Hello Folks,
I have installed the latest XPID Sabertooth Firmeware with 360° option and use the dcs-plugin.
If the signal goas from 360° to 0°, there is a jerk: the motor stops for about 0.5sec,
then there is a spike in the opposite direction, then the motor follows the target signal normally again.
I can rule out that this is not caused by power supply, Sabertooth, Arduino, Poti, Motor/Gear,
X-Sim axis settings, game values or x-sim dcs plugin.
Here is a video (Com3, position 1):
https://drive.google.com/open?id=1gTqcD ... m4n0ZUgE1G
The error is hardly visible here, i.e. it is lowest with the standard PID values (4.5, 0.4, .04).
and only noticeable as a short pause.
- The error is not recognizable in dcs plugin...values are smooth and steady while going from 360 to 0 and back.
- Also i can´t see an delay in the blue target from the xpid software.
But with other PID values, e.g. 6.0, 1, 0, the spike in the opposite direction becomes so strong that the sabertooth switches off (LED flashes red).
Here at sec 21 you can see the error clearly:
https://youtu.be/KlALa_obfkcWhen the axle is loaded with 220kg the effect squares and eveb with low PID values the whole system becomes unstable (runaways)!
- The error is only present in the transition if the signal is slow. At a roll rate of 30°/sec the error is nearly not present.
- The error occurs only in one signal direction! So from 360° to 0° everything is smooth. But not from 0° to 360°.
A simple solution would be to integrate a spike filter into the axis,
but then the 360° option will not work anymore (motor turns in opposite direction).
Same effect with smooth.
So i´m totally stuck. The Bug must be in the xpid software somewhere.
I´m really glad for any hint or help or maybe somebody who has the same problem.
Thank you all!