Return to KLUBNL.PL main page

rsgb_lf_group
[Top] [All Lists]

Re: LF: Confused over Jason coding

To: [email protected]
Subject: Re: LF: Confused over Jason coding
From: "David Willmore" <[email protected]>
Date: Mon, 05 Aug 2002 12:11:58 -0400
In-reply-to: Your message of "Mon, 05 Aug 2002 16:22:16 BST." <[email protected]>
Reply-to: [email protected]
Sender: <[email protected]>
Must have missed something, but in looking at the coding scheme for the
frequency shifts in Jason (as a preliminary to writing some PIC code to
drive a DDS chip for a beacon)  I came across a snag.

The 64 characters are coded 4 bits at a time (allowing for the MSB
identifier) as 16 frequency shifts.  OK so far, 16 shifts = 4 bits.  But,
what happens if it is impossible to get more than 8 bits of shift?   For
instance, say the frequency from the last character ended up at exactly
midway - frequency 8.  Then it is impossible to code any value greater than
8 without going outside the defined frequency limits.

I'm sure this was all explained last year when the mode was being developed
but can't find any reference to it now.  And the Tech Notes that come with
V0.94 don't cover this aspect.

A thought has just occurred while writing this - does the direction of
frequency shift matter, which can then be taken modulo 16, giving 16
possible shifts from 17 tones?  Which means the sideband in use is
sensitive.  Have I answered my own question ?

You 'wrap' around at the top of the 'band'.  So, all frequencies are available
at all times.  So, if you're at (zero based) tone 8 and want to shift +16, you
will end up on tone 8+16 mod 17 or tone 7.  The 'next tone' formula is just
current tone + desired shift mod 17.  Shifts are always thought of as 
'positive',
but the band is just a little mobius strip. :)

Cheers,
David N0YMV




<Prev in Thread] Current Thread [Next in Thread>