[Fwd: LF: Radio 4 in ZL]
Tue, 26 Jan 1999 23:06:07 -0500
Steve Olney wrote:
> You wrote...
> > I looked at the trace of 198kHz which you had on your site and I wondered
> > whether Simon LLoyd-Hughes may comment regarding the phase mod data on
> BBC Radio 4.
> > I would have thought this may have shown up as sidebands but I dont know
> > the parameters of it (electricity supply switching information). It could
> > be another BC TX on 198...
> Yes I know it looks a bit suspect at first look and I was a bit puzzled as
> it was not what I expected. But I think a number of factors need to be
> looked at. Firstly, the display BW (as shown on the image) is 0.3815Hz so
> any sidebands would have to within about +/-0.19Hz of the carrier. The
> second factor is the *average* power in the sidebands for typical program
> material from a station which has a high quality signal (little or no
> amplitude compression). We are so used to seeing the standard AM
> modulation spectrum with a single tone modulation (carrier with two
> sidebands each of 25% power of the carrier) that we expect something
> similar for the averaged spectrum of a typical BC program.
> In fact as you take a longer record (giving narrower and narrower FFT bin
> BW) the *lower* are the levels of the program modulation sidebands
> displayed in each FFT bin. This statement, which sounds at first to be
> either counter-intuitive or complete garbage, can be supported as follows
> by comparing what happens to a number of AM signal cases as the FFT bin BW
> is narrowed:
> 1. Modulated carrier with a single tone - as FFT bin BW is narrowed the
> level of the carrier remains the same and the power level of each of the
> sidebands remains at 25% (assuming that the frequencies land near the
> middle of FFT bins).
> 2. Modulated carrier with two tones - now the sideband power is shared
> between two sidebands so they appear 3dB lower on the display *except* if
> the two tones are so close together that they end up in the same bin.
> Adding more tones causes the available sideband power to be shared between
> those tones reducing their individual levels in the display.
> 3. Modulated carrier with noise of BW of say +/-5kHz - The carrier level
> will be undiminished but now we have (over time) an infinite number of
> tones from the noise modulation. Now as we reduce the FFT bin BW the
> carrier remains the same but the sideband power is spread over more and
> more FFT bins and so their display level reduces.
> Over time program material modulation contains all frequencies over the BW
> (unless someone whistles constantly into the microphone or some music) and
> so the *averaged* level in each narrow FFT bin is quite low, that is it
> looks like noise -:)
> This idea doesn't at first look sit well with intuition but is the basis
> for detecting constant tones in noise by using narrow BWs.
> To illustrate this effect I have taken a record of a local national BC
> station using both my software and Spectrogram (using narrow bandwidth and
> spectrum averaging of 128). In both cases they show that the *average*
> level of program modulation sidebands is over 40dB down on the carrier!
> These two images can be seen by going to
> and clicking on the "Other LF signal source" link and then clicking on the
> "Local BC station" link.
> Dave, could you verify this by using Spectrogram over there on BBC 4 ? I
> used 5500 sampling, 512 sample points, and spectrum averaging set to 128.
> The receiver was set to CW mode with a sidetone frequency of 600Hz.
> I will try and modify my software to record over 24 hours and analyse it in
> 1/2 hour blocks and draw a graph of the strength of the signal on 198kHz
> but my leave is over and I will have much less time to spend from now on.
> 73s Steve Olney (VK2ZTO)
> email: firstname.lastname@example.org
> Web Page: http://www.zeta.org.au/~ollaneg
> LowFer: http://www.zeta.org.au/~ollaneg/lowfer.htm
> LF Receiving - FRG-100, 8-turn 10sqm Loop
> or Ferrite Loop + K0LR Preamp