EME was contracted to determine how high some large motors could be uprated. We determined an uprating of over 50% could be achieved provided the motors be rewound with minimum insulation thickness thereby maximizing the copper while at the same time reducing the thermal impediment for getting heat out of the coils.

The manufacturer accepted to implement the uprate had no problem accepting the results of the study. An overpotential test performed after installing the coils was successfully performed. However, after wedging there were coil failures at a voltage lower than that at which it had previously passed. After making the connections there were additional failures at an even lower voltage.

What’s wrong?  The customer, manufacturer, and EME reviewed the situation and arrived at the following possibilities noting that all the failures occurred at the end of the slot.

A.) Insulation thickness is inadequate.

B.) The insulating tape was mica paper that might be mechanically inadequate.

C.) The pitch angle between the top and bottom sections had been adjusted to facilitate insertion in slots.

D.) All of the above.


The adjustment made on the pitch angle did indeed facilitate winding, but it does lead to stress at the ends of the slot. Although the adjustment was the manufacturer’s standard procedure, it had not been used with an insulation thickness at the rated voltage in this case. Tests showed that mica paper could compress and thereby becoming thinner as a result of the stress imposed.

The corrective action taken was as follows:

  • An additional layer of mica tape was applied resulting in slightly less copper in the slot. This required a reduction in the uprated but it still met the customer’s requirements.
  • The pitch angle adjustment was reduced such that the stress at the ends was less when the coil was in its final position.

Its worthy to notice that had any one of the contributing factors had not been there, the failures would not have occurred. The manufacturer had made many coils with no problems using the same techniques but with larger insulation thickness based on the volts/mil. Although this type of insulation would compress, the thicker insulation could tolerate the reduction.

Had the insulation been thicker, or had the insulation been of a different type such as mica splittings, or had there been no pitch angle adjustment, it is probable that there would have been no failures. It has been our experience that most failures are a result of multiple factors acting together that cause the problem. Taking the “shot gun” approach wherein all the factors are corrected is often justified.

Have an idea for a Quiz?!? Send us your suggestions at info@emeassociates.com