------------------------------------------------------------------------------------------
Split from S7-1200 with Load Cell.

huggy_d1

Does it improve if you change from 100ms to 20ms?  I would expect a 10ms cyclic interrupt to be significantly smoother. 

DimiGav77

I'm working on a new application for a winder-rewinder. I use an s7-1200(1214) cpu with a weighting module wp231. The Web (fabric) moving with speeds from 20-80m/min. I call a cyclic interrupt every 100ms to make pid(pid_compact). I already proceed to fine tuning of pid but I saw instability on fabric tension. Is it normal with 40m/min speed and tension setpoint about 5kg to see actual tension 4.5-5.4kg? I also use for motion a frequency converter and a motor with gearbox. For speed measuring I use an encoder 900pps.

 I will try it.

Jeff Casagrande

I would also add more filtering to the load cell input.  They tend to be noisy devices.  You also probably want to use little or no integral gain (I assume you modifying the speed loop to control tension) as the diameter calculator functions much like an integrator already.  If I remember way back when, we used to multiply the gain by diameter to compensate the RPM needed to make a tension change.

 I will try also to filter more the loadcells. I think I use 5Hz on WP231PR_DB, DR03 struct FREQ_LOW_PASS_1 value. You mean no integral gain on PID used for speed control.

For the diameter. I'm not calculating the diameter, I measure it with a laser that give me an analog value (0-10V) proportional to actual diameter.

Separately calculated diameter?  Not standard, but it should work.  This is also another potential source of your tension fluctuation as this directly influences the speed.

If you are separately calculation diameter, then your tension controller loop should be PI, not P as I suggested earlier.

Curious as to why you have a separate, expensive laser when speed reference/speed feedback gives a perfectly acceptable diameter calculation.

Jeff Casagrande

Separately calculated diameter?  Not standard, but it should work.  This is also another potential source of your tension fluctuation as this directly influences the speed.

If you are separately calculation diameter, then your tension controller loop should be PI, not P as I suggested earlier.

Curious as to why you have a separate, expensive laser when speed reference/speed feedback gives a perfectly acceptable diameter calculation.

 I measure the speed of the main machine with encoder and I also measure the winder diameter with laser in order to calculate the winder speed(RPM). So, the system is first synchronized with that. For steady tension results I use loadcells between main machine and winder and I proceed to trim the calculated synchrozined speed with PID loop that takes the calculated winder speed, tension setpoint and actual tension as input and gives me a multiplier between 0.8-1.2 (20%) for trimming the calculated speed.

I'm trying to change the philosophy.

I will use servo system.

Jeff Casagrande

Separately calculated diameter?  Not standard, but it should work.  This is also another potential source of your tension fluctuation as this directly influences the speed.

If you are separately calculation diameter, then your tension controller loop should be PI, not P as I suggested earlier.

Curious as to why you have a separate, expensive laser when speed reference/speed feedback gives a perfectly acceptable diameter calculation.

Speed Reference measured with 900ppr encoder from main machine.

Speed Feedback ??? I'm not using encoder on the motor or on the shaft of the winder.

The system use only a VFD and asynchronous motor.

DimiGav77

 

Jeff Casagrande

Separately calculated diameter?  Not standard, but it should work.  This is also another potential source of your tension fluctuation as this directly influences the speed.

If you are separately calculation diameter, then your tension controller loop should be PI, not P as I suggested earlier.

Curious as to why you have a separate, expensive laser when speed reference/speed feedback gives a perfectly acceptable diameter calculation.

 Speed Reference measured with 900ppr encoder from main machine.

Speed Feedback ??? I'm not using encoder on the motor or on the shaft of the winder.

The system use only a VFD and asynchronous motor.

 Dear DimiGav77,

Are you completed your winder application?

How to do that ?

Could you please share me your project file ?

Best Regards,

Vorapob