[time-nuts] nubie querie

[Magnus Danielson]

Hal Murray wrote:
> Concerning my query about what's good enough to count as a contact...
> 
>> We've done Moonbounce with 3mW (Hobart - Dwingeloo) in JT65 - but a
>> 26m and a 25m dish is stretching 'amateur' a bit again. 
> 
> Googling for JT65 finds a nice paper:
>   http://www.arrl.org/FandES/field/regulations/techchar/18JT65.pdf
>   The JT65 Communications Protocol
>   Joe Taylor, K1JT
> 
> It's a fun read.  17 pages.
> 
> The basic idea seems to be that amateurs (hams) have to exchange station IDs. 
>  That's more than a few bits, but not a huge number.
> 
> JT65 is a compact protocol for doing that in a (very) weak signal 
> environment.  Their packet format is 72 bits expanded to 378 by forward error 
> correcting.  On top of that, they use half of the time for a synchronizing 
> signal so the receiver can find the transmitter's time and frequency.  Each 
> 72 bit packet takes 1 minute to send.
> 
> Their modulation scheme is 1 of 65 tones.  6 bits per baud.  The extra tone 
> is the synchronizing signal.

6 bits per symbol. 1 baud is 1 symbol per second.

Happy to see someone using baud, just unhappy about seeing it being used 
incorrectly.

An amusing error was found in one of our early datasheets. For some 
reason they wanted to tell the signal rate, so they said 1,0625 GBaud/s. 
I found that very amusing to have symbol acceleration... it gets faster 
every second!!! Just keep a fixed bit/symbol ratio and you have a hell 
of a product. Later in life it will transport all of universe into a 
black hole.

Cheers,
Magnus