Suppose I measure both current and voltage going to an electric motor. So that I have two continuous streams of three phase data (current and voltage).

I want to use this data to diagnose potential problems with the motor. Some motor faults are supposed to have a small effect on the current pattern of the motor.

The way I understand it do changes in the grid mean small variations in the voltage. Either in the RMS or in the frequency of the measured voltage. If these differences are small over time, the grid quality is good. If the differences are big, the grid quality is poor.

Since I measure both current and voltage my intuition says I should be able to go to a representation which is more or less invariant to these small changes in voltage. I was thinking about the instantaneous electrical power.

I know instantanious electrical power is given by:

P(t) = I(t) * V(t)

So I could multiply the current waveforms with the voltage waveforms and get the instantanious power.

However, what I am not sure of is what the effect is of small variations in voltage due to changes in the grid. For example, how do these changes affect the current and the power? I assume the motor runs the same before and after a small variation in the voltage, is that correct? If I look at the instantaneous power instead of at the current, am I less susceptible to changes in the grid than when I look at just the current? Or do you know of other/better ways to compensate for changes in the grid?

  • \$\begingroup\$ I think you should add a paragraph to the beginning of your question explaining why you are measuring grid voltage in the first place. My initial thinking is that you only need to measure voltage to assess grid quality. Measuring motor current may help you detect problems with the motor. So what are you trying to accomplish? \$\endgroup\$ – mkeith Nov 13 '17 at 15:23
  • \$\begingroup\$ If grid quality is a big issue in your location, then unbalanced phase voltage can cause more heat in motor and more vibration. Speed changes from load shifts cause more power variation in consumption also abrupt changes in line f with inertia can cause big changes as well as tap changing power conditioners draw more current with tap increase and if grid impedance is higher than normal, causes more instability to grid voltage from V drop at higher current. %V change on motors affects RPM only slightly and often less than load changes unless >>10%. So beware of compensation stability issues. \$\endgroup\$ – Tony Stewart EE75 Nov 13 '17 at 15:41
  • \$\begingroup\$ @mkeith You guessed the application right. I added two sentences to clarify that I am trying to find current patterns that indicate a possible motor fault. \$\endgroup\$ – Semi Nov 13 '17 at 15:42
  • \$\begingroup\$ Analyze your Power Conditioner ESR (delta V/delta I) and compare with DCR of motor coil. This ratio of source/load DCR is the chief factor that determines load regulation from source V or load torque that cause RPM stability issues. Essentially adding to ESR of the grid. So tap changing or PWM compensation must be done wisely. \$\endgroup\$ – Tony Stewart EE75 Nov 13 '17 at 15:46

The following diagram shows the effect of induction motor operation of small to moderate changes in voltages. The load torque is assumed to be constant. Any change in starting and maximum torque can be interpreted as a torque variation over the entire torque vs. speed (T-S) curve. If the load torque is constant, the change in the T-S curve will result in a small change in operating speed. If the load torque increases with speed, the speed change will be smaller.

It will be very difficult if not impossible to determine if a change in motor current or power is due to a motor problem, a voltage change or a load change.

Image by Ed Cowern

Image by Ed Cowern

  • \$\begingroup\$ Thanks for the explanation and the image, that is very useful. So if I understand you correctly, will a small change in voltage cause the motor to run slightly faster/slower (if the load torque is constant). So it will change the current pattern as well. But even if I don't know what is the cause of changes in the current, is my intuition correct that changes in the grid have less of an effect on the power than on the current? \$\endgroup\$ – Semi Nov 13 '17 at 16:32
  • \$\begingroup\$ Yes. The motor will continue to provide the power required to turn the load at close to the same speed, so power will be less affected than current by a voltage change. \$\endgroup\$ – Charles Cowie Nov 13 '17 at 16:53

My comments above agree with @Charles 's answer.

Adding a shaft torque sensor with RPM tach feedback is necessary to determine cause of source or load variations with current for motor fault diagnosis.

Then compare expected power with actual unless some std. load profile is known. Another symptom is bearing fatigue and then you need a vibration sensor with harmonic filters to detect wear from ball frequency rise for preventive maintenance detection that before overcurrent in motor occurs from seized bearings. But if ok and just overload protection then current sensing with motor temp. sensing is effective in fault detection/prevention.

  • \$\begingroup\$ Thanks for the clarification. I understand that to know for sure what's the cause of a change in current you need to measure the torque as well. However, do you think the instantaneous power is less affected by changes in the grid than just the current is? \$\endgroup\$ – Semi Nov 13 '17 at 16:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.