Hello all!
What is the difference if any between a bearing "natural frequency"and the bearing defect frequencies. I ask this after reading in an attached file of another post describing the different stages of bearing failure.
Stage two says the appearance of bearing natural frequencies will begin to appear in the 30Kcpm to 120 Kcpm range, but goes on to say at stage three there will be present "bearing defect frequencies".Please advise
Thank you guys, this is a really good forum.

Personally I'd accept this as a poor statement overall. Generally this phenomenon is displayed with lack of lubrication and frequency may be damped or unrelated to the range eluded to.

Bearing natrual frequency is like the bearing being struck like a bell and ringing. Occurs when spalling is present. Bearing defect frequencies and calculated frequencies based on bearing geometry. Example typical outer race frequency is approx. .4 X #rollers x rpm. IR would then be .6 X.... Bearing defects occur equal to the rate that rollers run over defect on that race, or speed of the rollers, or speed of the cage.

Bearing natural freq. would be the first indication of either lack of grease or a small damage within the bearing. My experience with this is that this occurs normally around 5 to 10 khz for typical ball bearings.
As sam sais this could of course also be related to lack of grease, BUT should still be taken as a precaution. Grease could in my opinion on a short term dampen the natural freq, and therefor "hide" the defect. After a short while this freq would start to rise again.

I have seen spectrum plots take in acceleration prove high haystacks in these areas, where the bearing has had a corroded surface and also clear signs of damage in the surface.
On this bearing, the bearing DEFECT freq where not very clear, and of a small amplitude.

An identical motor suffered a serious bearing failure just two days after, with no clear signs of bearing defects. Unfortunatly, i never had the time to take a measurement in the higher freq range on this on.

Defects can and most often excite the natural frequencies, resonances of the rings in the bearings despite the load on them It´s used in bearing conditon systems like the simple broadband filtered values, what we call "g" value HP filtered signal in accel above 2kHz or our own L value, velocity with HP filtering, HFD or whatever or for lo speed things like rolling steel mills, a eddy probe pointed at the surface of the bearing outer ring, when it detects hi freq. signal from the outer ring, it´s the natural freq. that get excited, and you have a brg. damage. Yes in the FFT it may look like the haystack. Olov