Detect differences between cup anemometer bearings in a wind tunel

[AnalogKid2DigitalMan]

Instead of microphones, perhaps try an accelerometer mounted to anemometer near the bearings?

-AK2DM

[VDuarte]

I already did that, using an adxl3xx accelerometer, which was unable to detect any vibration from the anemometer. My guess is that it's not sensitive enough to detect such a low scale vibrations. Do you have any tip on how to properly select an accelerometer for an application like this?

[AnalogKid2DigitalMan]

Bummer, was the accelerometer oriented in the correct axis and signal conditioning used? 

I don't have any tips regarding selection, what is the expected frequency range of the vibration?

How about a contact microphone or stethescope type pickup? They may have too limited bandwidth though.

http://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=Contact_Microphone_C...

-AK2DM

[VDuarte]

I acquired both x and z axison the shaft of the anemometer, wasn't that the correct approach? What would be the correct axis in this case?

What do you mean by the signal conditioning I should use? All I got was noise varying over pretty much all the frequency ranges, there was no proper way to filter it.

The anemometer cups rotate between 4-16m/s, which translates to a frequency from ~60-250Hz. So I should expect something at that frequency and then some harmonics at higher frequencies, if I envelope the signal. Or is my line of thinking wrong?