joraiz91
Newbie level 2
- Joined
- Feb 11, 2013
- Messages
- 2
- Helped
- 0
- Reputation
- 0
- Reaction score
- 0
- Trophy points
- 1,281
- Activity points
- 1,300
Hi all,
Basically, I made a physical experiment and saved the output data from the oscilloscope in .csv file type. So now I have lots of data with the .csv format.
My next step is to process this data, and get the spectrums, SNR's, and power spectral density (PSD) plots. My first option was to use Matlab to process the data. I semi-followed the algorithm in the Matlab SDToolbox by Brigati (https://www.mathworks.com/matlabcentral/fileexchange/2460), i just edited the operating frequencies and the number of samples. Basically the steps I did were:
1. multiply the input data with the hanning window
2. generate the signal sinusoid
3. match this generated sinusoid with the input data, and extract the actual signal from the input data. This holds the signal, and this is from where the signal power is derived.
4. Subtract this signal data from the original input data to get the total noise.
5. Normalize, FFT and plot..
BUT! I am having problems with this procedure, in my PSD I expect a peak at 450kHz (my signal), however my matlab codes provide me a PSD with a peak at 6MHz++. I think there's something wrong with my normalizing and stuff. Could you please give me any tips for this?
ALSO, I AM THINKING OF IMPORTING THE CSV FILES TO CADENCE INSTEAD AND FFT AND PROCESS THEM USING CADENCE. But I can't find HOW TO IMPORT CSV FILES TO CADENCE. Do you know how?
Your replies would be highly appreciated. Thank you.
Basically, I made a physical experiment and saved the output data from the oscilloscope in .csv file type. So now I have lots of data with the .csv format.
My next step is to process this data, and get the spectrums, SNR's, and power spectral density (PSD) plots. My first option was to use Matlab to process the data. I semi-followed the algorithm in the Matlab SDToolbox by Brigati (https://www.mathworks.com/matlabcentral/fileexchange/2460), i just edited the operating frequencies and the number of samples. Basically the steps I did were:
1. multiply the input data with the hanning window
2. generate the signal sinusoid
3. match this generated sinusoid with the input data, and extract the actual signal from the input data. This holds the signal, and this is from where the signal power is derived.
4. Subtract this signal data from the original input data to get the total noise.
5. Normalize, FFT and plot..
BUT! I am having problems with this procedure, in my PSD I expect a peak at 450kHz (my signal), however my matlab codes provide me a PSD with a peak at 6MHz++. I think there's something wrong with my normalizing and stuff. Could you please give me any tips for this?
ALSO, I AM THINKING OF IMPORTING THE CSV FILES TO CADENCE INSTEAD AND FFT AND PROCESS THEM USING CADENCE. But I can't find HOW TO IMPORT CSV FILES TO CADENCE. Do you know how?
Your replies would be highly appreciated. Thank you.
Last edited:
