Optimum CW waveform recommendation??
 

CW Waveform Characteristics for no clix??

What would be the ideal output rise and decay times for a 25 wpm CW signal
with minimal clicks?

ak

King Zulu wrote:
CW Waveform Characteristics for no clix??

What would be the ideal output rise and decay times for a 25 wpm CW
signal
with minimal clicks?

ak


It depends what you call "ideal." To maximumize signal to noise ratio
by matching the receiver passband to the spectrum of the transmitted
signal, you can figure it out this way:

1. Find the pulsewidth (=T) of a 25 wpm "dit" and decide on a
bandwidth to pass it equal to BW= 1/T.

2. Then determine the rise and fall times from Tr=.35/BW.

3. To convert WPM to pulsewidth you need to know how many dit elements
(baud) there are in a word. Note 1 baud=T. The standard morse code
word is "PARIS" and it has 50 baud so if you send 25wpm you will be
sending 25 x 50 = 1250 baud per minute. So one baud=60/1250 =
..048Sec=T.

4. The bandwidth to pass it would be 1/T = 20.83Hz. (Note, this seems
low.)

5. So the 10%/90% rise and fall times should be Tr=.35/B = .35/20.83 =
16.8mS.

6. If you try to build a circuit to shape your signal to this rise
time, it will probably sound mushy and there will be a lot of
intersymbol interference (ISI). A better compromise is to have some
filtering in the transmitter and some in the receiver. Usually very
little is used in the transmitter to accomodate different code speeds
with one filter. So the bulk of the filtering for signal reception is
in the receiver.

7. Now the purpose of the transmitter filter is only to reduce the
transmitter spectrum so that nearby signals will not be interfered
with. Key clicks are the result of the transmitted spectrum exceeding
the receiver bandwidth (unsymmetrically), causing what is also referred
to as "rabbit ears" in radar detection terminology. The standard
communications voice bandwidth is .3 to 3kHz or 2.7kHz. Assuming this
is twice the baseband spectrum then the baseband spectrum must be
1.35kHz wide.

8. Applying the Tr=.35/B rule now yields Tr=259microseconds. This
will result in no clicks heard in a .3-3kHz SSB filter centered on the
cw carrier. However there could be key clicks heard depending on how
the receiver is tuned.

9. Don't ask me what time it is; I forgot how to make a watch (other
than a digital one..... :o)
73, Paul, AA1LL
Mason, NH
http://www.qsl.net/aa1ll

"AA1LL" wrote in message
oups.com...
King Zulu wrote:
CW Waveform Characteristics for no clix??

What would be the ideal output rise and decay times for a 25 wpm CW
signal with minimal clicks?

.. . .
5. So the 10%/90% rise and fall times should be
Tr=.35/B = .35/20.83 = 16.8mS.
.. . .

Thanks for the complete information, Paul. My next is to get the TR7 to do
that. But I now know where I should be going.

73, ak