[CQ-Contest] CW works better!

[Bill Coleman]

On 7/18/02 11:40, Bill Tippett at btippett at alum.mit.edu wrote:

>        1.  40 and 80 must operate split between Region 2 and
>Regions 1 & 3 for the most part...but that has not been the case
>for 160 which has identical favorable results for CW as on 
>the other two bands.  Of course, I maintain SSB scores on 160
>will actually go up if we segment and DX operates split.  The
>simple reason is that DX will not be buried underneath extremely
>strong local US stations continuously CQ-ing on top of them.  

I've seen this argument so many times, I'm somewhat sick of it.

It is a good technical point. My question is -- why is it only applied to 
SSB? Wouldn't this operation also be beneficial for CW? Of course it 
would. So, why not propose to use CW exclusively in the US from 1950-2000 
kHz and work all DX split?

>        2.  Part of the problem with SSB is that it is a 
>bandwidth hog.  When you try to crowd an equivalent number
>of contesters into the same low band frequencies, the narrow
>bandwidth mode will always win.  The inverse of this is 10-20
>meters where SSB usually wins.

SSB has bandwidth problems on the higher bands, with the possible 
exception of 10 meters.

One important effect on the higher bands is that the presence of skip 
zones tends to limit co-channel interference.

Bottom line, though, CW requires only a percent or so of the bandwidth of 
SSB. Therefore, the signal levels required for effective communications 
are definitely lower.

>On 10 meters, with effectively 
>no bandwidth limit on either mode, SSB wins by about 50% (my CQ 
>WW SSB record is 1.464M versus my CW record of 0.965M). 

I don't think these records are any indication of the inherent properties 
of the mode. On SSB, most likely it is due to the higher availability of 
stations to work than the properties of the mode itself. 

>        Not at all.  It has more to do with the fact that a
>narrow bandwidth mode allows better copy of weak signals
>because the narrower bandwidth allows better rejection of 
>interference, noise, etc.  This is the same reason that digital 
>modes work well in extracting signals from noise.  Programs 
>like WSJT, QRSS/Spectrascan, etc effectively make EXTREMELY narrow 
>bandwidths using DSP that allow copy even below the noise floor
>(of course I personally do not think a computer-to-computer 
>WSJT or QRSS "QSO" means much but that's another topic!)

These modes aren't anything alike. WSJT is 441 baud, which is actually a 
rather high signalling rate compared to CW. WSJT is designed for meter 
scatter work, and therefore has to transfer information at a high rate. 
(it also uses multi-bit FSK, to avoid some of the phase distortions 
present in the meteor pings)

So, WSJT is not extremely narrow. It is wider than typical RTTY or 300 
baud packet, even.

Point is, each of these modulation techniques is designed to meet certain 
channel goals. WSJT works much more effectively than high-speed CW. (high 
speed here meaning 100-800 wpm!)


Bill Coleman, AA4LR, PP-ASEL        Mail: aa4lr at arrl.net
Quote: "Not within a thousand years will man ever fly!"
            -- Wilbur Wright, 1901