Sign up Calendar Latest Topics Donate

  Author   Comment  

Sr. Member
Posts: 155
Reply with quote  #1 

Does anyone know if it's possible to see the belt frequency in demodulation? I have belt drive Fan that has a harmonic family that seems to relate to a sub run speed frequency that could be either the Belt Frequency or a bearing cage defect. I can see the same harmonic family in demodulation so does that discount the belt frequency as the source?

Walt Strong

Sr. Member
Posts: 889
Reply with quote  #2 
Did you calculate Belt Frequency based on dimensions and shaft speed? Did you use a Strobe light to visually inspect belt and measure belt speed (Belt Frequency) and to verify both shaft speeds?  Did you try airborne ultrasound to confirm belt defect? I would want to be accurate on evaluating belt condition before scheduling and replacing a bearing! Yes, I have seen Belt Frequency with PeakVue (Demodulation) and confirmed with other mentioned techniques.

Ron Brook

Posts: 35
Reply with quote  #3 
Demod is a highly localized measurement.
If the belt is bad, you should see it on both the drive and driven machines and it would be in the velocity spectrum.
If you see this freq only on one end of the drive or driven machine in demod, it is most likely a cage freq.
But seeing as how you have not provided any machine info, data, etc, this is all just for talking purposes.

Sr. Member
Posts: 163
Reply with quote  #4 
I hesitated to post since I could not offer anything useful (maybe still can't? [wink]) as I never heard of demod being useful for belt frequency identification before.

The demod process uses an HP filter to help remove the high level/low frequency content to help uncover low-level harmonic content typical of bearing degradation (sharp impacting). I have never read of this technique used for belt issues which I suspect do not have an "impacting" nature in the signal.....BTICBW.

Jim P

Posts: 63
Reply with quote  #5 
Unless the brg freq & cage are exactly the same, you should be able to take a reading with high enough resolution to separate the peaks. You should be able to do that in velocity.
Ralph Stewart

Avatar / Picture

Sr. Member
Posts: 235
Reply with quote  #6 
I sort of doubt that a belt frequency would show in the demod data, except under unusual circumstances. Like modulating slotpass.

I would first find the carrier of the demod data and then determine what and where it is with an acceleration conventional data set taken with an fmax 1k or 2k higher than my demod filter, unless you are using a band width filter. That would take a different approach.

This should show that the demod data is actually coming from the modulation of the carrier and might show that it is probably cage.

Just my thoughts without more info to use. [smile]

Thanks and Have a Great Day,

Ralph Stewart
Previous Topic | Next Topic

Quick Navigation:

Easily create a Forum Website with Website Toolbox.