convert time domain to frequency domain

[sgirgin]

hi all

I want to convert time domain to frequency domain.In attachment i am using two ways to do it but frequency stuck at zero point.Magnitudes are ok.

Am i underestimating Nyquest frequency?

[crossrulz]

What is the frequency of the signal you are sampling?

[sgirgin]

I have a microphone and i am hiting it which means i have no regular signal

[Bob_Schor]

It might be helpful to see the graph of the data ("Waveform Graph"), the FFT Peak ("Waveform Graph 2") and the (missing) Spectral Measurement Phase plot.  I'm not certain that the FFT Spectrum function accepts the (much-reviled) Dynamic Signal Wire (note the red Coercion Dot on this input).

To what is your DAQ Assistant connected?  Are you getting samples?  Are you really processing 100-point samples at a 10K sampling frequency (which means you are trying to run this loop every 10 msec)?

I'm assuming you know something about Signal Theory, so such concepts as the relationship between sampling frequency, number of samples, Nyquist frequency, frequency resolution, etc. are well-understood by you ...

Bob Schor

[sgirgin]

To what is your DAQ assistant connected?I dont understand this you mean the wiring of sensor.if so it wired to ai3. When i wire signal generator and run sin function i can see the frequency exactly but microphone tick and tacks are not accurate.Also i see in DAQ Assistant i have chosen acquisition mode as continuous sample, maybe i should choose N sample.

[crossrulz]

---

@sgirgin wrote:

To what is your DAQ assistant connected?I dont understand this you mean the wiring of sensor.if so it wired to ai3. When i wire signal generator and run sin function i can see the frequency exactly but microphone tick and tacks are not accurate.Also i see in DAQ Assistant i have chosen acquisition mode as continuous sample, maybe i should choose N sample.

---

Sounds like you need to setup using an analog trigger on the DAQ so that you actually have a signal to properly analyze.  Otherwise you are just getting a quick blip (not long enough for you to see) when you hit the microphone.

[sgirgin]

I am sending a screenshot.Last update i have chosen rate and n of samples as 10 000 for each. My signal has no ordinary shape and is not regular maybe because of this i cannot follow exact frequency.

[Bob_Schor]

Are you tapping your microphones at 5 Hz (5 times/second)?  That is what the left graph seems to show, while the right graph shows a peak at 5 Hz.  However, if you are really sampling at 10KHz and if the Express VI is doing something sensible with these data, I'd expect the frequency range to go from 0 to 5KHz, not 0 to 10.  There's obviously something here that I'm missing (but, then, I almost never use Express VIs) ...

Bob Schor

[sgirgin]

Yes i am tapping 5 times and peak shows at 5. x axis is chosen by me from 0 to 10

[Bob_Schor]

Aha, so you are displaying 10 out of 5000 frequency points (or 0.2% of the frequency data).  Are you aware that when plotting the amplitude portion of a spectrum, log/log coordinates are usually chosen?  Why don't you try plotting, say, the first 100 points, and at least making the Y scale logarithmic ...

Incidentally, if you save your Screen Shots as PNGs (or even the dreaded JPeg), you can attach them directly to your Post without requiring us to open a Word Document (only to find small embedded images).

Bob Schor