Hi friends

who can give an idea about relative phase measurement (procedure) using 2-channel data collector analyzer empac 2500 and e monitor odyssy software.

my idea is to fix one probe in a point and to move the other probe to other directions and points, but if i fix one probe in bearing no.1 vertical as a referance for example, then i move the other probe to other direction and points, how can i know the phase differance between bearing 1 Axial and bearing 2 axial for example.

thank you

The easy way is to use a keyphasor. Then you only need one transducer and you just subtract phase angles to get relative phase.....

Relative phase calcualtions require a single frequency and the same units (dont compare accel with velocity). If there are many frequencies it will be difficult, that is why filtering with the keyphasor makes it easy.

In your example, if you are lucky most of the vibration will be mostly 1X running speed. Then you lay the waveform for one transducer on top of the waveform for the other, and compare. If one waveform peaks, where the other one valleys, then the are 180 degrees out of phase. If on peaks where the other peaks, they are in phase, or 0 degrees relative phase.

Relative phase in this form is measured from 0 to 180 degrees...

Lastly, if you use a real time analyzer that has transer functions and phase functions you can look at relative phase at any frequency. I dont know enough about the enpac to know if it has those features.

You have to tell Enpac what frequency you are comparing phase to if you have no tach or other speed imput. I think there are three different ways it can monitor the frequency of interest. If you are going to compare axial readings to axial readings, then the transducers both have to be in the axial direction. The Enpac will not take into account transducer direction. You have to do that.

Dave

[QUOTE]Originally posted by Steve Schultheis:

Relative phase calcualtions require a single frequency and the same units (dont compare accel with velocity). If there are many frequencies it will be difficult, that is why filtering with the keyphasor makes it easy.


dear steve
I mean cross channel phase, using 2 channel data collector /FFT analyser

Say one has a reference signal R and two other signals A and B. A transfer function (cross channel phase) produces the phase difference at a particular (or each) frequency.

So, one can compute phase of A - phase of R (or it might be the other way around) and the phase of B - phase of R signals. If one wants to know the phase of a relative to B, i.e. the phase of A - phase of B then this is [phase of A -phase of R) - (phase of B - phase of R)]. This process gives relative phase using a common reference.

This takes care of the phase difference between signals (at each frequency). One may want to translate this into motion phase, which requires accounting for the physical orientation of the transducers from which the signals arise.