Electrical Motor Defect

This was a first for me in 17 years! It was one of those with the right experienced people at the right place at the right time beauties!

We were called to a site when the operators reported an unusual sound from a 4 Pole 1500RPM fan motor. We attended and through vibration analysis and temperature measurements recommended stopping the motor and checking the connections in the motor terminal connection box. This enabled the site electrical team to quickly pin point the defect.

The vibration data indicated a dominant 100Hz electrical noise, the motor felt like it was hunting or pulsing and in addition the whole motor was hot with the motor DE housing was over 100oC.

It would be interesting if anyone else has ever found a defect like this?

Vibration Data:

The overall velocity Increased from a normal level of 1mm/s RMS to 3.9mm/s RMS and then in two days increased to 8.88mm/s RMS.

The velocity spectrum below displayed a new peak at 100Hz:

This is a single frequency 100Hz Trend, and shows a marked increase at 100Hz:

The PeakVue spectrum displayed dominant 100Hz and harmonics.

The acceleration 10 KHz spectrum displayed two mounds of activity with 100Hz sidebands:

We then decided to look at the auto correlation data and found the following.

10 KHz Acceleration time waveform circle plot with four main peaks and inner lower peaks:


PeakVue Acceleration Time waveform circle plot showing 4 clear peaks:

And this is what was found. (Note cable broke when it was lifted for photo.)

We were surprised the motor was still running (very badly with high vibration) in this condition !!!!!

13 thoughts on “Electrical Motor Defect

  1. What a unique find. Have seen various winning and bearing faults found with vibration testing. I bet that motor would never start in that condition.

    • Hi Jim,
      Yes, this was a beaut of a unique find! They actually stopped and started the fan motor when we were there, this was so close to failure. Images are from the motor in the shop so the cable has probably been pulled about by then.

  2. The vshooter vibration analyser would have detected this problem in seconds and you would have been able to download all the data.
    Kind regards,

    • Hi Hugo,

      What automatic defect would the Vshooter have classed the 2xLF peak as, would it have taken the more dominate 1 order only?


    • Hi Skip,

      The velocity data collection was 2000Hz 3200LOR, we didn’t get time to take higher resolution data as it was felt to be too dangerous and shut down for electrical termination checks.

      It did feel as if the motor was pulsing at 4 orders and felt really bad. The whole motor was very hot with near 100oC at the DE, the three phase supply cables were also very hot to the touch. Could the pulsing at 4 orders be due to the shorting of one of the phase connections in the terminal box? This would then account for the 4 order peak at 100Hz with no sidebands. This was a brand new 4 pole motor only been running a few months on a VFD.

      Certainly we a new one on us, very interesting.


      • When the problem you discovered happens on a 4 pole motor it can appear to be 4X RPM but it is really 2X Line Freq. which is very close to 4X RPM in the spectrum. I have seen it drive the amplitude higher than 1 ips (Peak) at this frequency. It can be identified as the problem you found usually because the amplitude gets so high compared to other “electrical” problems. It is also usually surrounded by 1/3 Line Frequency sidebands. That is what led to my question since 50Hz/ 3=16.67Hz.

        Many people refer to this as “phasing”.

        It can create some awful torque reactions through tout the drive train of a machine especially if driving something through a gearbox.

        Skip Hartman

  3. Hi
    On my experience I lived this type of problem but through a routine of monitoring vibration and temperature under a schedule.
    The first parameter increased with the time was temperature (85 to 90 centi degrees, before to detect significant changes on vibration overall values. With this information, the first action was a solo run finding high temperature too, so we began reviewing terminal box, finding something like what you showed on photos but on a early stage, I mean not so extreme. The action was prettty simple fixing the failure on terminal box on site, not at shop, then started again at full load with normal temperature. The cause was a poor sealant on box terminal. The motor of four poles drived a centrifugal pump at the bottom of a reactor but in a very acid enviroment. I hope be useful. Best wishes.

  4. This appears to be a “six lead” motor connection and from what I read, starts and accelerates both the motor inertia and fan inertia from rest. It is unusual that it would have the torque, with one lead burned off unless it was an “unusual” connection.

    Is the motor type:
    1. Part Winding Start-Wye, but connected as RUN only
    2. External Delta connected

    Just curious…

    • Hi Brian,

      Thanks for the comment. This motor is a 400kw delta connection in the terminal box. This looks to be a bad connection, it was brand new and had only been in around 6-9 months.


      • 1.The Seasoned Analyst did good job with representation of information. I like it. Only sadly that he did mistakes (possible due to misunderstanding Basics of Correlation Analysis) in plot # 5 and plot # 6. It will be correct if instead “Acceleration Time Waveform plot” and “PeakVue Acceleration Time Waveform Plot” will be ” Autocorrelation Plot”.
        2. Need to pay attention to Power Electronics of VFD (especially filter part). Need to ask Maintenance did they something with this unit and if not, carefully check it.

        • Hi Leonid,

          Thanks for your comment.

          1. I did omit to put in the description of plot 5 and 6 that it was the autocorrrelation of the waveform data, but it does state that on the plot. Next time I will ensure I describe the plots fully. Good point.

          2. Agree with you on this point. I have seen many A.C. VFD drives causing vibration issues and once they have correctly set up the electronics the vibration dissipates.

          I appreciate your feedback.


Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.