AC Drive/Motor Speed Fluctuation?

[ataslaki]

Mind boggling thing, which is above my head to understand. Did a search on the forum, but couldn't find anything as crazy as what's going on with my system.

I have an AC drive and a continuous duty motor. Controlled by a few sensors, the motor starts on command, and stops when it reaches end of travel (horizontal screw table) as signaled by a prox sensor.

Everything runs beautiful in the hours of 6:00am-9:00PM (night), but at around 9:30PM-6:00AM (night shift crew), the motor overtravels when it reaches the end. It sees the prox sensor, then the prox sensor sends the signal to the motor to stop (via the AC drive), but the appears that the command to stop the motor takes slightly longer at night and hence the overtravel.

Nobody changes anything on the AC frequency driver. All accel/decel/other paramters are as-is in the morning and night; but somehow things slow down at night and electrons move slower?

No additional machines start to run around 9:30PM, to cause a surge of some sort. No machines are turned down. On top of everything, the power supplied to the system comes from a special huge power "cleaner", which delivers 110.00VAC on the money, every time.

Any help in solving this vudu magic?

Thanks

 

[Skogsgurra]

Yes, boggles the mind. Doesn't it?

A few questions:

1 Do you have regen or a braking resistor?

2 Do you have encoder feed-back or do you run sensorless?

3 If sensorless; do you run vector or U/f?

4 Does temperature change?

5 Why do you have to use a "power cleaner" in the first place? Does your "dirty" voltage vary a lot?

6 Have you tried with slower ramps? (Needs adjusting the prox sensor)

A lot of questions, yes. But rather a short and tough session than an eternal one. The answer to your problems may be in the answer to these questions.

Gunnar Englund

http://www.gke.org

 

[Barry1961]

That you have a "special huge power cleaner" leads me to believe you have power problems at your site. I have been in several places that experience large voltage changes from day to night. In most cases the voltage increases after the second shift when some near by plant shuts down.

If the voltage coming to the cleaner is high how does it lower the voltage to 110.00? You might want to look at the incoming voltage on a o-scope.

The best guess I have is some how the voltage to the sensor is changing and causing the point at which it sees the table to change. My other guess would be that the distance between the sensor and target is changing due to thermal expansion/contraction.

Barry1961

 

[ataslaki]

Thank you for asking all those questions. Let me try to answer some of them.

First of all, I've made certain that the problem is one of electricity/power related, and not mechanical or loose wiring, etc. The time of the day when the problem starts is also at around 9:30pm EVERY DAY.

1 Do you have regen or a braking resistor?
No regen or braking available. Accel/Decl set to shortest time for quick response. No faults reported on AC drive.


2 Do you have encoder feed-back or do you run sensorless?
No encoder feedback. Running sensorless.


3 If sensorless; do you run vector or U/f?
Running u/f.


4 Does temperature change?
Plant weather station reports no significant change of temperature. Temperature at 9:30PM is usually same as in 7:00AM, when there are absolutely no errors (or overtravel).


5 Why do you have to use a "power cleaner" in the first place? Does your "dirty" voltage vary a lot?
I'm in Southern California; a region where we have lots of power surges and power interruptions (lasting few mseconds), which fearing that may cause interruption in this highly safety-hazard prone machine. That's why.

6 Have you tried with slower ramps? (Needs adjusting the prox sensor)?
No; not tried slower ramps. I'm pretty sure it may work, but slows down the start/stop (total cycle time).

 

[Skogsgurra]

ataslaki,

I think that you should consider the idea put forward by Barry1961. Even if you say that your voltage is "110.00 V" for the drive, it is not sure that control voltages for the rest of the machine stays constant. Or is it?

Running without braking resistors in a positioning application is tricky. Needn't be much difference in friction or load to change overtravel.

Also, U/f is known to be difficult in such applications. A vector drive (many can be switched between U/f and vector) can handle the job much better.



Gunnar Englund

http://www.gke.org

 

[jraef]

Any error that is consistantly repeated every day at the same time is something environmental IMHO. Delayed execution is often caused by interference. Comm signals repeat themselves until the error cheks are satisfied, and when severe interference comes into play, they need to repeat themselves more than normal. I think you either have some sort of other machine or system that comes on during those hours and causes interference, or there is something else external causing it. The prinicpal of Occam's Razor would have me looking at the simplest causes first since they are the most likely, but from your description it sounds as if you have.

If you are satisfied that you have thoroughly eliminated all other sources of interference within your plant, I would begin looking at RFI problems. I have had a few systems that were affected by powerfull amatuer radio transmitters nearby, which were only on at night when the HAM operators were skipping to foreign lands. It sounds far fetched, but when you eliminate the obvious, only the unobvious remains. You may try just increasing or double checking your sheilding systems on signal carriers such as the wires to the sensors or even the drive keypad.

"Our virtues and our failings are inseparable, like force and matter. When they separate, man is no more."
Nikola Tesla