Phase Analysis with CTC's SB152

This blog post was written by Jake Ford, PFE Limited

Phase analysis is an important analytical technique used by many vibration analysts for various applications. Put simply, phase is the lag between one signal to another; these two signals could be a tachometer with an accelerometer, two accelerometer signals, or a key phasor with a proximity probe (often two to form an orbit).

Phase analysis capabilities remain one of the key advantages of wired accelerometer applications over wireless, as there is little to no transmission lag between two sensors wired, in comparison to wireless. Most often low frequencies are used in phase analysis, for example in cross-phase, run up-coast down, orbit analysis, and more. By analyzing the timing of the vibration with a reference, faults can be better detected, or a better degree of certainty attained from the analysis results.

To determine phase, two signals must be present with one of the signals acting as a trigger to time the lag to the other signal. In a tachometer with an accelerometer this would be a square wave for the trigger waveform and a sinusoidal waveform for the (filtered) accelerometer frequency:

chart depicting a square waveform overlaid with sinusoidal of the same delta frequency with an offset phase

This method of phase is often used for run-up/coast-down and balancing procedures and is usually referred to as ‘absolute’ phase as it is measuring the phase lag from an absolute position on the shaft (reflective tape for example) and timing the accelerometer signal against this reference.

For balancing this is useful for calculating the vectors required to counter an imbalance, which we won’t go into too much detail on. For RU/CD, this allows us to monitor the phase as a machine runs up or coasts down; if a machine runs through a resonance there will be a specific phase response to this resonance which is 90° into the resonance (typically would be the highest amplitude), and a further 90° out of resonance for a total of a 180° phase shift.

Alternatively, a method often called ‘relative phase’ is used whereas one accelerometer is used as the trigger for another. This is used in cross-phase applications and the phase timing works similarly, but with two filtered sinusoidal waveforms:

graph showing two filtered sinusoidal waveforms

Cross-phase, or relative phase, tells you the timing difference between the two accelerometers and not the lag to a fixed shaft reference. This method can be very useful in determining faults such as:
▶ Type of imbalance (static/coupled/dynamic)
▶ Structural faults
▶ Misalignments
▶ And more

Cross-phase with measurements in phase would look like this:

Graph showing cross-phase of two sign waves with measurements in phase

A very simplified example is that when a coupling is misaligned, often the two axial planes will be 180° out of phase. Another example would be if a machine bed is ‘rocking’ from the front to the back of the machine bed, the vertical accelerometers on the non-drive end to the drive end of the motor may act 180° out of phase. There can be other factors that change this phase reading to a less perfect phase shift, however, for this example at hand, the signal timing would look more like this:

graph showing showing data from a machine rocking from front to back on the machine bed, and vertical accelerometers acting 180 degrees out of phase

A simple way of thinking about this is when the drive-end accelerometer vibration is towards the floor, as the baseplate is rocking, the non-drive end accelerometer is acting upwards (towards the sky) and therefore opposing each other:

blue industrial motor on a brown foundation with two AC102 accelerometers

As can be seen, this extra information provided by phase can be incredibly useful in determining faults more accurately.  Most collectors can determine phase easily by setting up a cross-phase measurement, providing you with a phase readout in degrees (or radians, for those who like pi 😉).

When using switch boxes, often you only have one output, and then in the case of individual BNC junction boxes, the individual outputs may slow your routine data down. CTC’s SB152 Enclosure offers a bridge between these two use cases, with both a switched (BNC/2p mil) output for routine data and individual BNCs for each channel in the same box:

SB152 fiberglass switch box with open front showing a F2DB BNC connector attached

Therefore, when taking your routine data, you can use the switched output for efficiency, and should you need to dive into advanced diagnostics on the machine, you can easily perform cross-phase measurements from the individual BNC outputs provided!

This interface can also allow two measurement locations for two-plane balancing, some collectors may be able to power a permanently mounted tachometer (usually a hall-effect proximity switch) to acquire all the balancing data directly from the switch box in enclosed fans such as Air Handling Units, where running cables out can be difficult or unsafe.

Back to Blog Archive

CTC corporate logo with a brushed steel texture
7939 Rae Boulevard
Victor, NY 14564 USA
+01 585.924.5900
Send POs To:
[email protected]
Contact a Sales Rep:
[email protected]