Servo motor overheating

Learn the NEC with Mike Holt now!
Status
Not open for further replies.

S'mise

Senior Member
Location
Michigan
I've got a machine with an Allan Bradley AC 3p servo motor VPL-B1002 that runs well but occasionally overheats for no reason.

I have seen it heat up to 100c while in standby.
Looking at the kenetix 5700 drive motor diagnostics screen, it shows 0 amp motor current but a motor utilization of 28%.

Line, Bus and control voltage are fine.

How the heck can a motor overheat while drawing no current?

If I wait a few minutes, the utilization drops to <1% and the temperature drops to normal all on its own.

As I say, this is without the motor running.

Any idea what's going on?
 

synchro

Senior Member
Location
Chicago, IL
Occupation
EE
Have you measured the line currents when this is happening in standby?
How the heck can a motor overheat while drawing no current?
Just a guess, but perhaps the motor is getting a DC current but the drive only measures the AC part of the output current.
Taking a peak at the manual, that drive has a "flux on" mode that will apply DC to the motor, but that should only last a short time prior to starting.
 

S'mise

Senior Member
Location
Michigan
Thanks syscro,
No i haven't checked current. Even if is, what is causing it? A bad drive?

Yes, the "Flux on" mode should only apply a small current just before starting the motor. So that's not likely the problem.

Obviously the utilization percentage shouldn't be at 28% when it's not even moving.

Could it be that it's struggling to hold it's position?
I assume when it stops at its home position the brake would hold it there.

Do these motors just loose their tuning at times and fight to stay put?

The only thing I can think to do is swap kenetix drives and see if problem follows to another motor.
 

paulengr

Senior Member
The servo system is probably set up to flux up the motor on idle. I forget if the Kinetix system is AC, DC shunt (doubtful), brushless or PM, but regardless all motors, even AC induction, except PMDC, have two current paths; the magnetic field path and the torque path. There is significant time and energy involved in fluxing up the magnetic field so it’s common with servos to keep this energized.

Aside from that servos are very torquey by nature and run very hot, up to 150 C limits. And on top of that peak torque and current draw can be extremely high, enough to destroy a motor in seconds.

You cannot totally lock down programming from making mistakes without causing performance problems, and motor protection (fuses for instance) can’t protect servos from overheating under “normal” operation. It is on the programmer to do things correctly. It is unlikely the motor itself is bad.

For this very reason servo code must have torque or current limits set up. The goal is not to totally restrict peak torques but to control them. Particularly without S-curves right at the point where the motor is changing speeds or directions in a trapezoidal move motion control calls for almost infinite torques (current) and S-curve smoothing or simple current limiting is an absolute must. AB Kinetix makes servo programming simple for non-controls programmers but also opens the door for inexperienced ones. Most servo programmers destroy a few very expensive motors while learning what not to do, myself included.
 

MD Automation

Senior Member
Location
Maryland
Occupation
Engineer
I assume when it stops at its home position the brake would hold it there.

I have zero experience w/ Allen Bradley and Kenetix equipment - but wanted to mention in the servo world I am familiar with (Lenze, Beckhoff, B&R), the use of the brake to hold motor position is application dependent. Meaning, the programmer can decide when to apply the brake after some timeout. The reason for this is that the brake relay is typically a mechanical device (or at least they used to be) and had a MTBF of a few million or so cycles. While that might sound like a lot, depending on the application, you could eat up the life of a mechanical relay quite easily, and be replacing good drives every year or 2, simply because a ~5$ wave soldered relay failed on some PCB.

In most of my vertical applications, I would program an "inactive" timeout of 30 to 60 seconds, then if that expires I'd throw in the brake. But during that time, the motor is holding position against gravity and "warming" up.

This does not explain your zero current reading, but still thought to mention it.

Is it easy enough to put a voltmeter on the brake connections and see if you can correlate stationary motor heating with NO brake applied, then subsequent stationary motor cooling w/ Brake applied?

Good luck.
 

S'mise

Senior Member
Location
Michigan
Thanks guys.
What I do know is there have been no motor overheating problems in the past 2 years. So programming mistakes should have presented themselfs already.

This is an AC low inertia inverter motor. Similar motors next to it run much cooler.

I can understand settings may not be ideal for the operation but when machine is idol and drives are in standby, I should not see a motor go from 0 to 28% utilization.

Good point about motor brake being a separate function.

I wouldn't attempt without disconnect the linkage, but ....If I was to energize the 24v brake I imagine I might see the motor oscillate or drift because of feedback loop problems.

Can a bad encoder or feedback cable cause problems like this?
Regardless, I wouldn't expect any power going to motor while its in standby which is supposed to be a power safe condition.

I've read much of the manual without getting much insight. I was hoping it was a common problem with a simple fix.
Thanks for your insight.
 
Status
Not open for further replies.
Top