Need Help Please!!!

[insomnia]

I am doing vibration testing on engines using two different mounting methods. In the first case the accelerometer is isolated from the engine and in the second case the accelerometer is glued onto the engine. The problem i am having is when i calculate the power spectrum of the data collected from the glue mounted accelerometer i am seeing power at frequencies which are not even in the range of the accelerometer. The acclerometer range is 20Khz but i am seeing power at around 40Khz and 65Khz.
Any ideas why this could be happening would be greatly appreciated.

 

[GregLocock]

It could be (1) tribolitic noise (ie your cable is shaking and generating hf)

(2) flaky connector

(3) a real effect

(4) interference

You should be low pass filtering your signal anyway, at the charge amp or input to the A/D unit.

What frequency are you sampling at? if you aren't low pass filtering perhaps you aren't anti-alias filtering.

By the way normal piezo accelerometers don't stop responding at high frequencies, above their internal resonance, but they do roll off at 6 dB/octave. since you are only an octave or two above the resonance it'll still resond.

What is your instrumentation setup? ICP or charge coupled? if you give me manufacturers names and model numbers i can maybe help a bit more.

Incidentally getting vibration data off engines is often a bit of a miserable process, instrumentation wise.
Cheers

Greg Locock

 

You should specify lower and upper frequency range in setup of data aquisation unit.otherwise your data is useless.