displacement with noise

[MiamiVipul]

I'm collecting the displacements in two perpendicular directions of a rotating shaft at the mid-span. When the sampling rate is at  a lower value (less than 3000 samples/s), I get a signal as in Disp1.jpg. As I increase the sampling rate, I continue to get some sort of a noise on top of my signal (as in Disp3.jpg) and it will increase along the wave as I increase the sampling rate.

Can anyone explain what is really happening in here. Is this noise, or it is due to something due to ill-configured Sampling rate/number of data points combination?

I'm using Labview 7.1 on XP

Shaft is rotating at 140 RPM, with slight imperfections on the shaft, not a balanced shaft so it is sort of wobbling as you see on the displacement graphs.

Thanks a lot.

Vipul.

[Conseils]

Do you have any anti-alias filters?
Could just be that the signal is being missed by the lower sampling rate

From a vibration perspective, is there anything like a generator or 'firing' occuring at this instant?
Perhaps you have a C.V. / hookes joint or something knocking / loading, further up the system?
 
It looks like its quite a high frequency compared to your 2.3Hz fundamental. Try looking at it in the frequency / order domain as this wil be more indicative of the source of the 'noise'. If it's above 25Hz my guess is that you won't be overly worried about it.

See if it changes amplitude / frequency with respect to the rotaional speed (i.e. related to the rotational speed) and or if it's load related?

Hope these help.

Message Edité par Conseils le 09-12-2006 08:12 PM

[Ben]

This could be the cables/ckt response to a step stimulus.

A low pass filter, or using a buffer to make sure you do not have reflections on your transmision line may help that situation.

Ben

[unclebump]

What type of displacement sensors?? If it is just a linear pot, it could be a bad wiper inside the device.

[MiamiVipul]

Hi all:

Thanks a lot for all of your responses.

My setup has an eddy-current displacement sensor and NI 4472 DSA. I'm measuring the vertical displacement of a rotating shaft (at 120 RPM); a very basic setup.

I use PFI0 external trigger to collect data from optical encoder (one pulse/rev) at a certain angular location of the rotating shaft. The noise comes only when I use external trigger. I used a function generator instead of the encoder, but I still have the noise. I'm using screened wires and I kept them away from all other noises. I have another setup that collect data continuously without trigger; it does not have any noise at all.

Also, the noise appear only at high sampling rates.  If I go over 20,000 samples/sec, it shows visible noise modulated over the regular waveform (see figure). As I reach 40,000 samples/sec, you will see a large carrier frequency all over the signal.

Again, nothing happens when I do not use regular Labview code with continuous collection. It takes place when I use trigger and when the sampling rate is above 20k / sec.

I have attached my VI and a sub VI i use.

Thanks in advance for your help.

Vipul. 

[unclebump]

Jumper you input channel + and - terminals and see what you get. Then see if labview can determine the frequency of the noise in the signal. It sounds like it is in your hardware.