I'm trying to use this program I found to analyze a signal but I don't know what units
they're using for the y axis.  

http://www.mathworks.com/support/tech-notes/1700/1702.html?BB=1

any ideas of what the .5 something units are?  Or anybody know of a way to get a power vs
frequency plot accurately?

Yah, the y-axis is Power Spectral Density (PSD).

the y-axis values are just a relative thing....the relative value of
different 'frequency band' could be used to identify the 'dominating
frequency' or 'frequency band' from the signal. the plot shown in this
website is showing that the waveform is uniform, so, you can see only one
big PSD at 200Hz. if the waveform is a mixture of different source
generating this signal, then you would find different bands of frequency
from 0-512Hz (x-axis).

the details given on this Mathworks website are accurate to my
understanding, as it is a standard algorithm to plot PSD-Frequency for any
waveforms....

i hope it helps. plz let me know if you need further information....

with best regards
Faisal
On 06/07/07, x...@yahoo.com <x...@yahoo.com> wrote:
>
>   I'm trying to use this program I found to analyze a signal but I don't
> know what units they're using for the y axis.
>
> http://www.mathworks.com/support/tech-notes/1700/1702.html?BB=1
>
> any ideas of what the .5 something units are? Or anybody know of a way to
> get a power vs frequency plot accurately?
>

Ok so now I figured out the amplitude is fine for this program and is given in watts.

can anybody tell me why this program doesn't work when using large frequencies? say 10e9?
or a combo of frequencies?  It works well with the small ones, and I believe it has to do with
the "sampling frequency" they use but I'm not sure.

***************************************
I'm trying to use this program I found to analyze a signal but I don't know what units they're
using for the y axis.  
>
>http://www.mathworks.com/support/tech-notes/1700/1702.html?BB=1
>
>any ideas of what the .5 something units are?  Or anybody know of a way to get a power vs
frequency plot accurately?

Hi,

The reason you cannot do it for high frequency is 'fs'  which is 1024. Nyquist criterion
established that to accurately retrieve a frequency you have to sample it at atleast twice the
highest frequency component present in the signal. So for 10e9 you need atleast 20e9 sample
points. The .5 is due to the division of frequency components into +ve and -ve frequencies. The
program is using only half of the FFT so the power is multiplied by 2 to keep it same as the
case where full FFT is used. Hope this helps.
x...@yahoo.com wrote:                                  Ok so now I figured out the amplitude is
fine for this program and is given in watts.
 
 can anybody tell me why this program doesn't work when using large frequencies? say 10e9?
 or a combo of frequencies?  It works well with the small ones, and I believe it has to do with
the "sampling frequency" they use but I'm not sure.
 
 ***************************************
 I'm trying to use this program I found to analyze a signal but I don't know what units they're
using for the y axis.  
 >
 >http://www.mathworks.com/support/tech-notes/1700/1702.html?BB=1
 >
 >any ideas of what the .5 something units are?  Or anybody know of a way to get a power vs
frequency plot accurately?
 >
 >

Amit Pathania