Took a current spectrum on a 6MW 4 pole motor,Lf=50Hz motor speed was about 1461 RPM.expected the sidebands to be at pole pass frequency (0.65 x 4= 2.6Hz) but the plot shows sidebands at twice the value (5.2 Hz). Any ideas on the above.Load was only 40% which is not sufficient for accurate data.

Thanks

Motor_Current.doc (24 Kb, 56 downloads) Plot

I agree that is a weird place for sidebands.

If you had sidebands that are a weird frequency (but not exactly twice pole pass):

One possibility would be oscillations of the driven equipment (for example control valve oscillating on discharge of the pump or fan).

Another possibility might be load change between the time you took your speed measurement and the time you did your current signature.

Either of the above would be unlikely to produce EXACTLY twice pole pass (that would be quite a coincidence).

Is it exactly twice pole pass?

Sync Speed = 1500 * 50.04/50 = 1501.2
Slip Speed = 1501.2 - 1461 = 40.2
Pole Pass Frequency = 40.2 * 4 = 160.8 cpm = 2.68hz
Twice pole pass frequency = 5.36 hz.
But your sidebands are spaced at 5.19 and 5.20 hz.

On the other hand if you repeat the equation with speed=1462 you get pretty darned close to 5.2hz so I suspect 1 cpm error in speed is not unreasonable and certainly could be exactly twice line frequency.

In the event that it is exactly twice pole pass frequency, the only thing that comes to mind is it may be one of those weird situations where the rotor spider is creating an assymetry in the rotor field. It seems like in theory if you had an 8 legged rotor spider in a 4-pole field you could get twice pole pass frequency occuring.

Inspection to determine the number of rotor spider legs wouldn't be too difficult if the motor is tagged out for other reasons. For horizontal sleeve bearing motor, pulling the end covers should be easy. For vertical motors there is often a way to see through the air intakes to the rotor either directly or with a goose-neck boroscope (after all if you follow the air intake passage it will eventually lead you to the top or bottom of the rotor because that's where the rotor draws it's air from).

Then again if you don't find what you expected (something other than 8 spider arms), you are left with a lot of questions: "You thought it was important enough to investigate, so what now?" Been there/done that very recently in another situation. My lesson is to be careful what I ask for.

My situation was that in investigating a weird vibration problem on a 200 hp horizontal 4-pole motor with 6313 ball bearings, the team came up with a bunch of suggested checks including uncoupled run and alignment check. I suggested offhandedly while the motor was down, we might as well check the shaft runout and end-play of the motor shaft extension. Results came back 45 mils endplay when pushing on the shaft extension. I was surprised because I assume all motors of that style should have a fixed bearing on the inboard and floating with wavey washer on the outboard and I expected the endplay to be about the same as bearing axial movement from inboard bearing internal clearance which should be maybe 5 mils max. What followed was a frantic evaluation to decide whether we needed to go after the bearings. Vibration didn't show any sign of bearing problem which held me back from suggesting that. It might have been easy enough to pull the endbells and look but the environment was not conducive to that. After a few hours of investigation while the plant waited for an answer (naturally, this happened to be a very high profile critical job where everyone up to the top was waiting for a recommendation), I found out from the OEM this particular motor for some reason has fixed bearing on outboard and floating on inboard. My response to the plant was a sheepish "never mind" (spoken like Rosanne, Rosanna-Dana" from Saturday night live.

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

Thanks Pete
The RPM measurement was made at a different time compared to the current signature and the sidebands might not be 2X pole pass (i thought it was pretty close). If the sidebands at 5.2 Hz are really the pole pass sidebands then the motor speed would be around 1422 RPM which i think is very low at 40% load. i remember a year back i had taken a current spectrum on the same motor and the 5.2 Hz sidebands were also present. This motor seems to have devoloped a high vibration problem now the details of which i am not aware of, my part is only to analyze the current spectrum to see wheter it is normal.Again at 40% load i dont think i can see much.

Viki

I would say that even the 1461 RPM seems to be incredibly low. This speed would translate into 1753 RPM for a comparable 60 Hz motor. Unless it is a special high slip motor, the speed is way too low for 40% load. One would expect such a low speed with only few bars left on the rotor, yet the decibell ratio is not that bad.
The reason for the sidebands seems to be load related as Pete suggested. The event that modulates the load current must happen twice a revolution. But the question remains: Why such a high slip???

jank

Hmm, thinking about it little further, the frequency of the current modulating "event" should be higher than 2x a revolution. But no matter how hard I try at this time, I can't figure out how many times. I guess I have to sleep on it. It looks almost as a vane pass frequency. The current signature from a piston compressor would be a good lead.

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

I'd say the frequency of the torque-modulating event it it exists would have to be much lower... around pole pass frequency (once per 10 revolutions). But also the coincidence of being very close to twice pole pass frequency would argue against this scenario.

Jan - that's a great observation about the slip... I missed that one. If the slip is 39 rpm at 40%, then it would be 97.5 rpm at full load. That is 6.5% full-load slip which would be unheard of for a large motor (NEMA limits to 5% full load slip for NEMA-frame design B motors - does not apply to large motor, but large motors typically have much lower slip).

The high slip observation is every bit as significant an observation as the unusual sideband (maybe more significant considering we understand more about the expected behavior of slip than we do about the weird sideband). It is worth following up what could be causeing the high slip. Low or unbalanced voltage can cause high slip. High load can cause high slip, but I doubt this high.

Also open rotor bars can cause high slip, which is ominous considering this weird twice slip frequency observation. Now brings to mind another possibility... two open bars spaced at an exact distance of either 45 degrees, 135 degrees, 225 degrees, or 315 degrees from each other (K + 1/2 pole pitches for 4-pole motor where K is an integer)... would seem like quite a coincidence considering that typically the bars adjacent to the first broken bar deteriorate first, but you never know. If you rule out the other possibilities for high slip than I would say rotor bars are definitely suspect.

You could attempt to rule out unbalanced voltage fairly easily be comparing the currents with clampon. You could also attempt to rule out low voltage by looking at the associated bus voltage and considering effects of length of wire running from the bus to the motor.

Another smaller possibility would be as Viki mentioned speed was taken at a different time than the signature and perhaps load was higher at that time. But even if it were at full load, 1461 sounds like it would be too high a slip for a large motor such as this. Do you have the nameplate RPM handy?

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

My theory of load variation does not look too good. With the data provided by Viki, the load would have to change rougly every 5th revolution (calculated 4.68x). I think we can rule that out.
jank

Viki:

In Current Signature Analysis (MCSA), sidebands around line frequency will relate to two types of issues: 1) Rotor bars; and, 2) Problems from the rotor out through the load that would be seen in the low frequency spectrum.

Normally, if this were a fan or pump, if I had an alignment or impellor/fan blade problem that was severe enough, they would show in the sidebands of line frequency. If you have the ability to demodulate your spectrum, see how this signature shows and, in linear mode (current vs. dB), see if there are harmonics of this frequency at any level. Rotor bars will not show harmonics in current.

Based upon the severity of this signature, you can confirm whether you are looking at rotor bar problems, or not, by taking a look at the current readings from your analog display on the motor control. If there is a regular, sharp, ticking motion of a few amps (more indicates greater severity), then you are having a rotor bar problem.

Another option is to use an inductive based test for rotor testing during a future shutdown. If this is a rotor bar issue, you will want to take care of it as soon as possible based upon the level which, if taken from the low voltage side of the control, is dampened a bit, would indicate a severe condition. If it is the fan or pump impellor, then you will want to determine the cause.

Are you showing signs of a potential rotor problem in normal vibration test results?

Howard

Did you say what this motor is driving?

Sorry for the late reply, was away for a while.Could not get any new data.Now i have doubts about the rpm value it was measured by someone else i have to check it up again.The name plate says 1487 RPM and motor is driving an extruder posting the same spectrum with expanding the amplitude scale maybe more useful will try to get a demod spectrum soon.

motor_spec_3.doc (24 Kb, 28 downloads) Motor spec