On a 2 Pole AC motor (50 Hz) with sleeve bearings, sidebands are appearing as indicated by the spectrum

Spectrum of Motor

Where as the spectrum zoomed at 1X (2978 CPM) is

Zoomed spectrum

The side bands are 65 CPM appart from the fundamental frequency.

What could be the cause for these side bands?

Looks like pole pass frequency, could be a rotor bar issue.
Synchronous speed 3000-2978=22cpm x #of poles=44 cpm, this should be the spacing around 1x 2x 3x 4x etc..If this was a bearing fault, the only frequency that would be close would be cage frequency, it is typically .4 time turning speed which is close to 1200 cpm...
Is the motor running hot? Tripping out? Unable to do the same work it used too? These are syptoms of rotor bar failure. Is the motor rotor cast aluminum or copper bars with welded end ring?

Dave,

It looks like pole pass frequency but IT IS NOT because the sideband spacing is 65 CPM and not 44 CPM.

There is no antifriction bearing in the motor.

The motor is rated at 56 Amps and 6.3 KV. However it is operating around 61 amps. The rotor is copper bars with welded rings.

There are no other visible abnormal sysmptoms except a cyclic noise.

At the risk of being a PITA, what was the resolution? Just want to double-check to make sure the 2978 is correct. If you're real speed was 2967 (but not captured right due to resolution), than 65 cpm would be pole pass frequency.

Assuming it's not pole pass frequency, could be some kind of process oscillation at 65 cpm.

What is the load driven by this motor? Is this driven by a variable speed drive (that could also be a source of oscillations).

The resolution is around 2CPM and speed calculated from previous data collection also is the same.

This motor drives a gearbox which operates a centrifugal pump. No such frequency is visible on the train other than this motor.

Today the load was decreased to check its effect on the vibrations. There was no significant change in vibration amplitude or mpotor noise.

a 10 CPM increase in speed was noted with load decrease, the data is as under:

Motor Amps 63, Speed 2980 CPM, Side band spacing 64.79 CPM at resolution of 10 CPM

Motor Amps 60, Speed 2990 CPM, Side band spacing 65.56 CPM at resolution of 10 CPM

The amps and speed don't exactly make sense. 2990 should be about half the load of 2980, but the current went down by only 60/63. But maybe it was an increase from 2984 (round to 2980) up to 2986 (round to 2990).

But even so, it does seems to suggest as you said that it's not pole pass sideband since the sideband spacing appeared to go up (within our ability to measure at 10 cpm binwidht) when slip went down. And knowing your earlier measurements were based on 2cpm bin width seems to prove you had enough resolution to conclude it's not pole pass.

Is there a vfd? If so that can cause oscillation and sidebands.

If no vfd, my guess would still be that it could be originating from torsional oscillation induced by the load. Maybe the torsional oscillation is only coupled to lateral vibration at the motor (after all, the stator frame is the only stationary part that sees the torque... and if assymetric support H and V, that provides coupling from torsional to lateral). Or else could be broadband excitation exciting the train torsional resonance... and the motor again is the only place where that torsional behavior is coupled to lateral vibration that we measure.

Just guesses. I can't think of anything within the motor itself that would cause this.

This message has been edited. Last edited by: electricpete,

quote:

Originally posted by Ali Abrar:
The side bands are 65 CPM appart from the fundamental frequency.

I wonder as to how 65 cpm was measured. How accurate is 0.02 x RPM ? Third significant number will make all the difference in the world:
0.020 * 2978 = 60
0.024 * 2978 = 71
0.016 * 2978 = 47

Regarding the fact that side band pattern is present only on the motor suggests that it is due to a variable torque ( produced either by motor or load). Only in a device like motor torsional vibration can be sensed on a stationary part like stator.

Yes, but look at the zoomed spectrum (which as I understand it was developed from 2cpm bin width FFT). The 1.02 sidebands peak is more than halfway to 1.04 and the 0.98 peak is more than halfway to 0.96 (you can tell because the second sideband falls outside of 0.96 and 1.04). So the actual sideband spacing is more than 0.02 orders = 60 cpm, not less.

That's right, the SB is rather on the upper side of 0.20. Then, I agree, it must be load variation.

Is cyclic noise modulated at 65 cpm frequency?

Has any current analysis been done to see if there is any sidebands on the incoming power to the motor. This would rule out the rotor.

Current analysis would be useful, I agree 100%.

If you have sidebands, you must have modulation. When you have modulation it could come from load variations, two machines sitting next to each other operating at slightly different speeds, mechanical looseness, etc... Typically a pump has a constant load, to verify, guage reading for flow and pressure should be steady. I am assuming the vibration is sychronous, and you state the vibration is only on the motor, what type of coupling between motor and gearbox? I have seen gear/grid coupling's go dry and create this type of vibration...can you see the coupling? Do you have strobe light, this would help in looking at the coupling while the machine is running. There is also a fair amount of harmonics present in the motor spectrum indicating journal bearing clearance. What is the history of the motor, has the number of run speed harmonics increased? Has the amplitude of these harmonics increased?

Ali, do you have a spectrum (acceleration, preferably) out to about 100 orders? If we are going to discuss rotor bar problems, we need to see the rotor bar/slot pass frequencies.

The spectrum in acceleration units upto 100 order is


Spectrum Higher Order

There is no VFD.

Few days ago, the bearings of the motor were replaced to restore the clearances. The vibration amplitude decreased however, sidebands remained as such.

This message has been edited. Last edited by: Ali Abrar,

Ali, please use the "multiple point spectrum" option in your software. Find your machine on the left and select HVA readings on motor gearbox and pump. You can only have 18 spectrum at one time, and post for us all to see. When you changed the motor bearings was the coupling unbolted/removed on motor/gearbox? Need some more data, noise modulation can still indicate looseness and you still have a gearbox and pump that could have the looseness. Need to know motor speed and pump speeed please.
Dave