Scenario: 1,500HP DC motor driving gearbox. A first set of vibration data shows 360 Hz with its harmonics. Motor running approximately 550 RPM. No other significant peaks in spectrum. Waveform shows nothing abnormal. PeakVue – just low amplitude noise. Second set of data, taken 4 weeks later – no change. Recommendations – check DC drive for proper operations. Otherwise – continue to monitor.

Another vibration analysts, who routinely monitor this motor takes data at almost the same time and his recommendations are: “high vibration at 360 Hz indicative of a severe bearing defect. Replace motor as soon as possible.”

Motor has been removed from service and will be stripped down.

I was puzzled why “360 Hz is indicative of a severe bearing defect”. I am sure it was electrically related.

Anyone out there that can shine some light on this one? I am lost……

Kris,

If it is exactly 360 hz, then it is very probably scr firing frequency of a full wave rectified dc motor control and is quite possibly normal. If you can post some spectra, we can probably be more specific, but it is almost always there to some degree in a modern dc motor. You may find a few half wave rectified controls that make 180 hz but not that many.

Danny,

thanks for your reply. It was exactly 360HZ. I am 100% sure it was electrically related.
What I do not understand why the analyst called it “severe bearing defect”……

attached is a spectrum...

360HZ_DC.doc (36 Kb, 142 downloads) 360Hz

Kris,

On some bearings the BSF frequency is very close to 120HZ, maybe the analyst made a bad call and thought it was 3xBSF...

Kris,

The graph you posted shows 360.24 hz I think. I don't usually analyze spectra in acceleration, but the floor energy stuff could be something bearing related. If you have a PeakVue Spectrum and TWF, I could tell more. If you could change this plot to velocity and verify the cursor value, that would help.

Hi Kris,
Danny is right, it is probably the SCR. If it is new or amplitutes have incresaed it is likely to be a faulty diode in the rectifier circuit. Cheers

You always have 6xLF in a modern DC drive unless someone spent a lot of money to filter it and that is only required when it´s hi like 9mm/s due to local resonance in motor at that frequency as example. It will then shorten bearing life if left unmodified. If you replace the motor in that case it may change as the resonance is very individually dependent so if you change again it may come back, tested and verified with drives for papermachines. If using some automatic analysing it may fall into the category "everything not speed syncronous are bearing problems" but the world is not that simple. Olov

The magnitude is low thus not indicating a problem. Elec scr is probability but either way no real problem.

Danny,
This is exactly 360Hz. I checked it many times. So it definitely is electrically related. 360Hz as it is 6-pulse rectifier. Both waveform and PeakVue did not detect anything that could be indicative of an antifriction bearing fault.

Planty, OLI, Sam
Yes you are right – every DC drives produces 360Hz, or rather here in the states where line frequency is 60Hz . In Europe it would be 300Hz and 50Hz respectively.

I have seen the waveforms taken from the drive using digital oscilloscope – there is nothing wrong with the drive. The pulses appear to be normal. So it is quite likely that the combination of this particular drive (quite an old one) and the motor produces this spectrum.

I agree the amplitude is low and really not of a concern. What puzzles me why the analyst recommended replacement of the motor because of a severe bearing fault. What is more he did mention the 360Hz “is indicative of a bearing fault”. I spoke with a friend of mine who monitors many similar applications, if not identical ones, and most of them exhibit similar behavior. The 360Hz is present but does not warrant any corrective maintenance actions.

The question I still have is did anybody of you seen DC motor bearing fault that manifested itself as 360Hz (6 x line frequency) in a vibration spectrum?

As I said before I am still puzzled by the vibration analyst diagnosis……maybe he just made a mistake? Somebody also suggested to check who is he working for? A motor repair shop? But would be unethical, wouldn’t it….

I am posting a spectrum from a DC motor roughly the same size as yours. By no coincidence it is the outboard axial point. The vibration is very low. At the same time the 360 Hz is one of the two dominant peaks. The interesting part is that the power is not from SCR, but from 3 phase, full wave rectifier fed by a variable 3 phase voltage source. It means the current (644 amps) is as smooth as it can get. But the 360 Hz still shows and showed on all motors of this type in the past. I blame the large, flat endbell that certainly has some resonant frequencies and one of them may be close to 360 Hz.
The other peak at 765 Hz is the rotor slot pass frequency (57x running speed, 57 rotor slots). Again a peak that appears on every signature of this particular motor. I have never investigated those peaks closely because they do not seem to be a problem. Similarly in your case I believe that the problem is not a problem at all.
jank

MOA360Hz,57slots.pdf (134 Kb, 72 downloads) 360 Hz peak and slot pass peak

I had 360Hz come up about 15yrs ago on a paper machine DC drive. I concidered brg fault but the freq was fixed, did not move with speed, and exactly 360Hz. Cant remember amplitudes but it was a new peak, or maybe a jump up from a small peak. Sparkies found a faulty diode, replaced it a frequency disappeared. Funny thing is Im sure it was 360Hz and we are 50Hz line frq. Mind you it was 15yrs ago and my memory is a bit fuzzy.
Did the peak dissapear when the motor was changed?

In our plant we have several DC motors that run at various RPM's depending on product. Our SCR's are 6 pulse. With a line frequency of 60hz every DC motor here has a 360hz peak. The quick way that I can tell between a bearing fault frequency and an electrical frequency is by using Orders as well as Frequency when looking at my data. If a peak at 360hz is constant reguardless of RPM, its most likely electrical, if the same peak is viewed in Orders and is constant reguardless of RPM then it would be shaft related and not electrical.
If the motor runs at a constant RPM this doesn't help.

I have not had any bearing defects show up at 360hz yet.

quote:

As I said before I am still puzzled by the vibration analyst diagnosis……maybe he just made a mistake?

Kris,

He might have made a mistake. Question is, how much experience does he have? Has he never encountered this frequency before? Is he an on site employee or contractor? Everyone will make a mistake now and then (well almost everyone, I know one who claims he never does and it is not me .)

Why was the motor removed before getting a 2nd opinion if there was any doubt on what was wrong, well, after looking at the opening post again, I guess the guy who called the bearing defect is the 2nd opinion. Who wrote the first condition report where it said continue to monitor? Are there 2 or more analyst monitor the motor?

I see in the spectrum that the energy continues out past the 2K Fmax. Has any data been taken in that area to see what is happening? Since this drives a gearbox, there may be some gear noise present, huh? Looks like maybe gearmesh at about 900 Hz and 1800 Hz.

Let us know what the outcome is when the motor is repaired.

Jank,

You confirmed what at least three people told me.

REJ,
Thanks for the info.

Ralph,

The call was made by the contractor. I believe he had done monitoring for that plant for a while. I will try to find out more in next two weeks. The second opinion was dismissed….do not know why yet. But I will find it out. I had access only to 4 weeks of data. But given how this puzzled me I definitely will not rest till I find out answers to all my questions: why the guy called it bearing defect? Why the motor was removed from service so quickly, seems without giving second thought. Why the second opinion was simply ignored? Who made the decision? And last, probably the most important one – what, if anything, is wrong with the bearing?

I will definitely keep you updated on this one..

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

Kris,
I would guess that your "other guy" was operating under the assumption that everything non-synchronous is a bearing fault. You said that motor speed is 550 rpm, but the spectrum that you posted has it at 1150 rpm. Maybe the other guy thought it was a higher frequency from fluting.You would probably need to see his data to get a better idea of what he was thinking.

David Eason

down under we operate at 50Hz, I often see 300Hz on our DC motors, (6Xline frequency, equates to 360 for you guys) the levels change but its always there, something to do with how the power is processed in the switch room, I usually dont worry about it unless there is asignificant increase.

An update.

I had an opportunity to see both reports and spectral data. They do look identical! The one that called for bearing replacement has a note (from Csi software): 99% non synchronous energy….there is a good possibility that I will have access or get feedback on the bearing condition. I will keep you updated.

Kris,

The use of CSI's automatic analysis tells me more about the analyst than the analysis.

This looks more and more like a bad call.

Keep us posted.

Kris,

12 years ago when I was working for a Petrochemical plant in the Far East, I had this 6X-SCR Firing Frequency show up; during routine monitoring. This was a 800 KW DC Motor, which consequently failed after 4 weeks, with increased amplitudes and distinct "Humming" noise coming off the Motor. In this case, the bearings were OK however, the rotor, stator and commutator were damaged; with all the brushes turning into powder. Fortunately for me, I had made the call to get OEM's and Contractor's attention well ahead of time; who chose not pay attention saying that the 6X is inherent of the system....$12 Million loss in production and machine repairs...not to mention the loss of a job of a key personnel.

Regards....Rajan Muthukrishnan

RajSha

You observe the fault develop and increase over time, suddenly show up, or was inherent at the same level during your watch?

Thanks
Alan