O.T DC drives, stability, pid

Ok, first I'll describe the battle I have been in for the last two days and what I did to "fix" it. I hope at the end someone has some insight as to what might have happenned.

I have a machine that has 4 identical Siemens 6RA70 DC drives and motors with encoder feedback running in speed control. This machine has been in service for about 7 years without a hickup.

I got called down a couple of days ago because one of them was acting eratically (not running smoothly and jumping around). Initially it looked like a feedback problem so we disconnected the belt from the motor and tried to jog it. It wouldn't run smooth but would speed up and stop about 8 times per revolution. Here is what we did to troubleshoot.

1. Swap encoders from one of the other running motors. Same problem

2. Swap encoder wiring from another running motor. Same problem.

3. Run both encoders in lathe and check with Oscope. Both encoders seemed good.

4. Changed the drive from encoder feedback to armature feedback to eliminate encoder completely. Same thing.

5. Checked reference to drive to ensure it was steady.

6. Removed motor and took to local reputable rewind shop and set up on test bench. Motor ran like new and shop said nothing was wrong with it. Reinstalled motor.

7. Started suspecting a possible feedback problem in the drive itself. Verified all paramaters as being identical with drive beside it and then swapped the drive out. Still the same problem.

8. Started questioning the report from the repair shop so we installed the motor that was running fine from the adjacent station. Same problem.

So at this point everything from the system had been changed with parts from a known good system and the problem persists. Went over EVERY wire from end to end checking terminations, megger, you name it we did it and everything is fine. Keep in mind we are just trying to get the motor to jog smooth without a load!

While going over the drive paramaters I noticed that kP seemed high. It was set at 40 but so were all of the other perfectly functioning drives (factory default is 3). From the onset it looked to me like the motor was out of tune (tuned too hot), however it had been running with the same parameters for 7 years along with it's 3 sisters without a problem.

I am always leary of changing paramaters in anything that HAS been functioning fine be it a PLC or a Drive or anything, but at this point we don't know what to do. I changed kP from 40 to 20 to see what would happen. The problem went away. We put the machine back together and ran it but it seemed a little soft so I changed it to 30. Running fine ever since at 30.

The conclusion I have come up with is that this drive was tuned to the brink of being unstable and that something has changed ever so slightly to put it at the point of being unstable. What that something is or could be I don't know but I would be interested to see what others think.

Time for a beer, it's days like this I start to think I don't know anything about electricity.

In my epxerience with DC drives, I have usually found brushes/commutator problems but you already eliminated that.

The next most common root cause has been a gradual change in the load that finally broke the camel's back....but you eliminated that too...

Perhaps the load did change and, like you said, the drive was tuned so hot that it ran goofy unloaded too. Do the other drives act up unoaded as well?

Nevermind...enjoy your beer!

I had a similar situation a while back on an inverter.. but it was a simple speed, the hz setting was 90, after making a post here and reviewing the parameters I found that by reducing the hz from 90 to 60 it has ran fine since, (thanks to the one's that helped)

I don’t really know why it started to fail, suspect now that the components (drive, motor, bearings, etc) where 10ish years old putting more of a load on everything and at 90hz loosing the torque

BTW after I lowered the speed to 60 you could not tell the difference, the speed reducer/gear box had slowed everything down anyways, faster or balls to the wall is not always the best way to run.

My thread/issue http://www.plctalk.net/qanda/showthread.php?t=29007

Nevermind...enjoy your beer!

Click to expand...

Could not of said it better {{cheers}}

Obviously this happens but not a lot

Usually it is noise. Reducing the encoder resolution will have a big affect on how high the gains can be.

There are probably three PID controllers involved here, any one of them could be tuned too hot. First, I would check to see if the field current controller is stable. Then the current inner loop. Finally, the speed outer loop.

It sounds like the problem is in the speed outer loop from your posts but it wouldn't hurt to check stability in the others as well.

There are probably three PID controllers involved here, any one of them could be tuned too hot. First, I would check to see if the field current controller is stable. Then the current inner loop. Finally, the speed outer loop.

It sounds like the problem is in the speed outer loop from your posts but it wouldn't hurt to check stability in the others as well.