Hi Everyone,

Could use some help with a small motor I am trying to troubleshoot.

I have attached a PDF with an overall amplitude trend, a spectrum, and a zoomed in view of the same spectrum that shows the side bands I mention below.

This is a 250 HP motor (ball bearings) driving a small overhung pump (ball bearings) in reduced crude service (600+ F). This is a constant speed machine (3594 RPM) , and the load on the pump/motor never changes as the process condition does not vary.

In the vibration data I have collected, at the motor inboard horizontal location there is a dominant 2x peak (119.8 Hz) that varies in amplitude considerably through out the time I recorded vibration data. It is anywhere from .35 IPS peak all the way up to .65 IPS peak.

When I look at the spectrum, there is a dominant 2x peak with an amplitude of approximately .23 IPS peak... this peak amplitude stays fairly constant throughout the data that I have recorded. There are also side bands spaced at 4.6 Hz around this peak which is confusing to me.

I have checked for bearing frequencies in the motor and also for rotor bar related issues.... neither seem to be the source.

If you look at the time wave form, there are random and non-repeating impacts that occur... they are there and then gone... I may see them in one 4 second time block and then not again for another 8 to 12 seconds. Again.... not repeating and random frequencies... this appears to be where the amplitude variance is coming from but I am unsure of the cause.

I have looked at the cross channel phase data at 2x run speed, but there is nothing to suggest there is misalignment between the pump and motor.

 The current balance for this motor was even across the 3 phases and the current draw was fairly consistent.

I am skeptical to call this a potential electrical issue, as the dominant peak is not at 2x FL rather true 2x TS.

Does anyone have any ideas on what may be the cause of this?

Please let me know if you need any more vibe data.




Files (1)
Original Post

Mile High,

Yes, I have come across a soft foot condition causing a high 2X Line Freq. Get yourself a dial indicator, attach the magnet to the base and put the indicator on the foot and loosen the bolt and watch the needle. If it moves to the positive(or lifts) you have a soft foot. Take a large chisel and wedge the motor up and try to slide a shim in. And repeat the process until the dial indicator stays close to zero.

As John suggests, I once confirmed that soft foot was the root cause by loosening one bolt at a time and watching live instantaneous spectrum 2X line frequency essentially disappear as I loosened the bolt and its effect due to soft foot.   For our rather small hp motor, this was not a safety issue.

Motors this size with ball bearings sometimes need a wavy washer to control tracking in the ball bearing.  Check to make sure this motor doesn't require that.  The result is a 'bounce' of the rotor with the bearing cage acting as the spring.


Ron Brook


As others have stated, check for soft foot.

Your vibration data I do not believe you can see 2xLF, in order to see 2xLF you will need low frequency hi resolution data to separate 2x run speed from 2xLF. IF 3594 is 1x rpm X2=7188 and 2xLF is 7200.

What type of coupling are you using? could element be dry/hard/rattling

Sidebands, could be bearing spinning on shaft or spinning in housing or cage rattling

Impacting in time waveform could be flow induced, impacting from bearing is not typically random in nature

Typically soft foot is related to tilted-soft foot/hollow base which has a natural frequency very close to 7200 cpm. By loosening one motor foot bolt at a time, you will find the foot that is soft, the vibration will drop/disappear. When you tighten the bolt by hand the vibration will shoot back up with hand tightening the bolt, in some cases.....


Add Reply


HACKATHON in association with IMC-2019