec
First step: Determine the critical kP-value.
Set Ti to zero - so that we have a pure P-control.
Increase Kp until the behaviour gets unstable and over- and undershooting gets bigger.
This value for kP is called critical kP .
Basic Rule: Set 50% of the so found critical kP.
Now we have a look at the cyclic duration - with this we determine the value for Ti.
While working as a pure P-control we check the cyclic duration. We find approx. 10 minutes.
Basic Rule:
Set Ti to 85% of the so found cycle duration - so set this to 510sec in our example.
Now...result:
Looks good. Different to the initial situation our controller output has a quite constant output, which leads to a constant load and therefore to a good temperature behaviour.
The interesting part starts with a setpoint change:
No further optimization necessary.
We made a setpoint jump from 120°C to 200°C and had an overshooting of 0.5°C - this does not get any better. Also the controller output looks really good.