is it possible in Labview? Suggest me...

[johnsold]

I am not sure I completely understand your question.

When you are working on problems like this, you will always have a range of variation in the measurements.  You may need to measure many pumps of each type to determine what the criteria are for making the adjustments, and they may differ for different pumps.  When I did something similar for food mixers many years ago, we had detailed data on about 50 mixers (several weeks' production) before we designed the instrument and then ran about the same number afterwards to validate the performance.  Without knowing how mcuh variation you have from one pump to another, I cannot even guess how much data you will need to determine how to make the adjustment decisions.

Lynn

[shjukheter]

I have checked lot of pumps. When there is no noise the frequencies are in between 170 to 190HZ(You can see this in attached pictures on http://forums.ni.com/t5/LabVIEW/Sound-and-Vibration-pallet-is-only-for/td-p/1695594/page/2 ) and when metallic noise encounter it is warning also varying on pump frequencies and more until 1KHz for small and until 2KHz for big types. So, I just want to differentiate the noise and then I know to take decisions on that noise. Problem is only, How can I do these exact differentiation. Weather it is possible in LABVIEW or in another way.

[Yamaeda]

If you do a FFT-analysis you'll get amplitude values for different frequency bands, thus it would be very easy to see if 500Hz+ increase in value.

/Y

[shjukheter]

Yes, I have done this and you can see the pictures in my previous message link.

[dkfire]

And if you are looking for frequencys of 2kHz, then remember to have a sample frequency of no less than 8kHz ofr your DAQ.

EDIT: Can see you are sampling with 10kHz.

[shjukheter]

I want to differentiate the noise from spectrum or waveform. I don't care about what device is at other end. I just need to read the total frequencies (real + noise) and I would like separate the noise frequencies from that spectrum or waveform. I don't know the real or actual or original frequency of the device on other side but even the device changes, I need to separate the noise from the connected one. You might get my point now, Please, I have wasted lot of time on this but could not figure out. Help me.

[Yamaeda]

From what i understand, you only need to look at 2kHz+ frequencies. After the FFT you can split the generated data array and see if it's any noise, or are you looking for some other information? You could average some "good" FFT-signals and subtract them from the new one to almost only get the noise.

/Y

[johnsold]

Yameada has suggested what you should do.  I think I suggested soemthing similar in one of your other threads.

Have you tried any of these ideas? What were the results? If you have tried and it does not work, please post your VI with some typical data and an explanation of what is not working.

Lynn

[shjukheter]

Hello,

             I have been analyzed with all kinds of spectrums and you can see in pictures below. I have no idea to separate the noise and original frequencies and I would like to do the same for what ever the device is connected to my program.

[shjukheter]

More here.