Using 3rd Party drives with 735x Controllers

[Wes_Ramm1]

Academic question.

I have an NI 7354 connected to a Parker Aries drive via an NI UMI-7774. The Aries drive is connected to a Parker DC brushless motor. I have an absolute encoder for feedback, which Is connected to the 7354.

Interestingly, the Parker drive units run a control loop between the drive and the motor, and I am closing the control loop on the 7359 on position. That means I effectively have two PId controllers interacting on this system.

Frommm a philosophical standpoint, this is not desirable, I feel. I don't want to perform two tuning exercises. At the moment, I have very low gains on the 7354 side (which according to the NI Motion manual, chapter 3 is appropriate), and everything is sta le and happy, but I get a little bit of chattering at static positions.

Following the Parker install guide, I performed the auto tune on the drive units, and then connected to the controller and got everything working. But in my mind, wouldn't I rather have only one control loop than two? I feel that the appropriate configuration would be to set the gains on the Parker drive to zero, and use it only as an amplifier. But, it won't amplify without some control gains.

Ill be honest ,I am pretty darn comfortable with PID for motion control, but This situation is over my head. Again, everything is working more or less perfectly, but my engineering mind wants me to eliminate one of the loops.

Any advice?

Thanks,

Wes

[Highland_Controls]

Wes-

What feedback do you have going to the drive?

Traditionally, the drive is used only as an amplifier.  The current loop is closed in the drive, while the velocity and position loops are closed in the controller.  It is usually also possible to close both the current and velocity loops in the drive, and the position loop in the controller.  Nowadays, it is also somewhat common to close all the loops in the drive, and the controller is relegated to higher level control and motion profiling.  In this case, the drive is either in step and direction mode, or it is on a motion network(EPL, SERCOS, Mechatrolink, etc.).  So at the very least, you must tune the current loop.  Parker offers a utility to do this configuration, but you need to tell it you are using it in torque mode.  Yaskawa has a serial encoder that tells the drive what motor is connected to it, and there is no need to tune the current loop.  I recommend using the Aries as a dump amplifier and only tuning the current loop in the drive.  Close the other loops in the controller.

Brian

[Wes_Ramm1]

Well this is perfect testimonial for RTFM, as doing so allowed me to get everything up and running quite stably in the first place.  And, since I have already tuned the Aries unit, that "my work here is done". 

So I guess that the Parker tuning is really more a function of the drive/motor/encoder recognition rather than a tune of the overal system.

[Wes_Ramm1]

I am going to leave this thread open for a few more days just incase anyone wants to weigh in. 

This isn't really a question that can be answered, it's more of a discusssion, so I'll see if anyone else has any opinion.