On Jun 10, 2010, at 6/10 4:24 AM, Charlesjr@afgimail.com wrote:
> As far as loosing power at 850hz shift RTTY we never worred about
> that as we could run 40kw from
> some FRT-40's (transmitter).
The power problem when going from narrow shift to a wide shift is bad
but not terrible. The reason is that you are keeping the baud rate,
but just widening the separation of the two tones. The matched filter
for each tone remains narrow -- visualize two band pass filters whose
centers are separated by 850 Hz, each one having the equivalent
bandwidth that is equal to about 1/2 the baud rate (but drops off
slowly as 1/f).
Thus, the extra noise bandwidth is not much greater than using 170 Hz
shift, with the baud rates that we're talking about. I.e., and loss
in SNR has already been accomplished by raising the baud rate :-).
The problem of using excessive bandwidth is however a problem. If
everybody uses proper matched filter, you can start overlapping
signals and get along just fine. Notice that you can place quite a
few amateur RTTY of PSK31 signals in between the two tones of that
loud wide shift commercial RTTY that we hear on 30m and we don't
really suffer ill effects -- and they probably have the appropriate
filters to remove the weak gnats in between their two tones, so they
don't suffer either.
To overlap, you need a higher dynamic range receiving system. If you
are trying to copy a weak signal, the stronger signal that is
overlapping must not exceed the dynamic range of the receiver/sound
card/numerical range, etc. Therefore, in general, I don't think it
will work too well.
You can copy using mark-only or space-only of course. But when you do
that, you do lose that 3 dB of SNR which can affect copy of a very
weak station.
A good reason to use a wide shift is to counter the kind of selective
fading that you find on the lower frequency bands.
To see the effect, you can watch very wide signals, for example the
Coast Guard HF-FAX stations like NMC (in Europe, there are DDH and
DDK) or even a wide DominoEX or Olivia signal on a waterfall display.
What you will notice is that when selective fading occurs on the
higher bands, the QSB "notch" is often very narrow in frequency.
Narrower than 170 Hz. When selective fading occurs there, quite
often, only one of your two RTTY tones will be taken out. Thus, the
demodulator's slicer will automatically compensate for losing the more
attenuated tone (albeit at a loss of up to 3 dB in SNR).
The fading "notch" is much broader at the lower HF like 4 MHz (and I
suspect it is even worse at MF frequencies, but I have not found
appropriate signals to test with). When the fade take away both tones
of your signal, then all bets are off. The receiving system now looks
like it is copying a flat fading signal and you have lost the
advantage of using FSK.
73
Chen, W7AY
_______________________________________________
RTTY mailing list
RTTY@contesting.com
http://lists.contesting.com/mailman/listinfo/rtty
|