[RTTY] FREQUENCY IN USE BY ... (JT65/JT9)

[Kok Chen]

On Jan 8, 2014, at 9:33 AM, Peter Laws wrote:

> As long as everyone is at 31 (and a quarter!) bauds, who
> cares?  Yes it's slow compared to the full 45.45 bauds of regular
> amateur RTTY but ... so? 

Keep in mind too that PSK31 uses Varicode instead of a fixed-length Baudot code.

I have mentioned the "average" character rate for PSK31 Varicode with English text in the "Error Probabilities" paragraph here:

http://www.w7ay.net/site/Technical/DominoEX/Measurements/index.html

Steam RTTY is 7.5 bits per character at 22 ms (165 ms per character).

Then you need to add the overhead from LTRS and FIGS shifts to RTTY, which could reduce effective rates of a typical contest exchange by some 20%.  It adds two characters to pretty much everybody's callsigns, for example.

When you include LTRS and FIGS, my own call sign comes to 6 Baudot characters instead of 4 visible characters (50% overhead).  Hisami 7L4IOU has it worse, it adds 4 characters to his callsign if he uses USOS, going from 6 visible characters to 10 Baudot characters, for a whopping 66% overhead.

For English text, PSK31 varicode has an average of 6.33 bits per character. At 32 ms per bit, this comes to 203 ms per character.  

Without the need for LTRS and FIGS, *and* if you use lower case characters (upper case characters have longer Varicodes), you are really not doing badly with PSK31 -- as long as you also get rid of the long PSK31 squelch tail that most modems transmit (with no option to truncate).  (This is that beeeeep which you hear at the end of PSK31 transmissions.)

PSK63 is precisely twice as fast as PSK31; the "63" is rounded up from 62.5, while "31" is rounded down from "31.25". 

If you get rid of the squelch tail, PSK63 is definitely faster than Steam RTTY.  But, if you use upper case in a PSK contest, all bets are off.  

73
Chen, W7AY