Measurement Computing Medallion Rotate rev.2.3 - Manual (Page 37)

October 2000 Medallion Rotate Manual
37
O
RDER
N
ORMALIZATION
Medallion Rotate Plus can resample a time waveform to order normalize the
data when performing a Waterfall analysis. Order normalization cancels the
effect of frequency smearing across spectral bins when the shaft speed is
changing rapidly (high slew rate). Order normalization ensures that order-
related peaks are properly aligned when plotted against the orders.
Order normalization also allows you to use a smaller blocksize when
performing the Waterfall analysis. This can significantly shorten processing
time while maintaining good frequency and amplitude resolution.
A
PPLICATIONS
Order normalization is ideal for identifying order-related frequencies in the
Waterfall plot. Order-related frequencies are located at the running speed of the
component and multiples of the running speed. The running speed is the first
order. This is particularly useful when attempting to separate order-related and
non-order-related peaks in the Waterfall plot. Since the X axis is displayed in
orders, it is easy to pick out the order-related peaks.
I
NPUT
S
IGNAL
R
EQUIREMENTS
Order normalization requires a smoothed speed curve from performing a
Tachometer analysis or using the RPM from Waterfall function. Medallion
Rotate Plus processes the machine speed with the transducer data to create the
order normalized Waterfall analysis. When collecting the non-tachometer data
(vibration, temperature, …), note the following:
• If you want to see low frequency data in the Waterfall analysis, make
sure that the high pass filter setting in your analyzer is not excluding
the desired low frequencies.
• The sample rate in the analyzer must be at least 2.5 times the maximum
frequency of interest to avoid aliasing.
• The choice of window function for analysis depends on the type of
resolution you need (amplitude or frequency resolution). The Hanning
window is a good default choice. The Uniform window is only a good
choice when the orders are integer multiples of the shaft speed (1x, 2x,
3x, …).
Order normalization requires a clean tachometer signal for best results.
Processing the tachometer signal must produce a good spline-fit to provide a
smooth speed curve. Note that the resampling algorithm does not require that
the tachometer pulses correlate with the resampling angle.