Content-Transfer-Encoding: 8bit
Message-ID: <3.0.1.16.20010320155747.2b1f164c@mail.cc.kuleuven.ac.be>
Date: Tue, 20 Mar 2001 15:57:47
To: rsgb_lf_group@blacksheep.org
From: "Rik Strobbe" <rik.strobbe@fys.kuleuven.ac.be>
Subject: LF: DFCW
In-reply-to: <001d01c0b146$a6dcf720$9628893e@g3aqc>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

At 14:03 20/03/01 -0000, G3AQC wrote:
>I experimented last night with slightly shorter dots- 25secs rather than my 
>usual 30. This reduced my total message time (for three repeats of G3AQC)
from 
>56mins to44mins,including CW ident. I want to get shorter times without 
>reducing readability and both Rik and Alberto are helping with this. I think 
>DFCW is the most effective of the simpler techniques and well worth 
>developing. Last nights results are very encouraging.

Hello group,

Laurie had been busy in 'time optimizing' DFCW, mainly by reducing the
'idle' gaps. At the very slow speeds used these days for TA tests one can
indeed win a fair ammount of time by avoiding idle times. 
The main function of the DFCW gaps is to improve the readability of DFCW in
case of a series of dashes or dots. Or otherwise said, it is some kind of
'time synchronisation'.

This 'time synchonisation' has proven to be usefull, mainly for the reason
that at the RX side you otherwise had to guess when a dot or dash started
and stopped. If the RX would know at what times the TX was keying this gaps
wouldn't be nessecary ...

At the dot lengths used in recent TA tests (30 to 120 seconds) it shouldn't
be so difficult to get sufficient 'time synchronisation' between TX and RX :

1. Using DCF77 (or similar) and the appropriate software (such as provided
by DL4YHF) one can set the PC clock with a 1/10 second accuracy. Even by
setting the PC clock by hand an accuracy of better than 1 second can be
achieved. So the PC's at both sides (TX and RX) can have their clock
'synchronized' at 1 second or better.

2. QRS (v3.05) allows you start keying on exact multiples of 30 seconds (or
1 minute, 2 minutes, 5 minutes etc...). Of course this only makes sense if
you set the dotlength also at 30 seconds (or 1 minute etc..). 
So the TX can be 'time synchronized'.

3. If Alberto could adapt ARGO and let the 'time markers' also appear on
exact multiples of 30 seconds (or 1 minute etc..) and eventually replace
the 'time ticks' at the bottom by thin vertical lines (over the entire
window) then the RX would also have the required 'time synchronisation' :
'key up' or 'key down' of the TX would be at the same time when a 'time
marker' appears on the ARGO screen.

Result : no more need for time gaps (what should be a 20-25% time saving).

73, Rik  ON7YD