lower speed control

[Yahoo1]

I am fine tune my motion control system right now and I encountered some issue which frustrated me.
I fine tune my system at 9,000psi (which is in 2 in" diameter area). At high speed, it worked fine; but at low speed, I always got overload error and current jump up and down untill reached the amplifier's peak current(15A).
Is it because the fine-tuning is not very good or because the amplifier's peak current is not high enough (my motor's peak current is 31A)? If it is because the peak current, how come it draw such big current at low speed?

Any opinion is very welcome.

Thanks,

Wendy

 

[jbartos]

Suggestions marked ///\\I am fine tune my motion control system right now and I encountered some issue which frustrated me.
I fine tune my system at 9,000psi (which is in 2 in" diameter area).
///Please, would you elaborate\\ At high speed, it worked fine; but at low speed, I always got overload error and current jump up and down untill reached the amplifier's peak current(15A).
///Apparently, the motor shaft load increases with the decreasing RPM.\\\
Is it because the fine-tuning is not very good or because the amplifier's peak current is not high enough (my motor's peak current is 31A)?
///If the motor current increases with the RPM decreasing and the motor is rated for the RPM decrease, then the amplifier is not adequately rated.\\ If it is because the peak current, how come it draw such big current at low speed?
///If the motor shaft load increases, then the motor draws the higher current.\\Any opinion is very welcome.


Thanks,

Wendy

 

[Yahoo1]

Hi, Jbartos,

Thank you very much for your reply.

Actually, the load on the shaft is constant. I fine tune the system under constant load. I don't understand how come such big change in motor current. I am wondering what it means of "motor is rated for the RPM decrease"

Thanks again.

Wendy

 

[EdDanzer]

Most motors have a constant torque below a certain rpm. Have you checked the torque required to drive the load at the minimum rpm, and compared it to the rated motor torque at that rpm?
What kind of feedback is being used? If the feedback resolution is not high enough it can cause system instability at low speeds.

 

[jbartos]

Suggestion to Yahoo1 (Industrial) Nov 3, 2003 ///\\\
Actually, the load on the shaft is constant. I fine tune the system under constant load. I don't understand how come such big change in motor current.
///The constant torque under lower speed causes the motor current increase, if the voltage is used to decrease the motor speed. The constant torque usually implies constant HPs\\ I am wondering what it means of "motor is rated for the RPM decrease"
///There are various NEMA designs of squirrel-cage induction motors, namely, A,B,C,D, and E. D characteristics has torque increasing with decreasing speed. \\

 

[Yahoo1]

To EDanzer,

Hi, EDanzer. Thank you very much for your helpful point.
I guess that probably the instability of my system is caused by low resolution. Now I have 1024counts/rev quadrature encoder, so the controller see it is 4096 counts/rev. And the looptime is 2ms. If the speed is 1rpm, for one loop, encoder counts changing is (4096/60)/500=0.136count/loop. So the controller can see the movement after sevral loops. What do you think about it? Is the resolution too low to use for my system?

To jbartos,
The torque of my motor just decrease a little bit with speed increasing. From the torque and speed curve, it can be seen the torque is almost constant regardless of the speed of the motor . Thanks

 

[EdDanzer]

It may help to use a higher count encoder, but is sounds like it could be a tuning problem. Depending on the control you can slow the response time by other filter parameters than just gain.
What make of control, servo amp and servo motor are you using?
It is also possible that the other parts of the system are causing or amplifying the instability. Excessive clearance in any of the components can cause this, use a dial indicator on the positioned part and a bar to move the system with the motor shaft locked to check for clearance in the drive parts. If the slide way has excessive clearance or is tight you can get slip-stick at low movement speeds that increase motor torque requirements.

 

[Yahoo1]

I am using Baldor's AC servo motor, servo amplifier and controller. And the continuous current and peak current of the amplifier and motor are not match. For the amplifier, the continous current is 7.5A, peak current is 15A; for the motor, the continuous current is 8.6A and peak current is 31A. Is it a big problem?

 

[EdDanzer]

I’ve never used Baldor servos, but it is good practice to have the amplifier rated higher than the motor. This should not be a problem unless you are running the drive above the continuous rating of the amplifier to hold the load at low speed. The way to tell is look at the motor torque at 7.5 amps, and see if this is adequate to start moving the mass plus starting friction.
Have you looked at all the parameters for tuning the drive?

 

[Yahoo1]

I calculated the torque which is reflected to motor shaft by hydraulic force, it is 22.5lb-in. And the continuous torque output of the motor is 40lb-in. The torque is around 35lb-in at 7.5A. So I think the torque is enough, and the current is enough.
I tuned the controller's parameter, including Proportional gain, velocity feedback gain, velocity feedforward gain, integral gain and integral range. I got amplifier's parameter from auto tune, including current proportional gain, current integral gain, speed proportional gain and speed integral gain.
Maybe it is tuning problem. But the things are after the fine tuning, if the motor run at short time, it is ok, the current jump up and down, but it is still whthin continuous range; if the motor run for several minites, the current keep growing untill overload. I am wondering if it is because the mechanisim of the system.
It is really frustated.

Thanks

 

[EdDanzer]

Although I’ve had limited servo tuning experience, your description sounds like too much proportional gain. Lower that value 30% and see what the results are. There should be a parameter to determine acceleration and deceleration rates, (speed gains?) that could be reduced. My guess is that there is some slip stick that causes the gain to start oscillation.

 

[Yahoo1]

I tested the system at 100RPM with load. After 12 minites runing, the current and following error began increasing untill it is overload. The motor is not hot but it is very warm.

 

[EdDanzer]

Is there any work induced loads that are creating more force than calculated? The description of the problems you are having could also be noise. Make sure the encoder wires are shielded and do not go near high power AC wires. Another thing to do is remove the load from the motor and just let it run under no load to see if you have the same problem.

 

[Yahoo1]

I run the motor for 13 minites this time. The operating temperature is 25.1c at the start and 29.5c degree at the stop. the current increased from 3.8A to 11A.

 

[jbartos]

Suggestion to the original posting: It is not clear whether there is open loop control or closed loop (feedback) control. The feedback loop control can be designed to be very stable.

 

[Yahoo1]

To jbartos,

it is closed-loop control. there are velocity feedback, velocity feedforward and intergration in the controller. And current loop and speed loop in the amplifier. I am wondering how come the current increasing seems time related.

 

[dford]

Yahoo1:

You may want to look, with an oscilliscope, at the voltage waveform out of the amplifier. Since the energy used on the load isn't changing, but the supplied energy is increasing, I would be suspicious of a high frequency hidden oscillation being the cause. Digital controls are commonly afflicted with this sort of problem. Good luck.

Doug Ford

 

[jbartos]

Suggestion: An integrator (I element or filter) makes variable (possibly current) rising (possibly slowly) in time.
See regular undergraduate textbooks on controls.
Nothing or Differentiator (D element or filter) might cause some fuzziness in variable behavior.
Visit

http://cmosedu.com/cmos2/Figs_35.pdf

etc. for more info

 

[Yahoo1]

To EdDanzer,
I run the system without load, it worked fine. So Is it still tuning problem?

Thanks.

 

[EdDanzer]

If the system runs ok with the load disconnected, then the problem could be either tuning, or a problem with what you are trying to move. My guess is that the load and mechanical connection are causing the problem. What are all the components between the motor shaft and the load?