View Single Post
  #3   Report Post  
Old October 5th 06, 09:45 PM posted to rec.radio.amateur.homebrew
Roy Lewallen Roy Lewallen is offline
external usenet poster
 
First recorded activity by RadioBanter: Jun 2006
Posts: 1,374
Default PSD on a spectrum analyzer.

To add to what Tom wrote,

Ordinary spectrum analyzers sweep the filter across the band of
interest, so what you see is not only the power in that filter
bandwidth, but the power that's at each frequency at the moment the
filter is there. Movement of the filter frequency constitutes modulation
of the observed signal, which in itself creates sidebands. So what you
see on a spectrum analyzer is heavily influenced by the relationship
between the nature of the modulation causing the bandwidth spreading and
the sweep rate of the analyzer. With some broadband signals, you can get
vastly different displays by changing the resolution bandwidth and sweep
rate, or with different filter shapes.

This is covered in detail in _Modern Spectrum Analyzer Theory and
Applications_ by Morris Engelson (Artech House, 1984). I highly
recommend it to anyone needing to measure and interpret complex
waveforms on a spectrum analyzer.

Roy Lewallen, W7EL

K7ITM wrote:
A spectrum analyzer has filters with particular bandwidths. What you
see displayed is the power passed by the filter. If you have the
analyzer set for a 5kHz resolution bandwidth, what you see, then, is
the power in (nominally) 5kHz of spectrum. To make things accurate,
you need to understand the shape of the filter, as well. In addition,
you can't necessarily use that number to find power spectral density in
other than a gross sense unless you know that the power is distributed
more or less evenly across that piece of spectrum. That is, if you use
5kHz resolution bandwidth and the signal has 99% of its power in a 1Hz
segment within that 5kHz, the signal's actual power spectral density
(at its center frequency) is much higher than the indicated power per
5kHz.

Many modern spectrum analyzers will display power spectral density
directly for you, taking into account the filter shapes and such.
Also, beware that resolution bandwidth and video bandwidth (on
analyzers that have both) are not the same thing.

Cheers,
Tom

wrote:
Hello,
I think what we read on a spectrum analyzer is dBm for the specified
range.

dBm/MHz is the PSD.

Lets say if the signal bandwidth is occupying 500Mhz (UWB) and has a
PSD of -30dBm, then what is my power in the entire channel?

(-30dBm)*500 = (1microW)*500 = 500 microWatts = -3dBm

So do I see the -3dBm signal level on the spectrum analyzer in the
whole bandwidth?

Please correct me if I am wrong.


Thanks.