Il 12/03/2011 18.25, Michael Black ha scritto:
On Sat, 12 Mar 2011, Piero Soldi wrote:

"Ralph Mowery" ha scritto nel messaggio
m...
The bandwidth will depend on what shift is being used. Some of the more
common shifts are 170,425, and 850 Hz.

The actual bandwidth needed to receive them will be slightly larger.
Something like 200 for the 170 Hz shift and 1200 for the 850 Hz shift.
The transmitted bandwidth will be larger yet.

It is similar to an FM signal, it theory it is infinite,but in
practice it
can be much less.



"garyr" wrote in message
...
Anyone know what the bandwidth of a 50 baud RTTY signal might be?




Consider also the group delay of receiving filter...
with narrow Butterworth or Chebyshev crystal filters, group delay is big,
distorting
RTTY , CW or other digital signals, and ultimate result is error in
decoding.

So bandwidth need to be more than necessary. A Gaussian filter is
better for
group delay,
but is worse for steepness.


I figured the original question was the bandwidth the signal took,
rather than the bandwidth of the receiver.

I didn't think anyone went for a receiver filter to match the RTTY signal.
Well maybe the hardcore RTTYers. But generally one used a reasonable
filter in the receiver, and counted on the audio filters in the
demodulator. There again, relative few decoded at the IF, but even then
the articles I've seen didn't go for any special narrow filters for RTTY.

Michael VE2BVW

Hi Michael,
i was an hardcore RTTYer a lot of years ago... :-)

I've used Teletype Corp ( now IBM ) TG7, Kleinschmidt TT100,
Olivetti T2, Siemens ??? and after, first attempts of software
decoders.

Yes, normally you use a CW filter to receive RTTY, ie: 500/600 Hz BW for
170 Hz shift so the group delay is not problematic if rf tones are
centered in passband of filter ( stay away from edges ).

Cheers,
Piero I5SPO