[CW] QSK?

[N2EY@aol.com]

In a message dated 3/15/03 9:13:04 AM Eastern Standard Time, w2agn@w2agn.net 
writes:

> I guess what is considered "true" QSK varies. When I think of QSK, I think
>  of Ten Tec, Elecraft, etc. NO CLATTERING relays! 

Maybe we should first agree to what "QSK" and "full break-in" really mean!

For me, "QSK" and "full break-in" mean exactly the same thing: The sending 
station can hear what's happening on the receive frequency whenever the key 
is not down, up to the maximum speed of the rig being used. The quality of 
the transmitted signal should not be affected in any way by the QSK system.  

How QSK is implemented doesn't matter, it's the end result that is important. 
There are highspeed relays which will do the job, as well as PIN diodes and 
tube-type TR switches. 

QSK isn't new - hams were doing it at least as far back as the '30s, by means 
of a separate receiving antenna and a high speed relay that would unmute the 
receiver on key-up and key the transmitter on key-down. But I cannot recall a 
single manufactured ham transceiver or transmitter-receiver pair before the 
HW-16 that would do QSK right out of the box without modifications or 
add-ons. Many of the EFJ rigs (Ranger, Navigator, Valiant) had decent 
"differential" keying, but you still needed an electronic TR switch and some 
means of muting the receiver.

Many current rigs do QSK of one form or another, but in some cases the 
receiver recovery and/or keyed waveform leave something to be desired, 
particularly at high speeds. QST product reviews show this quite graphically.

>  What many seem to consider
>  QSK is just VOX type semi break in, with the delay set very short. 

Agreed - but that's not QSK, it's just TATOO (Transmitter Automatically Turns 
On & Off).
 
> This
>  would drive me nuts (nutsier?) very quickly. Also, I hate to think how long
>  those relays will last under that kind of abuse. 

Yup.

A bit of math....

The standard way of computing plain-language code speed that I learned was to 
see how many times the word "PARIS" could be sent in one minute, with correct 
spacing. Turns out that the word "PARIS" with a word space is exactly 50 dit 
times long, where a dit time is the time the key is down for a dit. 

So at 60 wpm, there are 3000 dit times in a minute, or 50 dit times in a 
second. A 60 wpm dit is therefore 20 milliseconds long. So the TR transition 
time should not be more than a few milliseconds to have really good high 
speed QSK. 

73 de Jim, N2EY