[RTTY] ARRL attack on current RTTY users

[Kok Chen]

On Nov 22, 2013, at 7:40 AM, Ben Antanaitis - WB2RHM wrote:

> IF it ain't broke, don't fix it. Put the super speed/data volume 2.8KHz signals in the band segments where 2.8KHz BW is already the 'standard'.

Ben,

While I too believe that the ARRL petition has no merit, if the petition passes FCC's rule making, the situation is really not that much more dire than it already is today.  Unless of course if sub-bands stop being segregated by modes and bandwidths.  

Personally, I really don't care if a 3 kHz wide signal is QRMing a 2.2 kHz wide signal in the subbands that 2.2 kHz signals already use.  My concern is that a 3 kHz signal is allowed to intermingle with a 300 Hz signal.  

Today, we already have 2.2 kHz signals intermingling with 300 Hz wide signals.  As happens often nowadays, when you answer an RTTY CQ, a loud Bzzzzt comes back at you instead.

Not all the Bzzzzt are 2.2 kHz wide either, some of them are only 500 Hz wide, but at a high symbol rate; that is why they sound like that -- you hear a similar sound with fast RTTY or PSK125, for example.

Perhaps I can use this opportunity to explain the symbol rate stuff (the gist of ARRL's petition) to the general audience of the reflector,  and how it relates to HF propagation.

Symbol rate (a.k.a. baud rate) is NOT directly related to occupied bandwidth.  Please don't make that mistake in your RM comments.  Your comments might otherwise be diminished if the ARRL lawyer convinces the Commissioners you don't understand the technical issues.

Again, let me repeat -- Symbol Rate is not directly related to bandwidth.

In general, it is directly related to bandwidth only when 1 symbol is equal to 1 bit -- as in the case of BPSK31 and amateur RTTY.

(For lack of better term, I use "amateur RTTY" to mean "2-tone 170 Hz shift FSK at 45.45 baud, with Baudot encoding" -- that, or my more affectionate term "steam RTTY" for it :-).

"Symbol Rate" is just a fancy way of saying how many times you switch modulation in each second.  And it is expressed in units of "baud" (symbols per second).

MFSK16, DominoEX and Olivia for example uses many tones (a weird number like 18 tones for DominoEX, even) to achieve better performance.  These signals occupy 500 Hz or more (but not for the sake of higher data rates).  

If you look at DominoEX's and MFSK16's philosophy (both created by ZL1BPU) they are designed with the understanding of HF channels by purposely making the symbol rate very *slow.*  And to make up for it, so it can achieve a practical typing rate, more tones are used (thus, more symbols), and that therefore uses more bandwidth.

The reason why a lower symbol rate is desirable is that when HF conditions become poor, the RF signal becomes smeared in time (visualize a rectangular pulse whose rise and fall times no longer have sharp skirts).  This smearing comes both from multipath and (more so) from what is called Doppler Spreading by the ionosphere.  

To overcome the smearing, you would want your symbols to change at a rate that is *slower* than the smearing.  Guys like ZL1BPU and G3PLX (the creator of Amtor and PSK31) really do understand the ionospheric effects.

This is contrary, for example to the ARRL petition claim that higher symbol rate somehow is more "modern" and by implication, more useful.    It is neither.

Thus, in the case of MFSK16, DominoEX and Olivia, we have the case of actually quite slow symbol rates (much slower than RTTY), but the signals have bandwidths a that are wider than RTTY, and they also, unsurprisingly, beats RTTY through poorer propagation conditions (RTTY standards were created before HF propagation became more understood).  

The modes above are the correct way of applying symbol rate tradeoffs under HF propagation (i.e., go to lower baud rates).  And that is why I have been saying that there is no *need* to have higher symbol rates.  

You can already, today, keep the symbol rate at 45.45 baud and create a 128 tone FSK signal that occupies many kHz of bandwidth.  There are no rules against that (since the symbol rate is under 1000 baud).  You are only limited by bandwidth rules from using something like that.

The relaxation of Symbol Rates only serves to allow certain modulation methods, which are not legal today, that happens to also be very wide.

73
Chen, W7AY