Product data

www.assetweb.com5
Machinery Health
™
Management October 2014
High Resolution Spectrum
The rst compression technique is FFT analysis, which
transforms the vibration waveform into a frequency
spectrum (shown below). This spectrum not only reduces
the le size by over 60%, it also presents the frequency
information in a more readable format. The smaller data
set accelerates responsiveness of the system while reducing
power consumption.
Thumbnail Spectrum
The thumbnail spectrum is derived from the same
waveform data. It contains the same frequency and amplitude
information as the high resolution spectrum, but the data set
has been compressed by an additional 98%. Now it is small
enough to transfer over the network in less than a second.
Energy Bands
As a nal data compression technique, the CSI 9420
divides the spectrum into three predetermined energy bands
(as shown below). It then calculates the vibration energy within
each energy band and passes these values to AMS Machinery
Manager for trending and alerts.
The elevated peaks are still clearly visible in the thumbnail spectrum
and indicate the presence of mechanical looseness.
Energy bands with frequency ranges. Optimized for a 4-pole motor
running between 1500 and 1800 RPM. Note: bands are xed.
Band Fault Types
Range
1
Rotor Vibration: Imbalance,
misalignment (also defects on
belt drives)
2–65 Hz
2
Rotor Harmonics: Looseness,
electrical faults, blade and
vane pass
65–300 Hz
3
High Frequency: Bearing and gear
defects, lubrication and cavitation
300–1000 Hz
By dividing the spectrum into energy bands, we can isolate
frequencies associated with different categories of faults.
The elevated peaks in this high resolution spectrum provide a clear
indication of mechanical looseness on the machine.