Speed controller drive

[Manglemender]

the print cylinder is 1000 mm diameter, 4 m long, there is a total reduction of (1/60) between motor en cylinder axis. The final gear is on the print cylinder and has 60 teeth on a diameter of approx. 25 cm, so very coarse.

The symptoms you have sound far more like an external disturbance but...

How much backlash is there in the drive train? If there is any amount of backlash then the load will float backwards and forward on the mechanics. If there is any compliance in the drive train then you have a "load on a spring" scenario.

How is your inertia match? As a servo application It should really be <= 5:1.

People often fail to appreciate the finer points of drive system design.

Nick

 

[aand74]

I also think about the backlash that causes the load to float backwards and forwards on the mechanics.
For the inertia match I should have to request more details about the cilinders.

 

[aand74]

Concerning the problem of the torque peaks, they seem to be present most when a beam that is used to lubricate the cylinder with ink is pushed against the cylinder. The effect can be seen on all of the 12 cilinders.
I have traced the P-action and I-action of the speed controller.
What supprises me is that most of the disturbance correction is done by the I-part. I would suspect this is not ideal and that most should be done by the P-part.
The P- and I- settings were obtained by an autotune process for the speed controller of the system. I did step responses and these autotune settings seemed to give the least overshoot, but for our process now I was surprised about the I-action.
Where should I begin : raise the I-time or raise the P-factor?

 

[aand74]

Just to add (because of bad picture quality), the light blue curve is the I-action, the dark blue curve the P-action, both have the same scaling, so you can compare their levels.

 

[Manglemender]

I would never recommend a speed loop autotune and it is generally worse on a high inertia load. What values did the autotune give you? As a starting point, set P-gain to 20 and I time to 200; this may not give servo performance but your mechanics may not permit higher gains.

Nick