[RTTY] K or carriage return

[Al Kozakiewicz]

A couple of things:

1.  PSK31 is 31.25 baud and  RTTY (in ham applications) is 45.45 baud.  RTTY is thus about 45% faster then PSK31.
2.  I don't know how effective simple error detection would be in a RTTY system.  The addition of a parity bit, for example, would add quite a bit of reliability (you could flag characters as known bad), but that only works where single bit errors are statistically the most likely.  I strongly suspect multi-bit errors are very common when HF radio is the medium and single bit parity would not be very useful..
3.  More sophisticated error detection (e.g. CRC) requires even more overhead data to be transmitted and is geared towards fixed message sizes.
4.  Backwards looking correction (ACK/NAK) is efficient and simple to implement in automated circuits, but the nature of amateur as a half duplex medium isn't really conducive to that. And again, geared to fixed size messages.
5.  Error correction requires a lot of data redundancy.  It can be made to work (JT65, e.g.) but that slows down the mesage rate considerably.It has the advantage of not requiring any handshaking.

Your brain does a surprising amount of error correction, because it understands context, that is difficult to do with a serial protocol that's required to function is a low bandwidth medium.

Al
AB2ZY

________________________________________
From: rtty-bounces at contesting.com [rtty-bounces at contesting.com] On Behalf Of Kok Chen [chen at mac.com]
Sent: Wednesday, February 16, 2011 1:00 PM
To: RTTY contest group
Subject: Re: [RTTY] K or carriage return

On Feb 15, 2011, at 2/15    6:56 AM, Bill, W6WRT wrote:

> Has anyone here ever operated a contest with MFSK or other
> error-correcting modes? It would be fascinating to see the actual
> tradeoff between speed and accuracy.

My gut feeling is that for stations that are running high power, RTTY
will be better, since power will help compensate for the reduced
sensitivity.  You can run so much power that even under QSB, the
weakest point of your signal is still way over the RTTY decoding
threshold, for example.  You can repeat until enough of your exchange
gets through for the other end to log.

But if you are running 25 watts into a dipole?  I suspect that CQ'ing
in DominoEX will be produce more callers.

One thing to keep in mind though, is that some of the better modes can
print a signal even when it cannot even be seen in a good waterfall.
So how would an S&P station find a weak CQ'ing station?  You cannot
see them in a waterfall much less hear them :-).  There is some cases
where you don't need that extra sensitivity.  Take a look at the top
waterfall here

http://homepage.mac.com/chen/w7ay/cocoaModem/UsersManual/mfskManual/mfskManual/dominoex.html#weak

That signal prints so close enough to 100% that you have to wait a
long time for an error hit, but yet it can barely be seen on a
waterfall.  The question is therefore, for something like a contest,
is the extra precision even needed.

As PSK125 contests have shown, exchanges in other modes can be fast
and adrenaline producing.   However, like RTTY, the PSK31 family has
nothing to help slow QSB, so does not work worth a plugged nickel when
conditions deteriorate.  With RTTY, you can at least jack up the power
to get through selective fading and flutter.

Also, some of the "slowness" that one perceive with PSK contesting has
to do with the awkward exchanges from some of the operators since they
attract many non-contesters.  If many people start sending you dots
and the color of their dog, average RTTY exchanges slow down too.

Just my humble opinion and worth the two pennies.

73
Chen, W7AY

_______________________________________________
RTTY mailing list
RTTY at contesting.com
http://lists.contesting.com/mailman/listinfo/rtty