[UK-CONTEST] CW sending speed on 160m

[Ian White GM3SEK]

Don Beattie wrote:
>Interesting thread !
>
>I agree with Peter that some of the SO2R brigade might be keeping their 
>channel open by frequent CQing.
>
>But the more interesting thing is "what is the optimum speed for DX on 
>160 ?"  I agree with Chris that the "send it slow" policy may well be 
>flawed. How many times have you waited for the missing digit in a call 
>only to find it missing in QRN or (more often) slow QSB ? I think 
>around a 24 wpm rate is probably optimal - not so fast that 
>flutter/auroral echo takes it out, but fast enough to be able to afford 
>call x 2 or even x 3 and have a good chance of copying it.
>
>Of course in the end, the answer will be "whatever works best" but I 
>must admit to frustration in recent contests when weak signals thought 
>that sending slower improved matters - it didn't - it just prolonged 
>the time needed to complete the QSO.
>
>I wonder whether there is a learned treatise on this ? Someone may have 
>done some modelling. I could not find anything via Google.
>

Matching the sending speed to the fading rate has been thought about a 
lot by moonbounce operators... who have plenty of time to think about 
such things. There is still no learned treatise, but some points may 
apply to Top Band, and some can be helpful whenever signals are weak and 
fading.

(There is quite an overlap between moonbouncers and Top Band DXers, by 
the way. If anyone remembers Jan, DL9KR(A), he recently made the first 
DXCC on 70cm, entirely on CW because that was the *only* way to do it 
until recent years. For traditionalists, it still is.)

Signals off the moon are extremely weak so high speeds are a complete 
waste of time. This may not always apply to Top Band of course.

A feature of moonbounce signals is a different short-term fading rate on 
each band. At say 15wpm a short fade would typically take out a whole 
word (or callsign) on 144MHz; but on 432MHz the same fade happens 3x 
faster so it would take out a single letter instead. On 1296MHz the 
fading is 3x faster still and can easily chop a slow dash into two dots, 
turning perfectly good Morse into complete gibberish.

This means that someone hopping from band to band would have to adopt 
three different sending styles - "whatever works best" on each band.

Another variable that hasn't been mentioned is the letter spacing, 
because the spaces carry information too. Standard Morse spacing was 
designed for landline circuits and S9 radio signals, but it isn't 
optimum for extremely weak signals. For example, the optimum way to get 
around the fading on 432MHz is to send individual letters at 20-25wpm so 
they stand a better chance of coming through complete; but also to leave 
very clear gaps between the letters - not too much, just enough to make 
the ar-tic-u-la-tion clearer.

On moonbounce this is always combined with a lot of repetition. What 
works best is to send EXACT repeats of the same short "frame" of 
information, so the other station knows exactly when a missing letter or 
number is coming up. Also leave a small extra gap between frames, which 
helps the receiving operator to re-sync with the pattern.

How much of all that may apply to Top Band contesting is up to the 
individual operator. I can only say helped to give my small moonbounce 
station a reputation for being "easy to work"... and in any kind of 
contesting, that counts for a lot.




-- 

73 from Ian GM3SEK