There is a strange 3600 CPM = 3xRPM peak standing out in spectrum on a motor driving a hammer mill at 1200 RPM through a gear type coupling. What catches one's eye and it is hard to explain is motor fault at 3x when both 1x and 2x are low. 2x is also present in V direction although of much smaller in magnitude.

Nothing comes to my mind other then possible resonance close to 3x, but I could not verify it on a running machine.

3600 CPM is also happened to be the second harmonic (of comparable magnitude) of another machine operating nearby at 1800 RPM. So, I checked coherence between those two sources at 3600 CPM and sure enough it was 100%.

Looks like a transmitted and possibly amplified signal!? Or is it a motor problem!?

Mill5_motor_3x.doc (46 Kb, 28 downloads)

I'd suggest taking a lower Fmax range to determine if the running speed is exactly 1200 cpm and the big peak exactly 3600 cpm.

3600 cpm = 60 Hz which could be power noise. A ground loop is unlikely with a portable data collector (but a real possibility in an installed system). A possible source is a bad cable or allowing the accelerometer cable to be close the the motor's power cables during the measurement.

Also take a lower Fmax on the 1800 RPM machine to verify that the 3600 cpm tone is really twice the run speed. If it is twice the run speed, is it possible to stop this machine briefly to see if the 3600 cpm tone disappears?

Jon
http://www.spintelligentlabs.com