DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=googlemail.com; s=gamma;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :content-type:content-transfer-encoding;
        b=tvTpyEvcapx/gmmzZE6kxyJeSLq7is37Gy3pUPghrRRoKnZ9Mxm57fS3HXB2vQPJbe
         bWJ0BWl3xO2Mq1VJjasb/AJb/rU3qNGpn5lyNkq/w0gSI4RjRDcPH25sn11kUsjGRoMe
         yUVwYcI8Wb1FdPJRFe2zPIwtSTfNOsHvrQciU=
MIME-Version: 1.0
In-Reply-To: <38A51B74B884D74083D7950AD0DD85E828AC66@File-Server-HST.hst.e-technik.tu-darmstadt.de>
References: <4B250D02.733.14E9A34@mike.dennison.ntlworld.com>
	 <38A51B74B884D74083D7950AD0DD85E828AC66@File-Server-HST.hst.e-technik.tu-darmstadt.de>
Date: Tue, 15 Dec 2009 10:59:14 +0000
Message-ID: <fc7eccec0912150259v4155e9f0pbbf8dea75353656b@mail.gmail.com>
From: Andy Talbot <andy.g4jnt@googlemail.com>
To: rsgb_lf_group@blacksheep.org
DomainKey-Status: good (testing) 
Subject: Re: LF: "Gain" between qrss3 and qrss10?
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org


Comparing QRSS with visual intepretation cannot be compared directly
with CW using Aural decoding - you have to normalise values.   Your
ears have an 'effective' bandwidth for CW of , probably, around 30Hz.
Within this bandwidth, most experienced operators can hear a tone, at
their optimum tone frequency of choice, at around 5 - 10dB S/N.  But
to decode CW reliably, this S/N has to be nearer 15dB.   CW experience
plays a big part here         (Mal, - stay...........!!)

Using QRSS, you can probably get reliable visual decoding at 10 - 15dB
S/N in the dot bandwidth.   For 20WPM CW, with a dot of say 50ms this
corresponds to 20Hz ( 1/0.05s) which is near-enough to the 30Hz noise
bandwidth assumed.  So aural CW and visual QRSS are roughly equal in
terms of   RATE / BANDWIDTH.

The QRSS advantage comes about since you are prepared to trade off
message duration against bandwidth.    This is basic information
theory as mostly worked out by Shannon in 1948.    If you normalise to
a standard form of Bits/second/Hz (which become Bits / cycle and is
rather more meaningful when using the old cycles per second
terminology of Shannon's day) then CW QRSS, and inded the various Hell
formats as well, are all roughly equal in signalling capability.
ie about 1B/s/Hz at 10 to 10dB S/N  Even  SSB voice fall into about
the same region since the ear/brain does a very good noise canceling
job there as well.  Remember that value - its pretty constant for
'human' modes (and the actual value very operator-experience
dependent)    [... I said 'Stay...." :-) ]

And yes, if the path was stable for several years, and the ionosphere
didn't move to give Doppler spreading on your signal, then dropping to
QRSSSSS....SSSSSSSS 1000000000000 you could indeed communicate with VK
using a nanowatt to a hairpin!

But note those caveats - if the path was stable and ionosphere...
It isn't.
Alan G3NYK and others have studied 137 / 500 propagation anomalies and
more-or-less come to the conclusion the stable period is only tens of
minutes, if that.   Which precludes any faster signalling than a few
tenths of a bit per second.

HOWEVER, and this is where the real future lies...

If you throw away the concept of simple signalling, forget CW, QRSS
etc, and use a properly designed multilevel waveform with heavy
forward error correction and soft decoding then you WILL gain a real
10dB or so over simple on/off. Yes you really will.

As an example take WSPR, which sends 72 bits of source information in
110 seconds, that's about 0.6B/s/Hz (0r 0.6Bits/cycle - I like that
terminology !) .  It can do that at -29dB S/N normalised it 2.5kHz -
we know that,  it is accurately measured and reported by the decoding
software.   Since the WSPR signalling bandwidth is 1.46Hz - the symbol
rate,  that -29dB at 2.5kHz increases by 32dB to +3dB.   And that's
completely error free

I reckon decent FEC can add 6dB to the signalling efficiency of human mode=
s.
Open to arguments.

See   www.g4jnt/MartleSham.htm    for a powerpoint presentation
including some simulated CW in defined S/N and more vlaues / details

Andy
www.g4jnt.com

This email has been scanned for damaging side-effects by the health
and safety police



2009/12/15 Stefan Sch=E4fer <schaefer@hst.tu-darmstadt.de>:
> Dear LF,
>
> I have done some further thoughts/questions about the "gain" between the
> qrss modes.
>
> If the "gain" or better said SNR between qrss30 and qrss3 is 10dB, so ca=
n I
> do a QSO to VK with 1W TX Power and an antenna efficiency of 1/1000 if=
 I
> choose QRSS10000000000? That=92s a litte jokingly formulated but you kno=
w what
> I mean. In other words, where are the limitations of this rule?
>
> In CW I think that rule cannot be valid because a dot in CW (say 20wpm)=
 is
> much shorter than 0,3s but to my experience the =93gain=94 compared to=
 qrss3 is
> not even 10dB. Furthermore a PC must be seen as a source of QRM, especia=
lly
> when connected to the antenna system (e.g. for switching a tx relay with=
 the
> qrs-programm) and RX (even when using a insulating transformer for the=
 AF),
> and so it always brings a =93negative gain=94, even when one uses 10s of=
 ferrite
> cores=85
>
> And: I always thought of the energy "waste" when applying 1kW (before 6=
 yrs
> when I became qrv with a much worse antenna) to the antenna and pumping
> 999,9W into the earth. But when I need 10x the time in qrss30 to do the=
 qso
> with 100W, the needed energy is the same (compared to 1kW in qrss3). So,
> qsos could be done much faster if choosing 10x the power and stay in qrs=
s3.
> The additional advantage would be, if I try to do a TA QSO I could just=
 stay
> in the region of 137.7 and those who are qrv in qrss3 and only watch the
> qrss3-sector would get a nice signal and thus more fascination for the
> LF-hobby (I often think about the LWLs watching my transmission when no=
 one
> is answering my call ;-)). In sum, the needed energy (=3Dmoney) is the=
 same
> but the needed time would be 10x less.
>
> Of course, there would be some problems like partial discharges due to=
 the
> high voltage at the antenna and for 10kW the kitchen stove must be conne=
cted
> in parallel, that could cause QRM with the XYL... ;-) But generally it=
 would
> be interesting!?
>
> What is the maximum power that is applied to an antenna yet? What have=
 you
> heard yet? I have applied 1kW in a bad antenna (=3D> abt 100mW ERP), so=
 10kW
> would be possible when considering to stay in the legal limit. 10kW woul=
d be
> possible since one doesn't need a ferrite core that has to keep the powe=
r
> (in Class E mode). In USA it would be even more easy due to the fact tha=
t
> the 110V of the mains could just be rectified. Then I would take some
> STW12NK90Z in parallel and a IXDD414 and try it with 15A antenna current=
 ;-)
>
> Sorry, but in that hobby it is really easy to become euphoric=85
>
>
>
> 73 and hope for fine qsos for us J
>
> Stefan / DK7FC
>
>