Difference between revisions of "NVGate Synchronous Order Analysis"

Jump to navigation Jump to search

NVGate Synchronous Order Analysis (view source)

Revision as of 10:06, 30 April 2020

5,519 bytes added ,  10:06, 30 April 2020
no edit summary
Line 1: Line 1:


Order analysis is a set of specialized measurement techniques often used when making vibration measurements on rotating machine. With the Order Analysis option, you can make RPM profiles, order track and order spectrum.
To make order measurements, you must use a proper tach signal which is synchronized with the rotational speed of the machine.
<font size = "4">'''6.1.2. What is Order Tracking?'''</font>
When doing measurement on a rotating machine, it is often useful to display spectrum in the behavior of harmonics or sub harmonics related with the shaft speed.
If the shaft speed varies, each harmonic of the shaft rate needs to appear at a fixed point into the spectrum, so called order.
The basics to make such analysis are to control the sampling rate of the analyzed signals in order to get an equal number of samples independently of the shaft speed.
The classical way uses a tracking ratio synthesizer based on phased locked loop oscillator which generates a constant number of sampling pulses during one shaft revolution. This synthesizer also controls variable cut off frequency of analog anti-aliasing filters. they are adapted to the variable sampling frequency that is itself depending on the frequency bandwidth.
<font size = "4">'''6.1.3. How the OROS Analyzer Works?'''</font>
'''Basics'''
The OROS analyzer is based on a variable digital resampling technique implemented as software in a digital signal processor chip.
This approach gives an improved performance over analog solutions along with reduced hardware cost and reduced complexity.
The figure below shows a general block diagram of digital resampling technique.
The signals to be analyzed are sampled at a constant rate with fixed antialiasing filters. These filters have a slope greater than 200dB per octave in the transition band.
The tach signal is sampled and processed in order to compute sampling rate of tach revolution pulses. The arrival time of a tach pulse is computed using interpolation between 2 samples. This tach processing also computes instantaneous speed in order to adjust the cut off frequency of the low pass digital filter used in resampling.
----
----
----
'''ORDER TRACKING ANALYZER'''
'''OUTLINES'''
<font color="#FF0000">''Figure 6-1. ''</font>''Block diagram of digital order analyzer''
After low pass digital filtering, the analyzed signals are resampled and a standard FFT analysis is done.
'''Resampling date computation'''
The tach processing first measures date arrival of tach pulses and next computes resampling dates as shown in the figure below:
[[Image:SOA_01.png|framed|none]]
For a revolution, resampling dates are computed taking into account current rev duration, next one and both preceding ones.
A function of sampling rate versus revolution or shaft position is evaluated and for the current revolution resampling rates are computed for equally spaced shaft positions.
If the number of tach pulses per revolution is not an integer, the tach processing calculates by interpolation the new dates of the tach pulses in order to always have an integer number of revolution for the resampling process.
[[Image:SOA_02.png|framed|none]]
----
----
'''ORDER TRACKING ANALYZER'''
'''OUTLINES'''
So using backward and forward tach information gives an accurate evaluation of resampling dates. But this method needs to store signals to be analyzed so that when a tach pulse arrives, the analyzer computes dates of the previous revolution (use of digital signal processors).
This method makes no basic assumption about shaft speed variation profile and compared to other methods, it always gives exact resampling dates at the start and the end of a revolution.
So in case of a large speed variation in a revolution, the analyzer can always deliver valuable spectrums. During implementation the operator can adjust the speed variation threshold over which measurements are automatically rejected.
'''Signal resampling'''
The digital resampling of analyzed signals needs a complementary antialiasing filter whose cut off is continuously adjusted to a frequency equal to the product of the maximum analyzed order by instantaneous shaft speed.
The resampling also needs X signal interpolation between samples. The variable filtering and interpolation is based on an OROS proprietary wich is the most accurate and powerful on the market.
Compared to other methods, the OROS one has the main advantage to be able to work with the full available frequency range of the original sampled signal, so the OROS order tracking analyzer is able to work with signals up to 40 kHz with initial sampling at 102.4 kHz.
'''Signal analysis'''
The signal analyzer does FFT processing on resampled signal.
This signal is always taken by starting synchronously on a tach revolution pulse in order to be able to deliver absolute phase information of each order.
Sub-ranging order analysis is available with order resolution down to 1/32. In this case the FFT processing is done by using signal corresponding to N consecutive shaft revolutions and so gives a 1/N order resolution.
The time and spectral averaging are implemented. Time averaging is necessary when absolute phase resolution and/or rejection of signals unrelated to shaft rotation are needed.
Spectral averaging can compute power spectrums and cross power spectrums with associated results. In this mode, weighting windows are available in order to reduce side lobe effects of signals not related to shaft rotation.


====Synch order====
====Synch order====
Retrieved from "https://wiki.oros.com/index.php?title=Special:MobileDiff/3061"

Navigation menu