[RTTY] 2Tone Chirp

[David G3YYD]

As you can see I have joined this reflector.

The start up sequence of 2Tone is just as Chen, W7AY, describes. This is 
intentional as I will explain.

When receiving using 2Tone in normal mode it asses the envelope value of 
both the mark and space tones. It uses this along with its noise 
assessment to generate a threshold value for each tone. These threshold 
values are used as part of the way 2Tone works out whether mark or space 
is being received at any instance in time.

At the moment a new transmission starts up 2Tone receive envelope and 
noise values will be incorrect as they will either be a value of the 
previous transmission or the value of the noise/QRM being received. In 
order to optimise 2Tone decode there is a need to send a space tone and 
a mark tone before characters are sent.

Being that mark tone is required to ensure the asynchronous decoder is 
ready for the start bit of the first character then the logical sequence 
is space tone for one character time; setting the space threshold value 
and mark noise value. This is then followed by mark tone for one 
character time setting the mark threshold value and space noise value 
and resetting the asynchronous decoder to wait for  first start bit (one 
bit time of space tone).

Most (all?) other means of sending RTTY normally start with a long, in 
many cases unnecessary long, mark tone and then sends the first 
character. This means the space tone threshold will be in error. But 
2Tone is a bit more sophisticated. It deliberately delays assessment by 
one character time so it can also use information from the second sent 
character to establish a space (and improved mark) threshold value for 
the first character. So not all is lost but for best results the 2Tone 
sequence is that fastest you can go and get the best decode accuracy.

As Chen says 2Tone transmission has a slow amplitude ramp up at the 
start and ramp down at the end of transmission so there are no clicks to 
cause QRM to other band users. Very few AFSK implementations  do this 
and no rig based FSK does it causing unnecessary QRM on adjacent channels.


73 David G3YYD