How measuring volts helps motors work with drives
It's increasingly popular to combine adjustable-speed drives (ASDs) with three-phase induction motors, but they can have a stormy relationship if not applied properly.For example, a major ASD-related problem can occur when repetitive high-voltage spikes bombard the motor, due to fast power switching in the drive.
It’s increasingly popular to combine adjustable-speed drives (ASDs) with three-phase induction motors, but they can have a stormy relationship if not applied properly.
For example, a major ASD-related problem can occur when repetitive high-voltage spikes bombard the motor, due to fast power switching in the drive. These spikes range from 650 V peak-just above the drive’s bus voltage-to as high as 2,100 V peak in worst-case situations, such as long lead length, high carrier frequency, or multiple motor applications.
Problems may arise when the level of these spikes is higher than the motor’s Corona Inception Voltage (CIV). Accelerated aging occurs when an insulation system is repeatedly stressed at levels even slightly above its CIV-a condition that unfortunately exists in many ASD systems.
Despite the need to be sure a particular motor will function well with an ASD, determining suitability has been difficult for some time. NEMA Standard MG 1 provides application guidelines, but the success of some motors still varies.
One possible solution is to use line conditioning equipment, such as reactors, filters, or terminating devices to ensure spikes won’t damage a motor’s insulation system. In some applications, lead length restrictions control this problem. Another remedy is to buy a motor specifically designed for use with an ASD. Part 31 of MG 1 says these motors can withstand significantly higher voltage (1,600 V peak) than typical motors. However, even these motors may not withstand worst-case, 2,100-V conditions, requiring continued use of line conditioning equipment in many applications.
Clearly, knowing a motor’s CIV can indicate if it needs added protection in a particular application. For example, if a motor’s CIV is 1,300 V and the overshoot at its terminals is 1,550 V, either line conditioning, lead length restrictions, or a different motor is then needed. Because it’s difficult to determine actual overshoot severity, selecting a motor with a CIV able to tolerate worst-case terminal voltage could allow reliable operation without line conditioning.
Smooth simulated current response results in no corona (left),but produces a response with corona ‘noise’ (right) when CIV is reached.
To accurately determine a motor’s CIV, it must be stressed with a controlled voltage pulse of adjustable magnitude. Each of a three-phase motor’s phases must be tested twice, a total of six measurements. A typical test would connect terminals 1 and 2 to the equipment, then connect 2 and 3, and finally connect 1 and 3. The test would be completed by repeating these connections with the terminals reversed to stress each phase with both positive and negative voltage pulses, which occur when the motor is connected to an ASD.
The test pulse must have a very quick risetime, between 50 and 100 nsec, to stress the motor’s insulation in a way similar to most IGBT-based ASDs. While test pulses are applied to a motor at low frequency, between 200 and 400 Hz, the voltage is increased. Monitoring current response to the test pulses will indicate onset of CIV (see before and after corona diagrams). When pulse magnitude surpasses the motor’s CIV, corona ‘noise’ will appear on the current waveform. CIV of the motor becomes the lowest CIV value of the six measurements taken.
Today, no industry standard exists for CIV measurement or even a reference typically used to determine it. In Lincoln Motors’ lab, a high-voltage, variable dc power supply is directed to a high-voltage switching circuit, and this output is fed to the motor. This allows adjustment of the voltage, frequency, and width of test pulses. Though visual determination of CIV with a high-bandwidth oscilloscope remains the most reliable method, several manufacturers are trying to make motor CIV measurements simpler and more consistent.
Determining a motor’s CIV before using it with an ASD can eliminate unneeded line conditioning and lessen costly downtime due to premature insulation failure. Though there is no standard for measuring CIV yet, users should still ask suppliers how they measure CIV and how they interpret test results.
Chris Lanier, engineering project manager – variable speed motors, Lincoln Motors, Cleveland, O.