Message-ID: <003101c1655e$a949c5c0$1bb51bca@xtr743187>
From: "Vernall" <vernall@xtra.co.nz>
To: rsgb_lf_group@blacksheep.org
References: <200111010724_MC3-414D-935B@compuserve.com>	 <001101c162d8$6ef36f30$0400000a@parissn2> <3BE31FFF.C0788E6E@usa.net> <160JaP-1LPangC@fwd02.sul.t-online.com> <3BE55DBA.BD1FB024@usa.net>
Subject: LF: Re: Re: spectrum 136kHz I2PHD/dj8wx
Date: Mon, 5 Nov 2001 07:28:52 +1300
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

Alberto and others,

> Hi Uwe,
>   yes, our spectra are quite different. And they seem also to differ from
what Ko
> has captured and sent to me. The only common factor is that at 136kHz
there
> seems to be a sort of clustering of spectral lines, albeit different from
location
> to location. I can only say that the spectrum at my location is
time-invariant, sort of.
> I have always seen those lines, night and day, winter and summer, with no
signs
> of modulation. Go figure...

The most plausible explanation, as already mentioned on the reflector, is
that the lines come from the sampling rate of 8 kHz used by digital
telephone systems.  There could be several telephone exchanges in a city and
while they likely have stable master oscillators, they are not necessarily
"locked".  Telephone systems still use a lot of copper pairs, covering
kilometres of distance, so it would be feasible for some radiation to occur
for any high frequency spectral content in the telephone signal.

The basic digital telephone circuit has a bit rate of 64 kbit/s (8 bit data
sampled at 8 kHz, analogue audio response to 4 kHz) so this also has
potential for "harmonic leakage" into the LF radio band.  64 kbit/s is still
related to the basic 8 kHz sampling rate, so it could be difficult to
distinguish which bit stream the harmonics are coming from.  Harmonics of
the 8 kHz sampling rate would not have modulation, but harmonics of 64
kbit/s would vary somewhat with baseband modulation.

No doubt the 136.00 kHz nominal lines have come up for discussion because
there are many amateurs tuning a narrow band for DX radio signals, and with
new tools like Argo it effectively gives bigger ears for detecting
sinusoidal types of emissions, even if they are below conventional noise
levels.  As others have suggested, there is probably a whole array of weak
LF lines that originate from digital telephony, and they would tend to be
worse in larger cities (highest number of telephone lines acting as
antennas, multiple exchanges, etc).

The practical application for DX radio testing is to avoid spot frequencies
that are multiples of 8 kHz !   Telephone exchanges are unlikely to respond
with QSL cards ;-)

73, Bob ZL2CA