Evaluate noise in electric motors
Determining the source can be more difficult than correcting it.
Just before the Christmas holidays last year, CFE Media & Technology hosted a webcast, “Evaluating noise in electric motors,” presented by EASA’s Tom Bishop, senior technical support specialist.
Bishop explained that determining the source of noise in a motor is often much more challenging than correcting it. However, a methodical approach to investigating noise can narrow down the possible causes and therefore make resolution easier. The session addressed the causes and characteristics of the primary sources of noise in AC motors.
Webcast learning objectives included the following:
- Possible sources of motor noise
- How to investigate motor noise
- Nature of magnetic, mechanical and windage noise
- Reducing noise intensity.
The webcast concluded with attendees asking their question for Tom Bishop. His answers to the questions follow.
Question: What are some salient points for anyone thinking about implementing predictive maintenance for motors in a plant or facility?
Bishop: Some of the initial things to consider are the cost of equipment such as spectrum analyzers, and the personnel that will perform the data gathering and analysis. That is, can it be done with in-house personnel, outsourcing with vendors, or a mix of both.
Question: Is it ever okay to overload a motor?
Bishop: Any overload on a motor increases heating and reduces thermal life of windings, bearings and lubricant. If the motor has a service factor it only applies if the motor is used in accordance with the conditions applicable to the service factor, typically what NEMA MG1 defines as usual conditions.
Question: When using a variable frequency drive (VFD), what type of meter should be used for measuring motor voltage and current?
Bishop: The meter needs to provide RMS (root mean squared) values so that it eliminates potential errors due to the high harmonic content of a VFD output.
Question: How far should a VFD be located from the motor using it?
Bishop: A VFD should be located as close to the motor as possible, so as to reduce the transient voltage increase “spike” at the motor terminals.
Question: Should vibration sensors be installed on motors for continuous monitoring of motor vibration?
Bishop: In most cases, vibration sensors should be installed at the bearing housings of motors being continuously monitored. Some larger motors with sleeve bearings use non-contact probes on the exterior of the bearing housings to measure the gap between the probe and the shaft and typically provide an output in mils (thousandths of an inch).
Question: Will longer core also increase overall size of the motor?
Bishop: The motor manufacturer typically extends the core length within a given frame; but in some cases, a longer frame is used to accommodate the longer core. The longer frames are usually above-NEMA where there are no standard frame sizes for a specific power and speed rating.
Question: Can you hear electrical harmonics in a motor?
Bishop: The sounds from harmonics that fall within the range of human hearing, 20Hz to 20kHz, can be heard. However, it would require a sound analyzer to determine the frequency and amplitude of the harmonics.
Question: How about noise from a VFD on a 3-phase motor?
Bishop: The most common noise from a motor on a VFD is a high pitched “singing” sound due to components in the motor being excited at their natural frequency.
Question: Do studies exist on motor noise reduction that can lead to reduction in energy consumption?
Bishop: I am not aware of any studies on this topic. Further, noise is essentially a low energy source, thus probably not accurately measurable.
Question: Can you play with a drive’s pulse width modulation (PWM) in order to decrease noise at the motor?
Bishop: Increasing the carrier frequency of the VFD above about 20kHz would make the noise non-perceptible to the human ear.
Question: If you have a slow speed wound rotor motor with dominant 120Hz vibration (.25 ips in vertical) and noise (95 dBA) that stops when power is cut, what would you investigate for root causes?
Bishop: The slow speed means that the motor winding has many poles, and that in itself can lead to noise. In general, I would check for loose laminations in the stator or rotor, and an unequal air gap.
Question: Are there affects from dV/dt that produce audible noise? (A dV/dt filter controls voltage spikes generated by VFDs.)
Bishop: The dV/dt (change in voltage versus change in time) does not result in noise unless it is so great as to cause partial discharge, which can produce audible noise. Partial discharge, also termed corona, is an electrical breakdown that partially bridges the insulation between conductors in a winding.
Question: Can acoustic and vibration sensors on a motor discriminate the different causes of noise?
Bishop: The sensors can identify the frequency of the noise, and then components of the motor can be tested to determine their natural (resonant) frequency. That is tedious, but sometimes necessary. If possible and practical, there are dual-microphone devices that work with spectrum analyzers. They can sweep across the motor checking for higher amplitude sources that line up with the microphones. Doing this from two different vantage points allows location of the source by triangulation.
Question: Larger motors are less noisy. Why is this not the case for 2-pole motors?
Bishop: The difference in noise for various poles is smaller for larger motors, but the base noise levels of larger motors are higher than for smaller motors.
Question: How can noise be reduced in a 2-pole motor?
Bishop: If possible, reduce the obstructions in the airflow paths.
Question: Do VFDs reduce noise?
Bishop: Motors supplied by VFDs typically are noisier than motors on sine wave (utility) power.
Question: How does a VFD affect motor noise?
Bishop: The VFD generates harmonics that excite various components in the motor, thus increasing noise.