Hi,
Does anyone know why current slip pole amplitude different between each phase?
Attached is current data taken 4200 Kw motor driving a screw mixer.

Y060127_4X-201_Current_Data.xls (18 Kb, 91 downloads)

I'm stumped by your question. Some rambling thoughts.

Those are higher sidebands than I have ever seen (even the one you call ok). Are you positive they are at pole pass frequency and not associated with the load? Do you suspect dynamic eccentricity or rotor problem? Time dependent increase after startup?

We typically only take one phase... the phase that goes to the ammeter. Would have to find another point (perhaps ct secondary input to relays) to get three phases.

I don't know why they would be a difference among phases.

I was trying to imagine if static eccentricity would make a difference. If rotor were near one phase than the ohters, then passing of broken bar, or of dynamic eccentric rotor, would have a bigger effect at the closest point. I might have believed that explanation with 2-pole motor, but 6 pole motor, 18 pole phase groups (3*6) spaced 20 degrees apart (360/18), and overlapping with other phase groups, I have a hard time imagining that much variation.

Thre are various errors associated with the clamp-on measurement. But since the 50hz is so close in frequency to the sidebands... and your 50hz component not affected, I have a hard time accepting measurement errors.

Another guess. perhaps the pole pass characteristic is changing over time for some reaonand you just happened to catch it at different part of it's variation on different phases. (although neither 50hz current magnitude nor side-band frequency changed between readings). I'm not sure what would make it change but stranger things have happened. **At any rate, you could check this possibility by going back for another check to see if you have the same pattern.**

Out of curiosity, what was the measurement setup Fmax, lines, averages, window?

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

I see now you also have speed recorded and sideband frequency is in fact pole pass frequency.

If this were a very severe rotor bar problem, we would expect slip to increase above expected and speed to decrease. 985rpm doesn't sound low, but we don't have enough data to judge it.

Do you have
1 - the exact nameplate speed?
2 - either the full-load-amperage or the voltage?
3 - optional - any idea of measured (vs nomrinal) voltage at a nearby bus and how far from bus to motor?

Your motor is running incredibly slow. Spinning at 985 rpm (rated 990) with only 355 amps out of 425 is not normal. You probably do not have too many bars left.
If the motor is in Delta connection, try analysis of the phase currents, not the line currents. There is probably something circulating in the delta.
jank

The 990 rpm was the nameplate speed I was looking for. I didn't see it on the spreadsheet. Also didn't see FLA.

I agree the low speed does tend to suggest broken bars. And when you combine that with those huge pole pass sidebands, the conclusion appears pretty close to certain.

To cover all the bases, it would help to verify that line frequency was not slightly low and voltage was not dramatically low. These could also cause the low rpm (although not the pole pass sidebands).

Once again I agree the very likely conclusion is what jank said.

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

This is refreshing.
I'd try to retake data with better resolution.
Note that attained resolution:

Resolution =(Fmax/Nlines)*2*WindowFactor

Current data was taken with SFK CMVA 60.
Fmax: 10Hz (45Hz to 55Hz)
No. Lines: 800
Window: Hanning
Will try to collect data again after Chinese New Year. We are close till 01/02/06 but motor still running.

Adam,

Looks like resolution is sufficient.

Question: Is data short term repeatable for each phase?

I suggest re-taking data with 8-12 averages, 50% overlap.

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

Gentlemen:

With MCSA, PPF sidebands (twice slip frequency) are rotor bars.

2(((1000-985)/1000)*50Hz)= 1.5Hz

There are no other issues that look the same. Rotor eccentricity signatures are much different.

The values of dB can only be the result of resolution and the filter.

See attached. This is a guide that I had developed (the AC portion) in 2004. Rotor frequencies and limits are on pages 9 and 10.

Additional details on MCSA and ESA interpretation can be found in my book "Motor Circuit Analysis: Electrical Motor Diagnostics" and downloads in http://www.motordiagnostics.com/presentations.htm

Howard

ATPOL_Desk_Guide_and_Pattern_Recognition_Manual_AC_only.pdf (652 Kb, 57 downloads)

quote:

Originally posted by MotorDoc:
Gentlemen:

With MCSA, PPF sidebands (twice slip frequency) are rotor bars.

2(((1000-985)/1000)*50Hz)= 1.5Hz

There are no other issues that look the same. Rotor eccentricity signatures are much different.

Doc - for this particular case I agree it's an easy call of rotor bar problem with the confirmation based on slip much higher than expected for the loading level as Jan pointed out.

Talking in general, I was under the impression that dynamic eccentricity can also produce pole pass sidebands around line frequency in the current spectrum. Correct or not?

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

Adam,

I don't know why the db amplitudes are different in each phase, but in my experience 41db is equal to one broken rotor bar and yours are all less than that suggesting severe rotor problems as verified by the actual slow RPM and low current draw versus the nameplate data at full load. I've experienced a 37db rotor problem that turned out to be a crack in the end ring! (Assuming the nameplate data is correct)

Do you know if this is a cast aluminum rotor or a copper rotor motor?

What instrument are you using to measure the pole pass frequency with? (Perhaps the CT ratios are different)

Please keep us posted with pictures of this rotor when they become available!

The Doc is right on with his All Test manual which explains the formulas for pole pass frequency, both eccentricities, stator mechanical and electrical as I have a copy of it at home and work for reference and use them often. Framatome also has an excellent manual that explains these issues. There are also a couple of articles under the "motor" section on ReliabilityWeb.com written by Aditya Korda and Prof. Bill Thompson about motor problems including rotor bar case studies along with some others.

The worst case rotor I've seen was on a small 25 HP 460 VAC 3 phase motor with a poured aluminum rotor that had the shorting ring seperated almost half way around the rotor from the rotor bars and it still ran! Of course it had no power because it couldn't produce the torque needed, didn't trip the overloads nor blew any fuses.

Doc,
Could the different in the values of the dB cause by other fault and masking it as rotor bars problems. This motor is wound rotor motor with steplesss autotransformer starter. Could bad contact (high resistance) on the slip rings or other component cause this differences in the dB?

quote:

Originally posted by adam:
This motor is wound rotor motor...

I was assuming squirrel-cage-rotor motor, as I believe most of the others were. Strike all my comments (except my question for Doc).

There are no rotor bars in a wound rotor motor. The pole pass sidebands may suggest some unbalance in the rotor circuit as you suggest. I have no idea what normal levels are for this test on wound rotor motor.

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

quote:

... motor is wound rotor motor with steplesss autotransformer...

The above explains it all and the observed pattern points to the autotransformer fault.

[/QUOTE]Originally posted by David_G:
the observed pattern points to the autotransformer fault.[/QUOTE]

David-Can you explain more indepth about autotransformet fault in wound rotor motor.
Are motor current analysis for squirrel cage motor only?
What other on-line test can be carry out on wound rotor motor?

Thanks

Sorry guys, been traveling. This week it is all RCFA's on utility machines. Everything comes in groups!

Actually, now that I know what the machine is (wound rotor motor), I would suspect the following:

1. Bad brushes or slip ring surface (check first);
2. Shorts in wound rotor (need MCA to check); or,
3. Concur with resistor bank issue.

You can verify the condition of the rotor circuit by measuring phase to phase voltage or current from the rotor circuit.

Howard

ElectricPete:

No. The issues that can occur around the line frequency include rotor bars (Must be sidebands of twice slip frequency)and faults external to the motor (torsional issues, such as looseness, broken gear teeth, etc.).

I think the best thing for me to do is to write up a short piece on the physics behind MCSA as it relates to detecting faults (ie: dynamic and static eccentricity). I have a study on 'The Skilled Workforce in the 21st Century' that I am doing with Terry O'Hanlon to complete, first.

I know Aditya, Bill and Don. Don and Aditya use the same type of system that I do, so the materials are the same, up to a point. My work on synchronous machines, DC machines, machine tools and generators is unique. They also deal with ESA - Electrical Signature Analysis - which is a combination of Voltage and Current Signature Analysis. This type of system was developed by Howard Haynes of Oak Ridge National Labs. Aditya has also attended my classes on Motor Diagnostics.

Bill Thompson has developed his system seperately and his algorithms drive the Iris Power Engineering MCSA tool and a special website that he runs.

All three have joined in with the Institute of Electrical Motor Diagnostics. IEMD is working with ISO on the development of the MCSA standard and I chair the working group, as of this year.

There is growing university, EPRI and IEEE interest in MCSA/ESA and papers are being published at a fairly good clip.

Howard

In the past I had always thought pole pass sidebands around LF in current spectrum could be either rotor bar problems or eccentricity... and further confirmation (such as thermal dependence, high slip) was required to separate the two.

After looking a little further I now think Doc and Jim are correct. It is only a symptom of rotor problems, not dynamic eccentricity. Thanks for setting me straight on that.

I have a pretty big collection of articles giving case studies on current signature analysis. I searched and couldn't find any reference to pole pass sidebands around LF in current spectrum related to dynamic eccentricity.

I did find a few older articles that state that the spectrum of the AIRGAP FLUX has pole-pass sidebands around LF both for dynamic eccentricity and for broken bars. One such article is "Noninvasive Detection of Broken Rotor Bars in Operating Induction Motors" by Stein, Endicott, and Madden, IEEE Transactions on Energy Conversion Vol 3, No 4, December 1988.

I think this is where I got my impression that dynamic eccentricity can cause pole pass sidebands in current spectrum. I assumed current spectrum would have similar behavior as they described for flux spectrum. Apparently not.

Lots of case studies show the pole pass sidebands in current as indication of rotor bars. None of the case studies I looked through show it as an indication of eccentricity (other frequencies tell that).

A revealing case study was in "25 Years of Troubleshooting" by Baxter, Part III, Case #10, he describes a rotor that was modified and was operating at it's first critical speed with a heavy bow (dynamic eccentricity). Heavy pole pass sidebands showed up around running speed in vibration spectrum, but pole pass sidebands around LF in the current spectrum were almost nonexistent (65db below LF). He didn't show any current spectra for sidebands of rbpf.

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

Adam,

You have started an excellent discussion here! Please let us know what you actually find wrong with this motor as soon as you nail the problem down.

Unfortunately, I have very little experience on testing wound rotor motors and have done only limited testing on them as ours are, for the most part, on overhead cranes and not easily accessable.

We have had a repetative problem with one hoist motor that likes to break off one of the three wound rotor lead lugs on the winding side of the slip ring. This open rotor circuit causes an audible noise when the motor tries to run. The motor control is called a static stepless setup.

So far the repair shop has tried heavier lugs and more lacing to help hold them in place.

By all means, after fixing this problem, take another online current signature of this motor and see if those line frequency sidebands are still there! I'm really curious as to the outcome. I'd also like to see this in the Explorer's torque spectrum. Does anyone out there have one they could share with us?

Can I detect the problem on the wound rotor motor by doing ESA using Framatome Empath 2000 system. Does anyone have experience using ESA on wound rotor?