Is it possible to average out synchronous vibration?

[Vibrationenthusiast]

Dear all,

I am new to the vibration analysis field. I am curious to know, is it possible to average "out" synchronous vibration data by using an analyser?
I know that synchronous time averaging with the help of a tracking filter helps in eliminating noise. Is the reverse possible? Can I use a tracking filter (key phasor or reflective tape laser sensor) to remove the synchronous components from a signal? In a hypothetical scenario where the GMF and any other high frequency vibration i.e. BPFI & harmonics or vane pass frequencies lie close to one another, will removing the synchronous components from the signal be helpful in identifying/emphasizing the non-synchronous bearing fault frequencies?

 

[Tmoose]

In the old TEC software of the 1980s it was possible to create spectra that were the difference between 2 spectra. I used it mostly for bump or impact tests.
I have no idea if modern software packages offer that manipulation.

 

[Strong]

There may be more than one way to do what you want. A Google search will bring up a lot of methods, some usefull and sone not. One way is to demodulate a bandpass filtered signal centered between synchronous harmonics, such as described here:

http://www.vibrotek.com/articles/

You could also search for Emerson-CSI PeakVue as another related method.

Walt

 

[IRstuff]

That wouldn't be much different than AM demodulation, so I think the answer is yes.

TTFN
faq731-376

 

[GregLocock]

I'll add to the snowball. Yes, it is possible, it is not however especially easy in the real world to totally get rid of a strong synchronous signal (if it was we wouldn't worry about ground loops and mains hum).

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?