Practical Solutions to Machinery and Maintenance Vibration Problems
Chapter 7, Misalignment
Section 15, Separating Electrical Hum from a True Harmonic
Depending on the type of vibration instrument available, there are several ways to determine whether the peak in question is an actual harmonic or electrically induced. If possible to shut the motor down, an electrical hum will disappear instantaneously when the power is cut off. For a fast enough reaction on an FFT, use a resolution with the least number of lines. Some instruments do not respond instantaneously to the elimination of electrical hum. In such situations, simply feel the vibration with your finger tips. If caused by electrical hum or buzz, the feeling will go away the instant the power is cut off.
Using a FFT, you can zoom in on the peak in question while providing maximum resolution. With this large expansion, you can usually see that the peak is actually two separate peaks, one at the harmonic frequency and the other at the electrically-induced frequency. You can view the portion of the spectrum that shows the peaks from zero frequency up to the peak in question. When the power is cut off, the frequencies of peaks at 1 x rpm and its harmonics will decrease in unison with each other as the motor decelerates. The electrical hum peak will not decrease in frequency at all. Instead, only its amplitude would drop to zero.
The synchronous time averaging feature of most FFT's can be used to
determine if the suspected
Focusing on the hum frequency, electrical hum vibration on a motor frame is much stronger in the torsional direction than in the typical vertical and horizontal measurement directions. For a motor that cannot be shut down at the time, this fact can be used to determine whether the vibration is due to a true harmonic or to electrical hum. The usual vibration amplitudes at frequencies due to unbalance or misalignment are revealed primarily in the radial direction, such as vertical radial and horizontal radial. However, as electrical hum vibration is strongest when measured in the torsional direction, simply place the vibration pickup on the motor's outer case, with the pickup stem tangent to its circular periphery.
For example, with the pickup stem positioned in the vertical direction, obtain a radial vibration reading. Keep the pickup stem positioned in the vertical direction, but this time place it in on the motor frame to obtain a tangent reading.
If the vibration is primarily radial vertical, it will result in approximately equal amplitudes as measured at both points. However, if the amplitude is considerably higher with the pickup in the tangential direction, it is an extremely strong symptom that the vibration is due to hum (and not a mechanically-induced harmonic). To further verify, repeat the procedure with the pickup in the horizontal radial and horizontal tangent directions.
There is another possibility for a peak induced by the electrical current that is relatively common. It is not a peak caused by actual vibration. Instead, this peak at electrical line frequency is caused by a bad electrical connection from the pickup to the FFT. Most often it is caused by frayed shielding. Sometimes it is caused by a stray electrical signal from the motor to the pickup base. The shielding problem is usually revealed by changing connecting cables. The stray electrical signal is usually corrected by placing insulating tape between the pickup and the motor. If this "false vibration" signal is suspected, you can determine if it shows up with another type of instrument.
To review, when it appears that there is an extra large amplitude harmonic
at a frequency that is at or near the electrical hum frequency, do not
use the harmonic information in your analysis until it is first determined
the extent to which the main contribution is a real harmonic or if it
is electrically induced.
This textbook contains only part of the information in our Practical Solutions seminar.