Sine wave data conditioning

[Vij@y]

Hello Everyone,

I acquired sin wave data from NI USB DAQ 6353 and found the signal is little distorted. I would like to further condition this signal data and get a linear output nearer to pure sine wave. Please find the snapshot of Sin wave on below:

Also I have attached Raw Sin data as an attachment. 

Wondering which filter will bring such an output. Kindly suggest on this.

Thanks in advance.

Regards,

Vijay

[santo_13]

If you know the exact frequency of the signal, the bandpass filter will filter out the frequencies outside the range.

BTW, the signal condition is something you should do before you feed the signal into the DAQ and not in the software.

[altenbach]

So you acquire a signal that you know is sinusoidal (is it?), but then you talk about "output".  Is the "output" just a graph/chart or e.g. an analog output?

Your sine has a significant offset and phase shift, but otherwise looks quite clean. Do you just want to eliminate the noise? Noise is NOT a distortion! Please explain what you mean by distortion and point them out. (extra harmonics? nonlinear response? etc.).

Your noise frequencies seem to be very well separated from the fundamental frequency and could easily be filtered in hardware before the DAQ.

Can you explain the entire experiment? Does the sine always have the same frequency or can it change over time? What is the expected frequency range? How about the offset? Is it fixed? What are the units of the time axis? Are you looking at a ptbypt solution, filtering as the data is acquired or post processing once you have all data.

[altenbach]

---

@altenbach wrote:

Your sine has a significant offset and phase shift, but otherwise looks quite clean. Do you just want to eliminate the noise? Noise is NOT a distortion! Please explain what you mean by distortion and point them out. (extra harmonics? nonlinear response? etc.).

.

---

As you can seem from the picture, there are no "distortions". It is just a simple sine with noise. I can manually simulate a sine that looks exactly the same, except for the noise. (Just a quick manual approximation. We could do a nonlinear fit for exact parameters)

Please explain exactly what you want to do. It's not clear.

[altenbach]

---

@altenbach wrote:
We could do a nonlinear fit for exact parameters).

---

Here are the best fit parameters to your data. Note that there is no structure in the noise!

(If you do a histogram of the residual, t's not exactly gaussian noise, but probably close enough)