Does the magnitude of Current Signature Analysis sidebands change significantly with load?

Is there any correltaion or correction for sideband magnitude measured at lower loads?

(We operate mostly at constant steady load, so I have not had any opportunity to observe variation with load).

Attached are 2 current spectra for 350 Hp 1780 RPM motor. In the first case the load is 35 Amps (full load), in the second case it is a very low 17 Amps. It can be seen that the load is really important. Not the change in the decibel ratio and the number of sidebands.
jank

sideband_variation_with_load.pdf (199 Kb, 26 downloads) variation of the current signature with load

Thanks, that is interesting. Quite a few less harmonics noticeable and db was 38 vs 35.

I see the label "1 broken bar"... is that what was found?

I will have to search out the data, but yes, the amplitudes do vary. I do remember atleast one case where PPF peaks were down at low load but rose with increase in load. That was with a faulty motor.

Conversely, good motors with low loads can show high PPF amplitudes. This occurs due to insufficient resolution because of which the PPF sidebands get influenced by the line frequency peak. At higher loads, the PPF peaks move sufficiently away to be assessed correctly. This problem is also seen with certain motors that have very low slip. For e.g. there are 6 pole motors here with full load RPM of 995. As these often run at 40-45 % load, I prefer to trend the PPF amplitudes rather than make a call with one measurement.

Regards,

Aditya