The Keyphasor
The Keyphasor
The Keyphasor
which provides a large change in gap in front of the proximity probe (Figure 2). Since the proximity
probe is a gap to voltage transducer, when the keyway passes the probe a voltage pulse
results. Since the keyway passes the probe once-per-revolution of the shaft, the Keyphasor pulse
occurs at a frequency equal to the speed of the machine. Therefore, its frequency is the same as
the running speed of the machine.
To properly understand the full operation of a Keyphasor, it is first necessary to understand the
waveform or orbital pattern on an oscilloscope (Figure 3). When vibration probe inputs are put into
a dual channel oscilloscope and observed on the cathode ray tube, the dynamic motion (vibration)
may be displayed in time base or orbital form. A waveform or orbit pattern on an oscilloscope is
simply a dot moving very rapidly such that to the eye it appears as a continuous line. The rapidly
The Keyphasor (shaft reference) technique tells where the rotor High Spot is at a particular instant
in time when used with a phase measuring instrument (Figure 4). If we define the phase angle as
the number of degrees from the Keyphasor pulse to the following positive peak of dynamic
motion (vibration) as sensed by a vibration probe, the use of the shaft reference technique allows
the direct phase angle readout to define the High Spot of the rotor system. The logic of this method
of phase measurement is as follows: the
Keyphasor pulse occurs the instant the Keyphasor transducer and the shaft reference (notch or
projection) are aligned. The positive vibration peak occurs "0" number of degrees (phase angle)
later in time as the shaft High Spot passes beneath the vibration input probe. The machine
rotor may then be stopped and rotated to align the Keyphasor transducer and its shaft reference. If
the rotor is then rotated in the direction of rotor rotation the number of degrees as defined by the
phase angle, the High Spot of the rotor will lie
directly under the vibration probe.
In a classical rotor dynamics situation, below the first balance resonance (critical) the phase angle
as referenced by the Keyphasor provides an accurate indication of the High Spot on the rotor when
the keyway is aligned with the Keyphasor probe. Well below the first balance resonance this High
Spot indicates the angular location of the residual unbalance on the rotor (Heavy Spot). Assuming
no thermal or permanent bows of the rotor, the trial weight could be added 1 80 from this angular
location and a finer degree of balance would be obtained. As the rotor increases in speed through
its balance resonance region an amplitude increase occurs with a corresponding 180 phase shift
(lag) of the High Spot from the Heavy Spot. Well above the first balance resonance the amplitude
has decreased and the full 1 80 phase shift has occurred. Since the 1 80 phase shift affects the
relationship of the High Spot on the rotor with respect to the unbalance, the balancing rules of the
Keyphasor change accordingly. Above the first balance resonance the phase angle, as referenced
by the Keyphasor, defines the location of the High Spot on the rotor where weight must be added
to obtain a finer balance (Figure 4). The phase angle as defined utilizing the shaft reference
(Keyphasor) technique always defines the High Spot on the rotor. With a fundamental knowledge
of rotor dynamics it is possible to more easily and more directly balance rotors without trial and
error techniques.
In a working machine an exact 1 80 phase shift may or may not be seen on Bode' plots. Every
POLAR PLOT
The Keyphasor provides a vital signal for techniques used to plot rotor unbalance response. Two
widely used techniques today are the Polar plot and the Bode' plot, which require primary
measurements of amplitude, rpm, and phase angle. The Keyphasor provides the rpm signal and a
reference pulse for phase angle measurements when a Vector Filter Phase Meter with 1 X rpm
filter is used. The Polar plot is a plot of the phasor (rotating vector) of the amplitude of rotative
speed (or forcing function speed) and phase angle as a function of speed. On rotating machinery,
the Polar plot is the amplitude and phase reading from a transducer showing the response of the
machine to its residual or deliberate unbalance as a function of speed. Mass unbalance manifests
itself as a phasor. The magnitude (p-p mils) and the direction (phase angle) of the resultant
motion with respect to a known reference point (Keyphasor) can be described on a Polar plot,
which displays the in-phase and quadrature components of the phasor. These component values
will change as the magnitude and direction of the phasor quantity changes with speed. A classic
Polar plot of a true balanced rotor will look like this:
BODE' PLOT
The Keyphasor pulse also allows a plot of rotative speed amplitude ( 1 X) of a given measurement
against rpm, along with the phase lag angle of that amplitude phasor against rpm. The Bode' plot
provides this relationship and is most useful in showing the rpm of various resonances of
a machine. The rotative speed absolute amplitude must be nulled with rotor speed below each
resonant rpm. Failure to do so always yields a "Brody" (false Bode').
The DVF 2 offers the capability of zero nulling for display of proper Bode' plots. Bode' plots may be
plotted directly from the DVF 2 with the use of an X-Y-Y plotter.
In attempting to utilize the Keyphasor and an oscilloscope for balancing on machines where a high
mechanical or electrical runout (glitch) level is present, it is sometimes necessary to filter the high
frequency glitch from the waveform before making the phase angle measurement (Figure 5). In
these instances it is extremely important that any phase angle shift of the electronic filters being
utilized be taken into account. If two matched tunable filters are utilized with both of them being
exactly tuned to 1 X rpm, zero phase
angle shift from the instrument can result. In this case, the 1 X rpm running speed frequency of a
machine vibration is obtained from the filters and input into an oscilloscope in conjunction with a
Keyphasor pulse input to the Z intensity axis of the oscilloscope. This provides 1 X rpm filtered
orbits with a phase angle reference. Dual low pass filtering is a successful means of observing
orbital patterns with relatively high glitch levels. However, caution must be used since phase angle
can be shifted by low pass filtering depending upon the filter cutoff frequencies. Utilizing "auto-
tuning" filters like the Digital Vector Filter Phase Meter eliminates the
problem of instrument phase distortion. The Digital Runout Compensator can also be used to
minimize problems associated with glitch. The DVF 2 also has the ability to null out glitch vectors in
addition to any bow vectors which may be present.
Besides the Keyphasor' s ability to define phase angle and help define natural rotor response, the
Keyphasor lends itself easily to determining whether a vibration motion is ( 1 ) at rotative speed or
an exact ratio thereof, or (2) is at a rate not related to rotative speed. If a motion that is an exact
ratio of rotative speed, such as %, 1, 2, etc., is present the Keyphasor dot will appear to be locked
onto the waveform of the vibration as displayed on an oscilloscope. If a nonsynchronous waveform
is present, the Keyphasor will appear to move over the waveform of the vibration. In many rotor
rub situations the Keyphasor will be unstable or jittery. This ability of the Keyphasor can be very
important in observation of whirl phenomena occurring below 50% X rpm.
The Keyphasor proximity probe should be installed observing a notch or projection of sufficient size
to create a large voltage pulse Figure 2). It is not desirable to simply put a scratch mark on the
shaft and expect the proximity probe to provide a large pulse. This pulse should be of
sufficient magnitude (nominally greater than 5 volts peak-to-peak) to trigger the Z intensity axis of
an oscilloscope and to trigger the time base sweep of the oscilloscope, as well as to trigger a digital
tachometer easily without fear of triggering the
tachometer on noise.
When the Keyphasor pulse is input to the Z intensity axis on an oscilloscope, it provides a
reference spot on the dynamic waveform and/or orbital pattern. This appears as a bright spot and a
blank. By determining which of these occurs first in time, in the time base mode of the
oscilloscope, it is possible to determine the direction of the dynamic motion of the shaft on the
orbital pattern. Also since the Keyphasor occurs at a frequency of once-per-revolution, we can
determine basic frequencies present in a vibration waveform on an oscilloscope by simply
counting the number of cycles of vibration per Keyphasor pulse. In complex waveforms, the basic
frequencies can be determined by counting the number of peaks that occur between Keyphasor
pulses.
Assuming there is no gear in a particular machine train, it isnecessary to have only one Keyphasor
It is desirable to have the Keyphasor probe installed on the driver of the machine train since it is not
uncommon to run the driver uncoupled from the load (for example, steam turbine drive). If the
Keyphasor probe was located on the driven equipment, then the uncoupled run would be without
the Keyphasor reference pulse.
SYSTEM INSTALLATION