starting noise 1

[jclough]

Does anyone know what causes electric motor starting noise? The noise to which I am refering is that short "whine" or "chirp" which happens when a motor is accelerating to no load speed from a dead stop. It seems to occur mainly 60 hz line fed 3 phase motors.

Any input would be appreciated.

 

[aolalde]

My opinion is that the typical starting winning or chirp of induction motors is due to the current frequency induced in the rotor bars, as the speed increases the rotor current frequency drops until it reaches full speed .

Fr start = F
F run = s * F

Were: Fr start = rotor circuit frequency at start.
F= Line frequency ( 50, 60 HZ)
s = Motor slip
F run = Motor steady running frequency ( almost zero at no load condition)

 

[vanskat]

I would assume that the noise is as a result of the action of alternating magnetic forces which are a function of flux density.

Principle of induction

 

[testtech]

I have recorded start up sound as wav files and then conducted spectral analysis to determine the component frequencies on two motors. These were motors with start up sounds that were of concern to the operators. The start-up sounds are low order harmonics of 120 Hz. Rotor speed X number of rotor bars is also present. I believe that the sound is a function of the rotor bars moving back and forth in the high magnetic field.

Of course various mechanical conditions can contribute to start up noise. These include belt vibration, bearing defects, resonances to VFD modulating frequencies and more.

Is your concern academic, or do you suspect a problem?

 

[SteveKW]

What HP and RPM is the motor?

 

[jclough]

Gentlemen,

Thanks for your replies. We just recently comlpeted a new design for a new customer. Its a 220V three phase subfractional hp model running at 3600 rpm. no load current is .75A. the stator core is laminated iron. The windings are connected Y. The rotor cores are cast laminated iron. I've noticed what I've always called "starting noise". While this is considered a defect here no one (myself included) can give a definitive answer as to what the cause is. We've tried various things to fix the problem in the past...all with limited success

 

[testtech]

There is no need to guess at this. You must obtain a sound spectrum if you want to determine the source. The spectral frequencies of the sound will correspond to motor components or motor actions. The data will probably be easily interpreted. You should be able to find a local sound or vibration consultant who can easily provide this information.

 

[jbartos]

Suggestion: Submarines are very meticulous about various equipment noises since they are supposed to be very quiet.
It is a good area to study noise therefrom.
A starting Reference for the motor noises is:
M. G. Say "Alternating Current Machines," John Wiley & Sons, 1978, page 293 "Noise":
Mechanical Forces
Magnetic Forces
Aerodynamic Noise
All of the above noises vary during the motor start. However, the Mechanical Forces are heavily dependent on the motor speed implying the biggest noisemaking origin.

 

[electricpete]

I believe it is not a simple question.

There are many forces generated at RBPF and RBPF +/-2*LF.

As machine speed increases these frequencies increase through resonances of various parts.

I would think spectral analysis by measuring the sound and attempting fft would be challenging when the frequencies you are trying to measure are changing over time (during a start).

Another thing is the unbalanced magnetic pull is higher during starting for a variety of reasons. If a rub is going to occur it is more likely at start.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[testtech]

Electripete:

You are correct that the frequencies would change as the rotor speed increases.

If the sound lasts long enough for you to notice, than I expect you will be able to capture it in a waterfall plot or perhaps a peak hold plot. The waterfall should do the job. You can trade off spectral resolution vs speed to ensure that you get the data you need.

If you capture a sound file than you can play around with analytical settings without the need to constantly restart the motor.

In addition, you can probably capture the data as vibration using an accelerometer.

After you capture data, you can use sound generating software to recreate the frequencies you have measured and compare them with what you are hearing. This will serve as a useful proof of the data.

 

[jbartos]

Comment: The power supply quality influences motor noise from the start and during the motor running. Has the motor been supplied from a different power supply source with the good power supply quality to make sure that it is actually the motor fault?

 

[electricpete]

One experiment: start the motor at reduced voltage.

If there is a dramatic reduction in noise, then it seem reasonable to conclude that electromagnetic noise is likely the cause. If not, then it seems reasonable to focus more on mechanical sources: windage, resonances, bearings.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[aolalde]

jclough

The problem could be the motor design.
The magnetic circuit should be constructed from insulated high permeability laminated steel ( not iron). The yoke cross section is very important for two poles it could saturate if marginal. The relation of stator and rotor slots is important for noise and torque. The air gap, span, coil grouping and rotor skew play a role with smooth operation too.

 

[jbartos]

Comment: It is not clear from the original posting whether or not there is a VFD. Visit

http://www.itee.uq.edu.au/~aupec/aupec02/Final-Papers/N-Hashemi1.pdf

for:
Using a PWM inverter makes it possible to run the machine at different speeds, but this will be noisier than with a purely sinusoidal supply [7, 11-13].